Spring框架参考文档 Spring Framework Reference Documentation

Authors

Rod Johnson , Juergen Hoeller , Keith Donald , Colin Sampaleanu , Rob Harrop , Thomas Risberg , Alef Arendsen , Darren Davison , Dmitriy Kopylenko , Mark Pollack ,Thierry Templier , Erwin Vervaet , Portia Tung , Ben Hale , Adrian Colyer , John Lewis , Costin Leau , Mark Fisher , Sam Brannen , Ramnivas Laddad , Arjen Poutsma ,Chris Beams , Tareq Abedrabbo , Andy Clement , Dave Syer , Oliver Gierke , Rossen Stoyanchev , Phillip Webb , Rob Winch , Brian Clozel , Stephane Nicoll , SebastienDeleuze

4.3.2.RELEASE

Copyright © 2004-2016

Copies of this document may be made for your own use and for distribution to others, provided that you do not charge any fee for such copies and further provided that each copy contains this Copyright Notice, whether distributed in print or electronically.

Table of Contents

I. Overview of Spring Framework
1. Getting Started with Spring
2. Introduction to the Spring Framework
2.1. Dependency Injection and Inversion of Control
2.2. Modules
2.2.1. Core Container
2.2.2. AOP and Instrumentation
2.2.3. Messaging
2.2.4. Data Access/Integration
2.2.5. Web
2.2.6. Test
2.3. Usage scenarios
2.3.1. Dependency Management and Naming Conventions
Spring Dependencies and Depending on Spring
Maven Dependency Management
Maven "Bill Of Materials" Dependency
Gradle Dependency Management
Ivy Dependency Management
Distribution Zip Files
2.3.2. Logging
Not Using Commons Logging
Using SLF4J
Using Log4J
II. What’s New in Spring Framework 4.x
3. New Features and Enhancements in Spring Framework 4.0
3.1. Improved Getting Started Experience
3.2. Removed Deprecated Packages and Methods
3.3. Java 8 (as well as 6 and 7)
3.4. Java EE 6 and 7
3.5. Groovy Bean Definition DSL
3.6. Core Container Improvements
3.7. General Web Improvements
3.8. WebSocket, SockJS, and STOMP Messaging
3.9. Testing Improvements
4. New Features and Enhancements in Spring Framework 4.1
4.1. JMS Improvements
4.2. Caching Improvements
4.3. Web Improvements
4.4. WebSocket Messaging Improvements
4.5. Testing Improvements
5. New Features and Enhancements in Spring Framework 4.2
5.1. Core Container Improvements
5.2. Data Access Improvements
5.3. JMS Improvements
5.4. Web Improvements
5.5. WebSocket Messaging Improvements
5.6. Testing Improvements
6. New Features and Enhancements in Spring Framework 4.3
6.1. Core Container Improvements
6.2. Data Access Improvements
6.3. Caching Improvements
6.4. JMS Improvements
6.5. Web Improvements
6.6. WebSocket Messaging Improvements
6.7. Testing Improvements
6.8. Support for new library and server generations
III. Core Technologies
7. The IoC container
7.1. Introduction to the Spring IoC container and beans
7.2. Container overview
7.2.1. Configuration metadata
7.2.2. Instantiating a container
Composing XML-based configuration metadata
7.2.3. Using the container
7.3. Bean overview
7.3.1. Naming beans
Aliasing a bean outside the bean definition
7.3.2. Instantiating beans
Instantiation with a constructor
Instantiation with a static factory method
Instantiation using an instance factory method
7.4. Dependencies
7.4.1. Dependency Injection
Constructor-based dependency injection
Setter-based dependency injection
Dependency resolution process
Examples of dependency injection
7.4.2. Dependencies and configuration in detail
Straight values (primitives, Strings, and so on)
References to other beans (collaborators)
Inner beans
Collections
Null and empty string values
XML shortcut with the p-namespace
XML shortcut with the c-namespace
Compound property names
7.4.3. Using depends-on
7.4.4. Lazy-initialized beans
7.4.5. Autowiring collaborators
Limitations and disadvantages of autowiring
Excluding a bean from autowiring
7.4.6. Method injection
Lookup method injection
Arbitrary method replacement
7.5. Bean scopes
7.5.1. The singleton scope
7.5.2. The prototype scope
7.5.3. Singleton beans with prototype-bean dependencies
7.5.4. Request, session, global session, application, and WebSocket scopes
Initial web configuration
Request scope
Session scope
Global session scope
Application scope
Scoped beans as dependencies
7.5.5. Custom scopes
Creating a custom scope
Using a custom scope
7.6. Customizing the nature of a bean
7.6.1. Lifecycle callbacks
Initialization callbacks
Destruction callbacks
Default initialization and destroy methods
Combining lifecycle mechanisms
Startup and shutdown callbacks
Shutting down the Spring IoC container gracefully in non-web applications
7.6.2. ApplicationContextAware and BeanNameAware
7.6.3. Other Aware interfaces
7.7. Bean definition inheritance
7.8. Container Extension Points
7.8.1. Customizing beans using a BeanPostProcessor
Example: Hello World, BeanPostProcessor-style
Example: The RequiredAnnotationBeanPostProcessor
7.8.2. Customizing configuration metadata with a BeanFactoryPostProcessor
Example: the Class name substitution PropertyPlaceholderConfigurer
Example: the PropertyOverrideConfigurer
7.8.3. Customizing instantiation logic with a FactoryBean
7.9. Annotation-based container configuration
7.9.1. @Required
7.9.2. @Autowired
7.9.3. Fine-tuning annotation-based autowiring with @Primary
7.9.4. Fine-tuning annotation-based autowiring with qualifiers
7.9.5. Using generics as autowiring qualifiers
7.9.6. CustomAutowireConfigurer
7.9.7. @Resource
7.9.8. @PostConstruct and @PreDestroy
7.10. Classpath scanning and managed components
7.10.1. @Component and further stereotype annotations
7.10.2. Meta-annotations
7.10.3. Automatically detecting classes and registering bean definitions
7.10.4. Using filters to customize scanning
7.10.5. Defining bean metadata within components
7.10.6. Naming autodetected components
7.10.7. Providing a scope for autodetected components
7.10.8. Providing qualifier metadata with annotations
7.11. Using JSR 330 Standard Annotations
7.11.1. Dependency Injection with @Inject and @Named
7.11.2. @Named: a standard equivalent to the @Component annotation
7.11.3. Limitations of JSR-330 standard annotations
7.12. Java-based container configuration
7.12.1. Basic concepts: @Bean and @Configuration
7.12.2. Instantiating the Spring container using AnnotationConfigApplicationContext
Simple construction
Building the container programmatically using register(Class<?>…​)
Enabling component scanning with scan(String…​)
Support for web applications with AnnotationConfigWebApplicationContext
7.12.3. Using the @Bean annotation
Declaring a bean
Bean dependencies
Receiving lifecycle callbacks
Specifying bean scope
Customizing bean naming
Bean aliasing
Bean description
7.12.4. Using the @Configuration annotation
Injecting inter-bean dependencies
Lookup method injection
Further information about how Java-based configuration works internally
7.12.5. Composing Java-based configurations
Using the @Import annotation
Conditionally include @Configuration classes or @Bean methods
Combining Java and XML configuration
7.13. Environment abstraction
7.13.1. Bean definition profiles
@Profile
7.13.2. XML bean definition profiles
Activating a profile
Default profile
7.13.3. PropertySource abstraction
7.13.4. @PropertySource
7.13.5. Placeholder resolution in statements
7.14. Registering a LoadTimeWeaver
7.15. Additional Capabilities of the ApplicationContext
7.15.1. Internationalization using MessageSource
7.15.2. Standard and Custom Events
Annotation-based Event Listeners
Asynchronous Listeners
Ordering Listeners
Generic Events
7.15.3. Convenient access to low-level resources
7.15.4. Convenient ApplicationContext instantiation for web applications
7.15.5. Deploying a Spring ApplicationContext as a Java EE RAR file
7.16. The BeanFactory
7.16.1. BeanFactory or ApplicationContext?
7.16.2. Glue code and the evil singleton
8. Resources
8.1. Introduction
8.2. The Resource interface
8.3. Built-in Resource implementations
8.3.1. UrlResource
8.3.2. ClassPathResource
8.3.3. FileSystemResource
8.3.4. ServletContextResource
8.3.5. InputStreamResource
8.3.6. ByteArrayResource
8.4. The ResourceLoader
8.5. The ResourceLoaderAware interface
8.6. Resources as dependencies
8.7. Application contexts and Resource paths
8.7.1. Constructing application contexts
Constructing ClassPathXmlApplicationContext instances - shortcuts
8.7.2. Wildcards in application context constructor resource paths
Ant-style Patterns
The Classpath*: portability classpath*: prefix
Other notes relating to wildcards
8.7.3. FileSystemResource caveats
9. Validation, Data Binding, and Type Conversion
9.1. Introduction
9.2. Validation using Spring’s Validator interface
9.3. Resolving codes to error messages
9.4. Bean manipulation and the BeanWrapper
9.4.1. Setting and getting basic and nested properties
9.4.2. Built-in PropertyEditor implementations
Registering additional custom PropertyEditors
9.5. Spring Type Conversion
9.5.1. Converter SPI
9.5.2. ConverterFactory
9.5.3. GenericConverter
ConditionalGenericConverter
9.5.4. ConversionService API
9.5.5. Configuring a ConversionService
9.5.6. Using a ConversionService programmatically
9.6. Spring Field Formatting
9.6.1. Formatter SPI
9.6.2. Annotation-driven Formatting
Format Annotation API
9.6.3. FormatterRegistry SPI
9.6.4. FormatterRegistrar SPI
9.6.5. Configuring Formatting in Spring MVC
9.7. Configuring a global date & time format
9.8. Spring Validation
9.8.1. Overview of the JSR-303 Bean Validation API
9.8.2. Configuring a Bean Validation Provider
Injecting a Validator
Configuring Custom Constraints
Spring-driven Method Validation
Additional Configuration Options
9.8.3. Configuring a DataBinder
9.8.4. Spring MVC 3 Validation
10. Spring Expression Language (SpEL)
10.1. Introduction
10.2. Feature Overview
10.3. Expression Evaluation using Spring’s Expression Interface
10.3.1. The EvaluationContext interface
Type Conversion
10.3.2. Parser configuration
10.3.3. SpEL compilation
Compiler configuration
Compiler limitations
10.4. Expression support for defining bean definitions
10.4.1. XML based configuration
10.4.2. Annotation-based configuration
10.5. Language Reference
10.5.1. Literal expressions
10.5.2. Properties, Arrays, Lists, Maps, Indexers
10.5.3. Inline lists
10.5.4. Inline Maps
10.5.5. Array construction
10.5.6. Methods
10.5.7. Operators
Relational operators
Logical operators
Mathematical operators
10.5.8. Assignment
10.5.9. Types
10.5.10. Constructors
10.5.11. Variables
The #this and #root variables
10.5.12. Functions
10.5.13. Bean references
10.5.14. Ternary Operator (If-Then-Else)
10.5.15. The Elvis Operator
10.5.16. Safe Navigation operator
10.5.17. Collection Selection
10.5.18. Collection Projection
10.5.19. Expression templating
10.6. Classes used in the examples
11. Aspect Oriented Programming with Spring
11.1. Introduction
11.1.1. AOP concepts
11.1.2. Spring AOP capabilities and goals
11.1.3. AOP Proxies
11.2. @AspectJ support
11.2.1. Enabling @AspectJ Support
Enabling @AspectJ Support with Java configuration
Enabling @AspectJ Support with XML configuration
11.2.2. Declaring an aspect
11.2.3. Declaring a pointcut
Supported Pointcut Designators
Combining pointcut expressions
Sharing common pointcut definitions
Examples
Writing good pointcuts
11.2.4. Declaring advice
Before advice
After returning advice
After throwing advice
After (finally) advice
Around advice
Advice parameters
Advice ordering
11.2.5. Introductions
11.2.6. Aspect instantiation models
11.2.7. Example
11.3. Schema-based AOP support
11.3.1. Declaring an aspect
11.3.2. Declaring a pointcut
11.3.3. Declaring advice
Before advice
After returning advice
After throwing advice
After (finally) advice
Around advice
Advice parameters
Advice ordering
11.3.4. Introductions
11.3.5. Aspect instantiation models
11.3.6. Advisors
11.3.7. Example
11.4. Choosing which AOP declaration style to use
11.4.1. Spring AOP or full AspectJ?
11.4.2. @AspectJ or XML for Spring AOP?
11.5. Mixing aspect types
11.6. Proxying mechanisms
11.6.1. Understanding AOP proxies
11.7. Programmatic creation of @AspectJ Proxies
11.8. Using AspectJ with Spring applications
11.8.1. Using AspectJ to dependency inject domain objects with Spring
Unit testing @Configurable objects
Working with multiple application contexts
11.8.2. Other Spring aspects for AspectJ
11.8.3. Configuring AspectJ aspects using Spring IoC
11.8.4. Load-time weaving with AspectJ in the Spring Framework
A first example
Aspects
'META-INF/aop.xml'
Required libraries (JARS)
Spring configuration
Environment-specific configuration
11.9. Further Resources
12. Spring AOP APIs
12.1. Introduction
12.2. Pointcut API in Spring
12.2.1. Concepts
12.2.2. Operations on pointcuts
12.2.3. AspectJ expression pointcuts
12.2.4. Convenience pointcut implementations
Static pointcuts
Dynamic pointcuts
12.2.5. Pointcut superclasses
12.2.6. Custom pointcuts
12.3. Advice API in Spring
12.3.1. Advice lifecycles
12.3.2. Advice types in Spring
Interception around advice
Before advice
Throws advice
After Returning advice
Introduction advice
12.4. Advisor API in Spring
12.5. Using the ProxyFactoryBean to create AOP proxies
12.5.1. Basics
12.5.2. JavaBean properties
12.5.3. JDK- and CGLIB-based proxies
12.5.4. Proxying interfaces
12.5.5. Proxying classes
12.5.6. Using 'global' advisors
12.6. Concise proxy definitions
12.7. Creating AOP proxies programmatically with the ProxyFactory
12.8. Manipulating advised objects
12.9. Using the "auto-proxy" facility
12.9.1. Autoproxy bean definitions
BeanNameAutoProxyCreator
DefaultAdvisorAutoProxyCreator
AbstractAdvisorAutoProxyCreator
12.9.2. Using metadata-driven auto-proxying
12.10. Using TargetSources
12.10.1. Hot swappable target sources
12.10.2. Pooling target sources
12.10.3. Prototype target sources
12.10.4. ThreadLocal target sources
12.11. Defining new Advice types
12.12. Further resources
IV. Testing
13. Introduction to Spring Testing
14. Unit Testing
14.1. Mock Objects
14.1.1. Environment
14.1.2. JNDI
14.1.3. Servlet API
14.1.4. Portlet API
14.2. Unit Testing support Classes
14.2.1. General testing utilities
14.2.2. Spring MVC
15. Integration Testing
15.1. Overview
15.2. Goals of Integration Testing
15.2.1. Context management and caching
15.2.2. Dependency Injection of test fixtures
15.2.3. Transaction management
15.2.4. Support classes for integration testing
15.3. JDBC Testing Support
15.4. Annotations
15.4.1. Spring Testing Annotations
@BootstrapWith
@ContextConfiguration
@WebAppConfiguration
@ContextHierarchy
@ActiveProfiles
@TestPropertySource
@DirtiesContext
@TestExecutionListeners
@Commit
@Rollback
@BeforeTransaction
@AfterTransaction
@Sql
@SqlConfig
@SqlGroup
15.4.2. Standard Annotation Support
15.4.3. Spring JUnit 4 Testing Annotations
@IfProfileValue
@ProfileValueSourceConfiguration
@Timed
@Repeat
15.4.4. Meta-Annotation Support for Testing
15.5. Spring TestContext Framework
15.5.1. Key abstractions
TestContext
TestContextManager
TestExecutionListener
Context Loaders
15.5.2. Bootstrapping the TestContext framework
15.5.3. TestExecutionListener configuration
Registering custom TestExecutionListeners
Automatic discovery of default TestExecutionListeners
Ordering TestExecutionListeners
Merging TestExecutionListeners
15.5.4. Context management
Context configuration with XML resources
Context configuration with Groovy scripts
Context configuration with annotated classes
Mixing XML, Groovy scripts, and annotated classes
Context configuration with context initializers
Context configuration inheritance
Context configuration with environment profiles
Context configuration with test property sources
Loading a WebApplicationContext
Context caching
Context hierarchies
15.5.5. Dependency injection of test fixtures
15.5.6. Testing request and session scoped beans
15.5.7. Transaction management
Test-managed transactions
Enabling and disabling transactions
Transaction rollback and commit behavior
Programmatic transaction management
Executing code outside of a transaction
Configuring a transaction manager
Demonstration of all transaction-related annotations
15.5.8. Executing SQL scripts
Executing SQL scripts programmatically
Executing SQL scripts declaratively with @Sql
15.5.9. TestContext Framework support classes
Spring JUnit 4 Runner
Spring JUnit 4 Rules
JUnit 4 support classes
TestNG support classes
15.6. Spring MVC Test Framework
15.6.1. Server-Side Tests
Static Imports
Setup Options
Performing Requests
Defining Expectations
Filter Registrations
Differences between Out-of-Container and End-to-End Integration Tests
Further Server-Side Test Examples
15.6.2. HtmlUnit Integration
Why HtmlUnit Integration?
MockMvc and HtmlUnit
MockMvc and WebDriver
MockMvc and Geb
15.6.3. Client-Side REST Tests
Static Imports
Further Examples of Client-side REST Tests
15.7. PetClinic Example
16. Further Resources
V. Data Access
17. Transaction Management
17.1. Introduction to Spring Framework transaction management
17.2. Advantages of the Spring Framework’s transaction support model
17.2.1. Global transactions
17.2.2. Local transactions
17.2.3. Spring Framework’s consistent programming model
17.3. Understanding the Spring Framework transaction abstraction
17.4. Synchronizing resources with transactions
17.4.1. High-level synchronization approach
17.4.2. Low-level synchronization approach
17.4.3. TransactionAwareDataSourceProxy
17.5. Declarative transaction management
17.5.1. Understanding the Spring Framework’s declarative transaction implementation
17.5.2. Example of declarative transaction implementation
17.5.3. Rolling back a declarative transaction
17.5.4. Configuring different transactional semantics for different beans
17.5.5. <tx:advice/> settings
17.5.6. Using @Transactional
@Transactional settings
Multiple Transaction Managers with @Transactional
Custom shortcut annotations
17.5.7. Transaction propagation
Required
RequiresNew
Nested
17.5.8. Advising transactional operations
17.5.9. Using @Transactional with AspectJ
17.6. Programmatic transaction management
17.6.1. Using the TransactionTemplate
Specifying transaction settings
17.6.2. Using the PlatformTransactionManager
17.7. Choosing between programmatic and declarative transaction management
17.8. Transaction bound event
17.9. Application server-specific integration
17.9.1. IBM WebSphere
17.9.2. Oracle WebLogic Server
17.10. Solutions to common problems
17.10.1. Use of the wrong transaction manager for a specific DataSource
17.11. Further Resources
18. DAO support
18.1. Introduction
18.2. Consistent exception hierarchy
18.3. Annotations used for configuring DAO or Repository classes
19. Data access with JDBC
19.1. Introduction to Spring Framework JDBC
19.1.1. Choosing an approach for JDBC database access
19.1.2. Package hierarchy
19.2. Using the JDBC core classes to control basic JDBC processing and error handling
19.2.1. JdbcTemplate
Examples of JdbcTemplate class usage
JdbcTemplate best practices
19.2.2. NamedParameterJdbcTemplate
19.2.3. SQLExceptionTranslator
19.2.4. Executing statements
19.2.5. Running queries
19.2.6. Updating the database
19.2.7. Retrieving auto-generated keys
19.3. Controlling database connections
19.3.1. DataSource
19.3.2. DataSourceUtils
19.3.3. SmartDataSource
19.3.4. AbstractDataSource
19.3.5. SingleConnectionDataSource
19.3.6. DriverManagerDataSource
19.3.7. TransactionAwareDataSourceProxy
19.3.8. DataSourceTransactionManager
19.3.9. NativeJdbcExtractor
19.4. JDBC batch operations
19.4.1. Basic batch operations with the JdbcTemplate
19.4.2. Batch operations with a List of objects
19.4.3. Batch operations with multiple batches
19.5. Simplifying JDBC operations with the SimpleJdbc classes
19.5.1. Inserting data using SimpleJdbcInsert
19.5.2. Retrieving auto-generated keys using SimpleJdbcInsert
19.5.3. Specifying columns for a SimpleJdbcInsert
19.5.4. Using SqlParameterSource to provide parameter values
19.5.5. Calling a stored procedure with SimpleJdbcCall
19.5.6. Explicitly declaring parameters to use for a SimpleJdbcCall
19.5.7. How to define SqlParameters
19.5.8. Calling a stored function using SimpleJdbcCall
19.5.9. Returning ResultSet/REF Cursor from a SimpleJdbcCall
19.6. Modeling JDBC operations as Java objects
19.6.1. SqlQuery
19.6.2. MappingSqlQuery
19.6.3. SqlUpdate
19.6.4. StoredProcedure
19.7. Common problems with parameter and data value handling
19.7.1. Providing SQL type information for parameters
19.7.2. Handling BLOB and CLOB objects
19.7.3. Passing in lists of values for IN clause
19.7.4. Handling complex types for stored procedure calls
19.8. Embedded database support
19.8.1. Why use an embedded database?
19.8.2. Creating an embedded database using Spring XML
19.8.3. Creating an embedded database programmatically
19.8.4. Selecting the embedded database type
Using HSQL
Using H2
Using Derby
19.8.5. Testing data access logic with an embedded database
19.8.6. Generating unique names for embedded databases
19.8.7. Extending the embedded database support
19.9. Initializing a DataSource
19.9.1. Initializing a database using Spring XML
Initialization of other components that depend on the database
20. Object Relational Mapping (ORM) Data Access
20.1. Introduction to ORM with Spring
20.2. General ORM integration considerations
20.2.1. Resource and transaction management
20.2.2. Exception translation
20.3. Hibernate
20.3.1. SessionFactory setup in a Spring container
20.3.2. Implementing DAOs based on plain Hibernate API
20.3.3. Declarative transaction demarcation
20.3.4. Programmatic transaction demarcation
20.3.5. Transaction management strategies
20.3.6. Comparing container-managed and locally defined resources
20.3.7. Spurious application server warnings with Hibernate
20.4. JDO
20.4.1. PersistenceManagerFactory setup
20.4.2. Implementing DAOs based on the plain JDO API
20.4.3. Transaction management
20.4.4. JdoDialect
20.5. JPA
20.5.1. Three options for JPA setup in a Spring environment
LocalEntityManagerFactoryBean
Obtaining an EntityManagerFactory from JNDI
LocalContainerEntityManagerFactoryBean
Dealing with multiple persistence units
20.5.2. Implementing DAOs based on plain JPA
20.5.3. Transaction Management
20.5.4. JpaDialect
21. Marshalling XML using O/X Mappers
21.1. Introduction
21.1.1. Ease of configuration
21.1.2. Consistent Interfaces
21.1.3. Consistent Exception Hierarchy
21.2. Marshaller and Unmarshaller
21.2.1. Marshaller
21.2.2. Unmarshaller
21.2.3. XmlMappingException
21.3. Using Marshaller and Unmarshaller
21.4. XML Schema-based Configuration
21.5. JAXB
21.5.1. Jaxb2Marshaller
XML Schema-based Configuration
21.6. Castor
21.6.1. CastorMarshaller
21.6.2. Mapping
XML Schema-based Configuration
21.7. XMLBeans
21.7.1. XmlBeansMarshaller
XML Schema-based Configuration
21.8. JiBX
21.8.1. JibxMarshaller
XML Schema-based Configuration
21.9. XStream
21.9.1. XStreamMarshaller
VI. The Web
22. Web MVC framework
22.1. Introduction to Spring Web MVC framework
22.1.1. Features of Spring Web MVC
22.1.2. Pluggability of other MVC implementations
22.2. The DispatcherServlet
22.2.1. Special Bean Types In the WebApplicationContext
22.2.2. Default DispatcherServlet Configuration
22.2.3. DispatcherServlet Processing Sequence
22.3. Implementing Controllers
22.3.1. Defining a controller with @Controller
22.3.2. Mapping Requests With @RequestMapping
Composed @RequestMapping Variants
@Controller and AOP Proxying
New Support Classes for @RequestMapping methods in Spring MVC 3.1
URI Template Patterns
URI Template Patterns with Regular Expressions
Path Patterns
Path Pattern Comparison
Path Patterns with Placeholders
Suffix Pattern Matching
Suffix Pattern Matching and RFD
Matrix Variables
Consumable Media Types
Producible Media Types
Request Parameters and Header Values
HTTP HEAD and HTTP OPTIONS
22.3.3. Defining @RequestMapping handler methods
Supported method argument types
Supported method return types
Binding request parameters to method parameters with @RequestParam
Mapping the request body with the @RequestBody annotation
Mapping the response body with the @ResponseBody annotation
Creating REST Controllers with the @RestController annotation
Using HttpEntity
Using @ModelAttribute on a method
Using @ModelAttribute on a method argument
Using @SessionAttributes to store model attributes in the HTTP session between requests
Using @SessionAttribute to access pre-existing global session attributes
Using @RequestAttribute to access request attributes
Working with "application/x-www-form-urlencoded" data
Mapping cookie values with the @CookieValue annotation
Mapping request header attributes with the @RequestHeader annotation
Method Parameters And Type Conversion
Customizing WebDataBinder initialization
Advising controllers with @ControllerAdvice and @RestControllerAdvice
Jackson Serialization View Support
Jackson JSONP Support
22.3.4. Asynchronous Request Processing
Exception Handling for Async Requests
Intercepting Async Requests
HTTP Streaming
HTTP Streaming With Server-Sent Events
HTTP Streaming Directly To The OutputStream
Configuring Asynchronous Request Processing
22.3.5. Testing Controllers
22.4. Handler mappings
22.4.1. Intercepting requests with a HandlerInterceptor
22.5. Resolving views
22.5.1. Resolving views with the ViewResolver interface
22.5.2. Chaining ViewResolvers
22.5.3. Redirecting to Views
RedirectView
The redirect: prefix
The forward: prefix
22.5.4. ContentNegotiatingViewResolver
22.6. Using flash attributes
22.7. Building URIs
22.7.1. Building URIs to Controllers and methods
22.7.2. Building URIs to Controllers and methods from views
22.8. Using locales
22.8.1. Obtaining Time Zone Information
22.8.2. AcceptHeaderLocaleResolver
22.8.3. CookieLocaleResolver
22.8.4. SessionLocaleResolver
22.8.5. LocaleChangeInterceptor
22.9. Using themes
22.9.1. Overview of themes
22.9.2. Defining themes
22.9.3. Theme resolvers
22.10. Spring’s multipart (file upload) support
22.10.1. Introduction
22.10.2. Using a MultipartResolver with Commons FileUpload
22.10.3. Using a MultipartResolver with Servlet 3.0
22.10.4. Handling a file upload in a form
22.10.5. Handling a file upload request from programmatic clients
22.11. Handling exceptions
22.11.1. HandlerExceptionResolver
22.11.2. @ExceptionHandler
22.11.3. Handling Standard Spring MVC Exceptions
22.11.4. Annotating Business Exceptions With @ResponseStatus
22.11.5. Customizing the Default Servlet Container Error Page
22.12. Web Security
22.13. Convention over configuration support
22.13.1. The Controller ControllerClassNameHandlerMapping
22.13.2. The Model ModelMap (ModelAndView)
22.13.3. The View - RequestToViewNameTranslator
22.14. HTTP caching support
22.14.1. Cache-Control HTTP header
22.14.2. HTTP caching support for static resources
22.14.3. Support for the Cache-Control, ETag and Last-Modified response headers in Controllers
22.14.4. Shallow ETag support
22.15. Code-based Servlet container initialization
22.16. Configuring Spring MVC
22.16.1. Enabling the MVC Java Config or the MVC XML Namespace
22.16.2. Customizing the Provided Configuration
22.16.3. Conversion and Formatting
22.16.4. Validation
22.16.5. Interceptors
22.16.6. Content Negotiation
22.16.7. View Controllers
22.16.8. View Resolvers
22.16.9. Serving of Resources
22.16.10. Falling Back On the "Default" Servlet To Serve Resources
22.16.11. Path Matching
22.16.12. Message Converters
22.16.13. Advanced Customizations with MVC Java Config
22.16.14. Advanced Customizations with the MVC Namespace
23. View technologies
23.1. Introduction
23.2. Thymeleaf
23.3. Groovy Markup Templates
23.3.1. Configuration
23.3.2. Example
23.4. Velocity & FreeMarker
23.4.1. Dependencies
23.4.2. Context configuration
23.4.3. Creating templates
23.4.4. Advanced configuration
velocity.properties
FreeMarker
23.4.5. Bind support and form handling
The bind macros
Simple binding
Form input generation macros
HTML escaping and XHTML compliance
23.5. JSP & JSTL
23.5.1. View resolvers
23.5.2. 'Plain-old' JSPs versus JSTL
23.5.3. Additional tags facilitating development
23.5.4. Using Spring’s form tag library
Configuration
The form tag
The input tag
The checkbox tag
The checkboxes tag
The radiobutton tag
The radiobuttons tag
The password tag
The select tag
The option tag
The options tag
The textarea tag
The hidden tag
The errors tag
HTTP Method Conversion
HTML5 Tags
23.6. Script templates
23.6.1. Dependencies
23.6.2. How to integrate script based templating
23.7. XML Marshalling View
23.8. Tiles
23.8.1. Dependencies
23.8.2. How to integrate Tiles
UrlBasedViewResolver
ResourceBundleViewResolver
SimpleSpringPreparerFactory and SpringBeanPreparerFactory
23.9. XSLT
23.9.1. My First Words
Bean definitions
Standard MVC controller code
Document transformation
23.10. Document views (PDF/Excel)
23.10.1. Introduction
23.10.2. Configuration and setup
Document view definitions
Controller code
Subclassing for Excel views
Subclassing for PDF views
23.11. JasperReports
23.11.1. Dependencies
23.11.2. Configuration
Configuring the ViewResolver
Configuring the Views
About Report Files
Using JasperReportsMultiFormatView
23.11.3. Populating the ModelAndView
23.11.4. Working with Sub-Reports
Configuring Sub-Report Files
Configuring Sub-Report Data Sources
23.11.5. Configuring Exporter Parameters
23.12. Feed Views
23.13. JSON Mapping View
23.14. XML Mapping View
24. Integrating with other web frameworks
24.1. Introduction
24.2. Common configuration
24.3. JavaServer Faces 1.2
24.3.1. SpringBeanFacesELResolver (JSF 1.2+)
24.3.2. FacesContextUtils
24.4. Apache Struts 2.x
24.5. Tapestry 5.x
24.6. Further Resources
25. Portlet MVC Framework
25.1. Introduction
25.1.1. Controllers - The C in MVC
25.1.2. Views - The V in MVC
25.1.3. Web-scoped beans
25.2. The DispatcherPortlet
25.3. The ViewRendererServlet
25.4. Controllers
25.4.1. AbstractController and PortletContentGenerator
25.4.2. Other simple controllers
25.4.3. Command Controllers
25.4.4. PortletWrappingController
25.5. Handler mappings
25.5.1. PortletModeHandlerMapping
25.5.2. ParameterHandlerMapping
25.5.3. PortletModeParameterHandlerMapping
25.5.4. Adding HandlerInterceptors
25.5.5. HandlerInterceptorAdapter
25.5.6. ParameterMappingInterceptor
25.6. Views and resolving them
25.7. Multipart (file upload) support
25.7.1. Using the PortletMultipartResolver
25.7.2. Handling a file upload in a form
25.8. Handling exceptions
25.9. Annotation-based controller configuration
25.9.1. Setting up the dispatcher for annotation support
25.9.2. Defining a controller with @Controller
25.9.3. Mapping requests with @RequestMapping
25.9.4. Supported handler method arguments
25.9.5. Binding request parameters to method parameters with @RequestParam
25.9.6. Providing a link to data from the model with @ModelAttribute
25.9.7. Specifying attributes to store in a Session with @SessionAttributes
25.9.8. Customizing WebDataBinder initialization
Customizing data binding with @InitBinder
Configuring a custom WebBindingInitializer
25.10. Portlet application deployment
26. WebSocket Support
26.1. Introduction
26.1.1. WebSocket Fallback Options
26.1.2. A Messaging Architecture
26.1.3. Sub-Protocol Support in WebSocket
26.1.4. Should I Use WebSocket?
26.2. WebSocket API
26.2.1. Create and Configure a WebSocketHandler
26.2.2. Customizing the WebSocket Handshake
26.2.3. WebSocketHandler Decoration
26.2.4. Deployment Considerations
26.2.5. Configuring the WebSocket Engine
26.2.6. Configuring allowed origins
26.3. SockJS Fallback Options
26.3.1. Overview of SockJS
26.3.2. Enable SockJS
26.3.3. HTTP Streaming in IE 8, 9: Ajax/XHR vs IFrame
26.3.4. Heartbeat Messages
26.3.5. Servlet 3 Async Requests
26.3.6. CORS Headers for SockJS
26.3.7. SockJS Client
26.4. STOMP Over WebSocket Messaging Architecture
26.4.1. Overview of STOMP
26.4.2. Enable STOMP over WebSocket
26.4.3. Flow of Messages
26.4.4. Annotation Message Handling
26.4.5. Sending Messages
26.4.6. Simple Broker
26.4.7. Full-Featured Broker
26.4.8. Connections To Full-Featured Broker
26.4.9. Using Dot as Separator in @MessageMapping Destinations
26.4.10. Authentication
26.4.11. User Destinations
26.4.12. Listening To ApplicationContext Events and Intercepting Messages
26.4.13. STOMP Client
26.4.14. WebSocket Scope
26.4.15. Configuration and Performance
26.4.16. Runtime Monitoring
26.4.17. Testing Annotated Controller Methods
27. CORS Support
27.1. Introduction
27.2. Controller method CORS configuration
27.3. Global CORS configuration
27.3.1. JavaConfig
27.3.2. XML namespace
27.4. Advanced Customization
27.5. Filter based CORS support
VII. Integration
28. Remoting and web services using Spring
28.1. Introduction
28.2. Exposing services using RMI
28.2.1. Exporting the service using the RmiServiceExporter
28.2.2. Linking in the service at the client
28.3. Using Hessian or Burlap to remotely call services via HTTP
28.3.1. Wiring up the DispatcherServlet for Hessian and co.
28.3.2. Exposing your beans by using the HessianServiceExporter
28.3.3. Linking in the service on the client
28.3.4. Using Burlap
28.3.5. Applying HTTP basic authentication to a service exposed through Hessian or Burlap
28.4. Exposing services using HTTP invokers
28.4.1. Exposing the service object
28.4.2. Linking in the service at the client
28.5. Web services
28.5.1. Exposing servlet-based web services using JAX-WS
28.5.2. Exporting standalone web services using JAX-WS
28.5.3. Exporting web services using the JAX-WS RI’s Spring support
28.5.4. Accessing web services using JAX-WS
28.6. JMS
28.6.1. Server-side configuration
28.6.2. Client-side configuration
28.7. AMQP
28.8. Auto-detection is not implemented for remote interfaces
28.9. Considerations when choosing a technology
28.10. Accessing RESTful services on the Client
28.10.1. RestTemplate
Working with the URI
Dealing with request and response headers
Jackson JSON Views support
28.10.2. HTTP Message Conversion
StringHttpMessageConverter
FormHttpMessageConverter
ByteArrayHttpMessageConverter
MarshallingHttpMessageConverter
MappingJackson2HttpMessageConverter
MappingJackson2XmlHttpMessageConverter
SourceHttpMessageConverter
BufferedImageHttpMessageConverter
28.10.3. Async RestTemplate
29. Enterprise JavaBeans (EJB) integration
29.1. Introduction
29.2. Accessing EJBs
29.2.1. Concepts
29.2.2. Accessing local SLSBs
29.2.3. Accessing remote SLSBs
29.2.4. Accessing EJB 2.x SLSBs versus EJB 3 SLSBs
29.3. Using Spring’s EJB implementation support classes
29.3.1. EJB 3 injection interceptor
30. JMS (Java Message Service)
30.1. Introduction
30.2. Using Spring JMS
30.2.1. JmsTemplate
30.2.2. Connections
Caching Messaging Resources
SingleConnectionFactory
CachingConnectionFactory
30.2.3. Destination Management
30.2.4. Message Listener Containers
SimpleMessageListenerContainer
DefaultMessageListenerContainer
30.2.5. Transaction management
30.3. Sending a Message
30.3.1. Using Message Converters
30.3.2. SessionCallback and ProducerCallback
30.4. Receiving a message
30.4.1. Synchronous Reception
30.4.2. Asynchronous Reception - Message-Driven POJOs
30.4.3. the SessionAwareMessageListener interface
30.4.4. the MessageListenerAdapter
30.4.5. Processing messages within transactions
30.5. Support for JCA Message Endpoints
30.6. Annotation-driven listener endpoints
30.6.1. Enable listener endpoint annotations
30.6.2. Programmatic endpoints registration
30.6.3. Annotated endpoint method signature
30.6.4. Response management
30.7. JMS Namespace Support
31. JMX
31.1. Introduction
31.2. Exporting your beans to JMX
31.2.1. Creating an MBeanServer
31.2.2. Reusing an existing MBeanServer
31.2.3. Lazy-initialized MBeans
31.2.4. Automatic registration of MBeans
31.2.5. Controlling the registration behavior
31.3. Controlling the management interface of your beans
31.3.1. the MBeanInfoAssembler Interface
31.3.2. Using Source-Level Metadata (Java annotations)
31.3.3. Source-Level Metadata Types
31.3.4. the AutodetectCapableMBeanInfoAssembler interface
31.3.5. Defining management interfaces using Java interfaces
31.3.6. Using MethodNameBasedMBeanInfoAssembler
31.4. Controlling the ObjectNames for your beans
31.4.1. Reading ObjectNames from Properties
31.4.2. Using the MetadataNamingStrategy
31.4.3. Configuring annotation based MBean export
31.5. JSR-160 Connectors
31.5.1. Server-side Connectors
31.5.2. Client-side Connectors
31.5.3. JMX over Burlap/Hessian/SOAP
31.6. Accessing MBeans via Proxies
31.7. Notifications
31.7.1. Registering Listeners for Notifications
31.7.2. Publishing Notifications
31.8. Further Resources
32. JCA CCI
32.1. Introduction
32.2. Configuring CCI
32.2.1. Connector configuration
32.2.2. ConnectionFactory configuration in Spring
32.2.3. Configuring CCI connections
32.2.4. Using a single CCI connection
32.3. Using Spring’s CCI access support
32.3.1. Record conversion
32.3.2. the CciTemplate
32.3.3. DAO support
32.3.4. Automatic output record generation
32.3.5. Summary
32.3.6. Using a CCI Connection and Interaction directly
32.3.7. Example for CciTemplate usage
32.4. Modeling CCI access as operation objects
32.4.1. MappingRecordOperation
32.4.2. MappingCommAreaOperation
32.4.3. Automatic output record generation
32.4.4. Summary
32.4.5. Example for MappingRecordOperation usage
32.4.6. Example for MappingCommAreaOperation usage
32.5. Transactions
33. Email
33.1. Introduction
33.2. Usage
33.2.1. Basic MailSender and SimpleMailMessage usage
33.2.2. Using the JavaMailSender and the MimeMessagePreparator
33.3. Using the JavaMail MimeMessageHelper
33.3.1. Sending attachments and inline resources
Attachments
Inline resources
33.3.2. Creating email content using a templating library
A Velocity-based example
34. Task Execution and Scheduling
34.1. Introduction
34.2. The Spring TaskExecutor abstraction
34.2.1. TaskExecutor types
34.2.2. Using a TaskExecutor
34.3. The Spring TaskScheduler abstraction
34.3.1. the Trigger interface
34.3.2. Trigger implementations
34.3.3. TaskScheduler implementations
34.4. Annotation Support for Scheduling and Asynchronous Execution
34.4.1. Enable scheduling annotations
34.4.2. The @Scheduled Annotation
34.4.3. The @Async Annotation
34.4.4. Executor qualification with @Async
34.4.5. Exception management with @Async
34.5. The Task Namespace
34.5.1. The 'scheduler' element
34.5.2. The 'executor' element
34.5.3. The 'scheduled-tasks' element
34.6. Using the Quartz Scheduler
34.6.1. Using the JobDetailFactoryBean
34.6.2. Using the MethodInvokingJobDetailFactoryBean
34.6.3. Wiring up jobs using triggers and the SchedulerFactoryBean
35. Dynamic language support
35.1. Introduction
35.2. A first example
35.3. Defining beans that are backed by dynamic languages
35.3.1. Common concepts
The <lang:language/> element
Refreshable beans
Inline dynamic language source files
Understanding Constructor Injection in the context of dynamic-language-backed beans
35.3.2. JRuby beans
35.3.3. Groovy beans
Customizing Groovy objects via a callback
35.3.4. BeanShell beans
35.4. Scenarios
35.4.1. Scripted Spring MVC Controllers
35.4.2. Scripted Validators
35.5. Bits and bobs
35.5.1. AOP - advising scripted beans
35.5.2. Scoping
35.6. Further Resources
36. Cache Abstraction
36.1. Introduction
36.2. Understanding the cache abstraction
36.3. Declarative annotation-based caching
36.3.1. @Cacheable annotation
Default Key Generation
Custom Key Generation Declaration
Default Cache Resolution
Custom cache resolution
Synchronized caching
Conditional caching
Available caching SpEL evaluation context
36.3.2. @CachePut annotation
36.3.3. @CacheEvict annotation
36.3.4. @Caching annotation
36.3.5. @CacheConfig annotation
36.3.6. Enable caching annotations
36.3.7. Using custom annotations
36.4. JCache (JSR-107) annotations
36.4.1. Features summary
36.4.2. Enabling JSR-107 support
36.5. Declarative XML-based caching
36.6. Configuring the cache storage
36.6.1. JDK ConcurrentMap-based Cache
36.6.2. EhCache-based Cache
36.6.3. Caffeine Cache
36.6.4. Guava Cache
36.6.5. GemFire-based Cache
36.6.6. JSR-107 Cache
36.6.7. Dealing with caches without a backing store
36.7. Plugging-in different back-end caches
36.8. How can I set the TTL/TTI/Eviction policy/XXX feature?
VIII. Appendices
37. Migrating to Spring Framework 4.x
38. Spring Annotation Programming Model
39. Classic Spring Usage
39.1. Classic ORM usage
39.1.1. Hibernate
The HibernateTemplate
Implementing Spring-based DAOs without callbacks
39.2. JMS Usage
39.2.1. JmsTemplate
39.2.2. Asynchronous Message Reception
39.2.3. Connections
39.2.4. Transaction Management
40. Classic Spring AOP Usage
40.1. Pointcut API in Spring
40.1.1. Concepts
40.1.2. Operations on pointcuts
40.1.3. AspectJ expression pointcuts
40.1.4. Convenience pointcut implementations
Static pointcuts
Dynamic pointcuts
40.1.5. Pointcut superclasses
40.1.6. Custom pointcuts
40.2. Advice API in Spring
40.2.1. Advice lifecycles
40.2.2. Advice types in Spring
Interception around advice
Before advice
Throws advice
After Returning advice
Introduction advice
40.3. Advisor API in Spring
40.4. Using the ProxyFactoryBean to create AOP proxies
40.4.1. Basics
40.4.2. JavaBean properties
40.4.3. JDK- and CGLIB-based proxies
40.4.4. Proxying interfaces
40.4.5. Proxying classes
40.4.6. Using 'global' advisors
40.5. Concise proxy definitions
40.6. Creating AOP proxies programmatically with the ProxyFactory
40.7. Manipulating advised objects
40.8. Using the "autoproxy" facility
40.8.1. Autoproxy bean definitions
BeanNameAutoProxyCreator
DefaultAdvisorAutoProxyCreator
AbstractAdvisorAutoProxyCreator
40.8.2. Using metadata-driven auto-proxying
40.9. Using TargetSources
40.9.1. Hot swappable target sources
40.9.2. Pooling target sources
40.9.3. Prototype target sources
40.9.4. ThreadLocal target sources
40.10. Defining new Advice types
40.11. Further resources
41. XML Schema-based configuration
41.1. Introduction
41.2. XML Schema-based configuration
41.2.1. Referencing the schemas
41.2.2. the util schema
<util:constant/>
<util:property-path/>
<util:properties/>
<util:list/>
<util:map/>
<util:set/>
41.2.3. the jee schema
<jee:jndi-lookup/> (simple)
<jee:jndi-lookup/> (with single JNDI environment setting)
<jee:jndi-lookup/> (with multiple JNDI environment settings)
<jee:jndi-lookup/> (complex)
<jee:local-slsb/> (simple)
<jee:local-slsb/> (complex)
<jee:remote-slsb/>
41.2.4. the lang schema
41.2.5. the jms schema
41.2.6. the tx (transaction) schema
41.2.7. the aop schema
41.2.8. the context schema
<property-placeholder/>
<annotation-config/>
<component-scan/>
<load-time-weaver/>
<spring-configured/>
<mbean-export/>
41.2.9. the tool schema
41.2.10. the jdbc schema
41.2.11. the cache schema
41.2.12. the beans schema
42. Extensible XML authoring
42.1. Introduction
42.2. Authoring the schema
42.3. Coding a NamespaceHandler
42.4. BeanDefinitionParser
42.5. Registering the handler and the schema
42.5.1. 'META-INF/spring.handlers'
42.5.2. 'META-INF/spring.schemas'
42.6. Using a custom extension in your Spring XML configuration
42.7. Meatier examples
42.7.1. Nesting custom tags within custom tags
42.7.2. Custom attributes on 'normal' elements
42.8. Further Resources
43. spring JSP Tag Library
43.1. Introduction
43.2. The argument tag
43.3. The bind tag
43.4. The escapeBody tag
43.5. The eval tag
43.6. The hasBindErrors tag
43.7. The htmlEscape tag
43.8. The message tag
43.9. The nestedPath tag
43.10. The param tag
43.11. The theme tag
43.12. The transform tag
43.13. The url tag
44. spring-form JSP Tag Library
44.1. Introduction
44.2. The button tag
44.3. The checkbox tag
44.4. The checkboxes tag
44.5. The errors tag
44.6. The form tag
44.7. The hidden tag
44.8. The input tag
44.9. The label tag
44.10. The option tag
44.11. The options tag
44.12. The password tag
44.13. The radiobutton tag
44.14. The radiobuttons tag
44.15. The select tag
44.16. The textarea tag
 

The Spring Framework is a lightweight solution and a potential one-stop-shop for building your enterprise-ready applications. However, Spring is modular, allowing you to use only those parts that you need, without having to bring in the rest. You can use the IoC container, with any web framework on top, but you can also use only theHibernate integration code or the JDBC abstraction layer. The Spring Framework supports declarative transaction management, remote access to your logic through RMI or web services, and various options for persisting your data. It offers a full-featured MVC framework, and enables you to integrate AOP transparently into your software.

Spring is designed to be non-intrusive, meaning that your domain logic code generally has no dependencies on the framework itself. In your integration layer (such as the data access layer), some dependencies on the data access technology and the Spring libraries will exist. However, it should be easy to isolate these dependencies from the rest of your code base.

This document is a reference guide to Spring Framework features. If you have any requests, comments, or questions on this document, please post them on the user mailing list. Questions on the Framework itself should be asked on StackOverflow (see https://spring.io/questions).

1. Getting Started with Spring

This reference guide provides detailed information about the Spring Framework. It provides comprehensive documentation for all features, as well as some background about the underlying concepts (such as "Dependency Injection") that Spring has embraced.

If you are just getting started with Spring, you may want to begin using the Spring Framework by creating a Spring Boot based application. Spring Boot provides a quick (and opinionated) way to create a production-ready Spring based application. It is based on the Spring Framework, favors convention over configuration, and is designed to get you up and running as quickly as possible.

You can use start.spring.io to generate a basic project or follow one of the "Getting Started" guides like the Getting Started Building a RESTful Web Service one. As well as being easier to digest, these guides are very task focused, and most of them are based on Spring Boot. They also cover other projects from the Spring portfolio that you might want to consider when solving a particular problem.

2. Introduction to the Spring Framework

The Spring Framework is a Java platform that provides comprehensive infrastructure support for developing Java applications. Spring handles the infrastructure so you can focus on your application.

Spring enables you to build applications from "plain old Java objects" (POJOs) and to apply enterprise services non-invasively to POJOs. This capability applies to the Java SE programming model and to full and partial Java EE.

Examples of how you, as an application developer, can benefit from the Spring platform:

  • Make a Java method execute in a database transaction without having to deal with transaction APIs.
  • Make a local Java method a remote procedure without having to deal with remote APIs.
  • Make a local Java method a management operation without having to deal with JMX APIs.
  • Make a local Java method a message handler without having to deal with JMS APIs.

2.1 Dependency Injection and Inversion of Control

A Java application — a loose term that runs the gamut from constrained, embedded applications to n-tier, server-side enterprise applications — typically consists of objects that collaborate to form the application proper. Thus the objects in an application have dependencies on each other.

Although the Java platform provides a wealth of application development functionality, it lacks the means to organize the basic building blocks into a coherent whole, leaving that task to architects and developers. Although you can use design patterns such as Factory, Abstract Factory, Builder, Decorator, and Service Locator to compose the various classes and object instances that make up an application, these patterns are simply that: best practices given a name, with a description of what the pattern does, where to apply it, the problems it addresses, and so forth. Patterns are formalized best practices that you must implement yourself in your application.

The Spring Framework Inversion of Control (IoC) component addresses this concern by providing a formalized means of composing disparate components into a fully working application ready for use. The Spring Framework codifies formalized design patterns as first-class objects that you can integrate into your own application(s). Numerous organizations and institutions use the Spring Framework in this manner to engineer robust, maintainable applications.

Background

"The question is, what aspect of control are [they] inverting?" Martin Fowler posed this question about Inversion of Control (IoC) on his site in 2004. Fowler suggested renaming the principle to make it more self-explanatory and came up with Dependency Injection.

2.2 Modules

The Spring Framework consists of features organized into about 20 modules. These modules are grouped into Core Container, Data Access/Integration, Web, AOP (Aspect Oriented Programming), Instrumentation, Messaging, and Test, as shown in the following diagram.

Figure 2.1. Overview of the Spring Framework

Spring框架参考文档 Spring Framework Reference Documentation

The following sections list the available modules for each feature along with their artifact names and the topics they cover. Artifact names correlate to artifact IDs used inDependency Management tools.

2.2.1 Core Container

The Core Container consists of the spring-core, spring-beans, spring-context, spring-context-support, and spring-expression (Spring Expression Language) modules.

The spring-core and spring-beans modules provide the fundamental parts of the framework, including the IoC and Dependency Injection features. TheBeanFactory is a sophisticated implementation of the factory pattern. It removes the need for programmatic singletons and allows you to decouple the configuration and specification of dependencies from your actual program logic.

The Context (spring-context) module builds on the solid base provided by the Core and Beans modules: it is a means to access objects in a framework-style manner that is similar to a JNDI registry. The Context module inherits its features from the Beans module and adds support for internationalization (using, for example, resource bundles), event propagation, resource loading, and the transparent creation of contexts by, for example, a Servlet container. The Context module also supports Java EE features such as EJB, JMX, and basic remoting. The ApplicationContext interface is the focal point of the Context module. spring-context-support provides support for integrating common third-party libraries into a Spring application context for caching (EhCache, Guava, JCache), mailing (JavaMail), scheduling (CommonJ, Quartz) and template engines (FreeMarker, JasperReports, Velocity).

The spring-expression module provides a powerful Expression Language for querying and manipulating an object graph at runtime. It is an extension of the unified expression language (unified EL) as specified in the JSP 2.1 specification. The language supports setting and getting property values, property assignment, method invocation, accessing the content of arrays, collections and indexers, logical and arithmetic operators, named variables, and retrieval of objects by name from Spring’s IoC container. It also supports list projection and selection as well as common list aggregations.

2.2.2 AOP and Instrumentation

The spring-aop module provides an AOP Alliance-compliant aspect-oriented programming implementation allowing you to define, for example, method interceptors and pointcuts to cleanly decouple code that implements functionality that should be separated. Using source-level metadata functionality, you can also incorporate behavioral information into your code, in a manner similar to that of .NET attributes.

The separate spring-aspects module provides integration with AspectJ.

The spring-instrument module provides class instrumentation support and classloader implementations to be used in certain application servers. Thespring-instrument-tomcat module contains Spring’s instrumentation agent for Tomcat.

2.2.3 Messaging

Spring Framework 4 includes a spring-messaging module with key abstractions from the Spring Integration project such as Message, MessageChannel,MessageHandler, and others to serve as a foundation for messaging-based applications. The module also includes a set of annotations for mapping messages to methods, similar to the Spring MVC annotation based programming model.

2.2.4 Data Access/Integration

The Data Access/Integration layer consists of the JDBC, ORM, OXM, JMS, and Transaction modules.

The spring-jdbc module provides a JDBC-abstraction layer that removes the need to do tedious JDBC coding and parsing of database-vendor specific error codes.

The spring-tx module supports programmatic and declarative transaction management for classes that implement special interfaces and for all your POJOs (Plain Old Java Objects).

The spring-orm module provides integration layers for popular object-relational mapping APIs, including JPA, JDO, and Hibernate. Using the spring-orm module you can use all of these O/R-mapping frameworks in combination with all of the other features Spring offers, such as the simple declarative transaction management feature mentioned previously.

The spring-oxm module provides an abstraction layer that supports Object/XML mapping implementations such as JAXB, Castor, XMLBeans, JiBX and XStream.

The spring-jms module (Java Messaging Service) contains features for producing and consuming messages. Since Spring Framework 4.1, it provides integration with the spring-messaging module.

2.2.5 Web

The Web layer consists of the spring-web, spring-webmvc, spring-websocket, and spring-webmvc-portlet modules.

The spring-web module provides basic web-oriented integration features such as multipart file upload functionality and the initialization of the IoC container using Servlet listeners and a web-oriented application context. It also contains an HTTP client and the web-related parts of Spring’s remoting support.

The spring-webmvc module (also known as the Web-Servlet module) contains Spring’s model-view-controller (MVC) and REST Web Services implementation for web applications. Spring’s MVC framework provides a clean separation between domain model code and web forms and integrates with all of the other features of the Spring Framework.

The spring-webmvc-portlet module (also known as the Web-Portlet module) provides the MVC implementation to be used in a Portlet environment and mirrors the functionality of the spring-webmvc module.

2.2.6 Test

The spring-test module supports the unit testing and integration testing of Spring components with JUnit or TestNG. It provides consistent loading of SpringApplicationContexts and caching of those contexts. It also provides mock objects that you can use to test your code in isolation.

2.3 Usage scenarios

The building blocks described previously make Spring a logical choice in many scenarios, from embedded applications that run on resource-constrained devices to full-fledged enterprise applications that use Spring’s transaction management functionality and web framework integration.

Figure 2.2. Typical full-fledged Spring web application

Spring框架参考文档 Spring Framework Reference Documentation

Spring’s declarative transaction management features make the web application fully transactional, just as it would be if you used EJB container-managed transactions. All your custom business logic can be implemented with simple POJOs and managed by Spring’s IoC container. Additional services include support for sending email and validation that is independent of the web layer, which lets you choose where to execute validation rules. Spring’s ORM support is integrated with JPA, Hibernate and JDO; for example, when using Hibernate, you can continue to use your existing mapping files and standard Hibernate SessionFactory configuration. Form controllers seamlessly integrate the web-layer with the domain model, removing the need for ActionForms or other classes that transform HTTP parameters to values for your domain model.

Figure 2.3. Spring middle-tier using a third-party web framework

Spring框架参考文档 Spring Framework Reference Documentation

Sometimes circumstances do not allow you to completely switch to a different framework. The Spring Framework does not force you to use everything within it; it is not anall-or-nothing solution. Existing front-ends built with Struts, Tapestry, JSF or other UI frameworks can be integrated with a Spring-based middle-tier, which allows you to use Spring transaction features. You simply need to wire up your business logic using an ApplicationContext and use a WebApplicationContext to integrate your web layer.

Figure 2.4. Remoting usage scenario

Spring框架参考文档 Spring Framework Reference Documentation

When you need to access existing code through web services, you can use Spring’s Hessian-, Burlap-, Rmi- or JaxRpcProxyFactory classes. Enabling remote access to existing applications is not difficult.

Figure 2.5. EJBs - Wrapping existing POJOs

Spring框架参考文档 Spring Framework Reference Documentation

The Spring Framework also provides an access and abstraction layer for Enterprise JavaBeans, enabling you to reuse your existing POJOs and wrap them in stateless session beans for use in scalable, fail-safe web applications that might need declarative security.

2.3.1 Dependency Management and Naming Conventions

Dependency management and dependency injection are different things. To get those nice features of Spring into your application (like dependency injection) you need to assemble all the libraries needed (jar files) and get them onto your classpath at runtime, and possibly at compile time. These dependencies are not virtual components that are injected, but physical resources in a file system (typically). The process of dependency management involves locating those resources, storing them and adding them to classpaths. Dependencies can be direct (e.g. my application depends on Spring at runtime), or indirect (e.g. my application depends on commons-dbcp which depends on commons-pool). The indirect dependencies are also known as "transitive" and it is those dependencies that are hardest to identify and manage.

If you are going to use Spring you need to get a copy of the jar libraries that comprise the pieces of Spring that you need. To make this easier Spring is packaged as a set of modules that separate the dependencies as much as possible, so for example if you don’t want to write a web application you don’t need the spring-web modules. To refer to Spring library modules in this guide we use a shorthand naming convention spring-* or spring-*.jar, where * represents the short name for the module (e.g. spring-core, spring-webmvc, spring-jms, etc.). The actual jar file name that you use is normally the module name concatenated with the version number (e.g. spring-core-4.3.2.RELEASE.jar).

Each release of the Spring Framework will publish artifacts to the following places:

  • Maven Central, which is the default repository that Maven queries, and does not require any special configuration to use. Many of the common libraries that Spring depends on also are available from Maven Central and a large section of the Spring community uses Maven for dependency management, so this is convenient for them. The names of the jars here are in the form spring-*-<version>.jar and the Maven groupId is org.springframework.
  • In a public Maven repository hosted specifically for Spring. In addition to the final GA releases, this repository also hosts development snapshots and milestones. The jar file names are in the same form as Maven Central, so this is a useful place to get development versions of Spring to use with other libraries deployed in Maven Central. This repository also contains a bundle distribution zip file that contains all Spring jars bundled together for easy download.

So the first thing you need to decide is how to manage your dependencies: we generally recommend the use of an automated system like Maven, Gradle or Ivy, but you can also do it manually by downloading all the jars yourself.

You will find bellow the list of Spring artifacts. For a more complete description of each modules, see Section 2.2, “Modules”.

Table 2.1. Spring Framework Artifacts

GroupId ArtifactId Description

org.springframework

spring-aop

Proxy-based AOP support

org.springframework

spring-aspects

AspectJ based aspects

org.springframework

spring-beans

Beans support, including Groovy

org.springframework

spring-context

Application context runtime, including scheduling and remoting abstractions

org.springframework

spring-context-support

Support classes for integrating common third-party libraries into a Spring application context

org.springframework

spring-core

Core utilities, used by many other Spring modules

org.springframework

spring-expression

Spring Expression Language (SpEL)

org.springframework

spring-instrument

Instrumentation agent for JVM bootstrapping

org.springframework

spring-instrument-tomcat

Instrumentation agent for Tomcat

org.springframework

spring-jdbc

JDBC support package, including DataSource setup and JDBC access support

org.springframework

spring-jms

JMS support package, including helper classes to send and receive JMS messages

org.springframework

spring-messaging

Support for messaging architectures and protocols

org.springframework

spring-orm

Object/Relational Mapping, including JPA and Hibernate support

org.springframework

spring-oxm

Object/XML Mapping

org.springframework

spring-test

Support for unit testing and integration testing Spring components

org.springframework

spring-tx

Transaction infrastructure, including DAO support and JCA integration

org.springframework

spring-web

Web support packages, including client and web remoting

org.springframework

spring-webmvc

REST Web Services and model-view-controller implementation for web applications

org.springframework

spring-webmvc-portlet

MVC implementation to be used in a Portlet environment

org.springframework

spring-websocket

WebSocket and SockJS implementations, including STOMP support

Spring Dependencies and Depending on Spring

Although Spring provides integration and support for a huge range of enterprise and other external tools, it intentionally keeps its mandatory dependencies to an absolute minimum: you shouldn’t have to locate and download (even automatically) a large number of jar libraries in order to use Spring for simple use cases. For basic dependency injection there is only one mandatory external dependency, and that is for logging (see below for a more detailed description of logging options).

Next we outline the basic steps needed to configure an application that depends on Spring, first with Maven and then with Gradle and finally using Ivy. In all cases, if anything is unclear, refer to the documentation of your dependency management system, or look at some sample code - Spring itself uses Gradle to manage dependencies when it is building, and our samples mostly use Gradle or Maven.

Maven Dependency Management

If you are using Maven for dependency management you don’t even need to supply the logging dependency explicitly. For example, to create an application context and use dependency injection to configure an application, your Maven dependencies will look like this:

<dependencies>
    <dependency>
        <groupId>org.springframework</groupId>
        <artifactId>spring-context</artifactId>
        <version>4.3.2.RELEASE</version>
        <scope>runtime</scope>
    </dependency>
</dependencies>

That’s it. Note the scope can be declared as runtime if you don’t need to compile against Spring APIs, which is typically the case for basic dependency injection use cases.

The example above works with the Maven Central repository. To use the Spring Maven repository (e.g. for milestones or developer snapshots), you need to specify the repository location in your Maven configuration. For full releases:

<repositories>
    <repository>
        <id>io.spring.repo.maven.release</id>
        <url>http://repo.spring.io/release/</url>
        <snapshots><enabled>false</enabled></snapshots>
    </repository>
</repositories>

For milestones:

<repositories>
    <repository>
        <id>io.spring.repo.maven.milestone</id>
        <url>http://repo.spring.io/milestone/</url>
        <snapshots><enabled>false</enabled></snapshots>
    </repository>
</repositories>

And for snapshots:

<repositories>
    <repository>
        <id>io.spring.repo.maven.snapshot</id>
        <url>http://repo.spring.io/snapshot/</url>
        <snapshots><enabled>true</enabled></snapshots>
    </repository>
</repositories>

Maven "Bill Of Materials" Dependency

It is possible to accidentally mix different versions of Spring JARs when using Maven. For example, you may find that a third-party library, or another Spring project, pulls in a transitive dependency to an older release. If you forget to explicitly declare a direct dependency yourself, all sorts of unexpected issues can arise.

To overcome such problems Maven supports the concept of a "bill of materials" (BOM) dependency. You can import the spring-framework-bom in yourdependencyManagement section to ensure that all spring dependencies (both direct and transitive) are at the same version.

<dependencyManagement>
    <dependencies>
        <dependency>
            <groupId>org.springframework</groupId>
            <artifactId>spring-framework-bom</artifactId>
            <version>4.3.2.RELEASE</version>
            <type>pom</type>
            <scope>import</scope>
        </dependency>
    </dependencies>
</dependencyManagement>

An added benefit of using the BOM is that you no longer need to specify the <version> attribute when depending on Spring Framework artifacts:

<dependencies>
    <dependency>
        <groupId>org.springframework</groupId>
        <artifactId>spring-context</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springframework</groupId>
        <artifactId>spring-web</artifactId>
    </dependency>
<dependencies>

Gradle Dependency Management

To use the Spring repository with the Gradle build system, include the appropriate URL in the repositories section:

repositories {
    mavenCentral()
    // and optionally...
    maven { url "http://repo.spring.io/release" }
}

You can change the repositories URL from /release to /milestone or /snapshot as appropriate. Once a repository has been configured, you can declare dependencies in the usual Gradle way:

dependencies {
    compile("org.springframework:spring-context:4.3.2.RELEASE")
    testCompile("org.springframework:spring-test:4.3.2.RELEASE")
}

Ivy Dependency Management

If you prefer to use Ivy to manage dependencies then there are similar configuration options.

To configure Ivy to point to the Spring repository add the following resolver to your ivysettings.xml:

<resolvers>
    <ibiblio name="io.spring.repo.maven.release"
            m2compatible="true"
            root="http://repo.spring.io/release/"/>
</resolvers>

You can change the root URL from /release/ to /milestone/ or /snapshot/ as appropriate.

Once configured, you can add dependencies in the usual way. For example (in ivy.xml):

<dependency org="org.springframework"
    name="spring-core" rev="4.3.2.RELEASE" conf="compile->runtime"/>

Distribution Zip Files

Although using a build system that supports dependency management is the recommended way to obtain the Spring Framework, it is still possible to download a distribution zip file.

Distribution zips are published to the Spring Maven Repository (this is just for our convenience, you don’t need Maven or any other build system in order to download them).

To download a distribution zip open a web browser to http://repo.spring.io/release/org/springframework/spring and select the appropriate subfolder for the version that you want. Distribution files end -dist.zip, for example spring-framework-{spring-version}-RELEASE-dist.zip. Distributions are also published for milestones and snapshots.

2.3.2 Logging

Logging is a very important dependency for Spring because a) it is the only mandatory external dependency, b) everyone likes to see some output from the tools they are using, and c) Spring integrates with lots of other tools all of which have also made a choice of logging dependency. One of the goals of an application developer is often to have unified logging configured in a central place for the whole application, including all external components. This is more difficult than it might have been since there are so many choices of logging framework.

The mandatory logging dependency in Spring is the Jakarta Commons Logging API (JCL). We compile against JCL and we also make JCL Log objects visible for classes that extend the Spring Framework. It’s important to users that all versions of Spring use the same logging library: migration is easy because backwards compatibility is preserved even with applications that extend Spring. The way we do this is to make one of the modules in Spring depend explicitly on commons-logging(the canonical implementation of JCL), and then make all the other modules depend on that at compile time. If you are using Maven for example, and wondering where you picked up the dependency on commons-logging, then it is from Spring and specifically from the central module called spring-core.

The nice thing about commons-logging is that you don’t need anything else to make your application work. It has a runtime discovery algorithm that looks for other logging frameworks in well known places on the classpath and uses one that it thinks is appropriate (or you can tell it which one if you need to). If nothing else is available you get pretty nice looking logs just from the JDK (java.util.logging or JUL for short). You should find that your Spring application works and logs happily to the console out of the box in most situations, and that’s important.

Not Using Commons Logging

Unfortunately, the runtime discovery algorithm in commons-logging, while convenient for the end-user, is problematic. If we could turn back the clock and start Spring now as a new project it would use a different logging dependency. The first choice would probably be the Simple Logging Facade for Java ( SLF4J), which is also used by a lot of other tools that people use with Spring inside their applications.

There are basically two ways to switch off commons-logging:

  1. Exclude the dependency from the spring-core module (as it is the only module that explicitly depends on commons-logging)
  2. Depend on a special commons-logging dependency that replaces the library with an empty jar (more details can be found in the SLF4J FAQ)

To exclude commons-logging, add the following to your dependencyManagement section:

<dependencies>
    <dependency>
        <groupId>org.springframework</groupId>
        <artifactId>spring-core</artifactId>
        <version>4.3.2.RELEASE</version>
        <exclusions>
            <exclusion>
                <groupId>commons-logging</groupId>
                <artifactId>commons-logging</artifactId>
            </exclusion>
        </exclusions>
    </dependency>
</dependencies>

Now this application is probably broken because there is no implementation of the JCL API on the classpath, so to fix it a new one has to be provided. In the next section we show you how to provide an alternative implementation of JCL using SLF4J as an example.

Using SLF4J

SLF4J is a cleaner dependency and more efficient at runtime than commons-logging because it uses compile-time bindings instead of runtime discovery of the other logging frameworks it integrates. This also means that you have to be more explicit about what you want to happen at runtime, and declare it or configure it accordingly. SLF4J provides bindings to many common logging frameworks, so you can usually choose one that you already use, and bind to that for configuration and management.

SLF4J provides bindings to many common logging frameworks, including JCL, and it also does the reverse: bridges between other logging frameworks and itself. So to use SLF4J with Spring you need to replace the commons-logging dependency with the SLF4J-JCL bridge. Once you have done that then logging calls from within Spring will be translated into logging calls to the SLF4J API, so if other libraries in your application use that API, then you have a single place to configure and manage logging.

A common choice might be to bridge Spring to SLF4J, and then provide explicit binding from SLF4J to Log4J. You need to supply 4 dependencies (and exclude the existing commons-logging): the bridge, the SLF4J API, the binding to Log4J, and the Log4J implementation itself. In Maven you would do that like this

<dependencies>
    <dependency>
        <groupId>org.springframework</groupId>
        <artifactId>spring-core</artifactId>
        <version>4.3.2.RELEASE</version>
        <exclusions>
            <exclusion>
                <groupId>commons-logging</groupId>
                <artifactId>commons-logging</artifactId>
            </exclusion>
        </exclusions>
    </dependency>
    <dependency>
        <groupId>org.slf4j</groupId>
        <artifactId>jcl-over-slf4j</artifactId>
        <version>1.5.8</version>
    </dependency>
    <dependency>
        <groupId>org.slf4j</groupId>
        <artifactId>slf4j-api</artifactId>
        <version>1.5.8</version>
    </dependency>
    <dependency>
        <groupId>org.slf4j</groupId>
        <artifactId>slf4j-log4j12</artifactId>
        <version>1.5.8</version>
    </dependency>
    <dependency>
        <groupId>log4j</groupId>
        <artifactId>log4j</artifactId>
        <version>1.2.14</version>
    </dependency>
</dependencies>

That might seem like a lot of dependencies just to get some logging. Well it is, but it is optional, and it should behave better than the vanilla commons-logging with respect to classloader issues, notably if you are in a strict container like an OSGi platform. Allegedly there is also a performance benefit because the bindings are at compile-time not runtime.

A more common choice amongst SLF4J users, which uses fewer steps and generates fewer dependencies, is to bind directly to Logback. This removes the extra binding step because Logback implements SLF4J directly, so you only need to depend on two libraries not four ( jcl-over-slf4j and logback). If you do that you might also need to exclude the slf4j-api dependency from other external dependencies (not Spring), because you only want one version of that API on the classpath.

Using Log4J

Many people use Log4j as a logging framework for configuration and management purposes. It’s efficient and well-established, and in fact it’s what we use at runtime when we build and test Spring. Spring also provides some utilities for configuring and initializing Log4j, so it has an optional compile-time dependency on Log4j in some modules.

To make Log4j work with the default JCL dependency ( commons-logging) all you need to do is put Log4j on the classpath, and provide it with a configuration file (log4j.properties or log4j.xml in the root of the classpath). So for Maven users this is your dependency declaration:

<dependencies>
    <dependency>
        <groupId>org.springframework</groupId>
        <artifactId>spring-core</artifactId>
        <version>4.3.2.RELEASE</version>
    </dependency>
    <dependency>
        <groupId>log4j</groupId>
        <artifactId>log4j</artifactId>
        <version>1.2.14</version>
    </dependency>
</dependencies>

And here’s a sample log4j.properties for logging to the console:

log4j.rootCategory=INFO, stdout

log4j.appender.stdout=org.apache.log4j.ConsoleAppender
log4j.appender.stdout.layout=org.apache.log4j.PatternLayout
log4j.appender.stdout.layout.ConversionPattern=%d{ABSOLUTE} %5p %t %c{2}:%L - %m%n

log4j.category.org.springframework.beans.factory=DEBUG
Runtime Containers with Native JCL

Many people run their Spring applications in a container that itself provides an implementation of JCL. IBM Websphere Application Server (WAS) is the archetype. This often causes problems, and unfortunately there is no silver bullet solution; simply excluding commons-logging from your application is not enough in most situations.

To be clear about this: the problems reported are usually not with JCL per se, or even with commons-logging: rather they are to do with binding commons-logging to another framework (often Log4J). This can fail because commons-logging changed the way they do the runtime discovery in between the older versions (1.0) found in some containers and the modern versions that most people use now (1.1). Spring does not use any unusual parts of the JCL API, so nothing breaks there, but as soon as Spring or your application tries to do any logging you can find that the bindings to Log4J are not working.

In such cases with WAS the easiest thing to do is to invert the class loader hierarchy (IBM calls it "parent last") so that the application controls the JCL dependency, not the container. That option isn’t always open, but there are plenty of other suggestions in the public domain for alternative approaches, and your mileage may vary depending on the exact version and feature set of the container.

Part II. What’s New in Spring Framework 4.x

3. New Features and Enhancements in Spring Framework 4.0

The Spring Framework was first released in 2004; since then there have been significant major revisions: Spring 2.0 provided XML namespaces and AspectJ support; Spring 2.5 embraced annotation-driven configuration; Spring 3.0 introduced a strong Java 5+ foundation across the framework codebase, and features such as the Java-based @Configuration model.

Version 4.0 is the latest major release of the Spring Framework and the first to fully support Java 8 features. You can still use Spring with older versions of Java, however, the minimum requirement has now been raised to Java SE 6. We have also taken the opportunity of a major release to remove many deprecated classes and methods.

A migration guide for upgrading to Spring 4.0 is available on the Spring Framework GitHub Wiki.

3.1 Improved Getting Started Experience

The new spring.io website provides a whole series of "Getting Started" guides to help you learn Spring. You can read more about the guides in the Chapter 1, Getting Started with Spring section in this document. The new website also provides a comprehensive overview of the many additional projects that are released under the Spring umbrella.

If you are a Maven user you may also be interested in the helpful bill of materials POM file that is now published with each Spring Framework release.

3.2 Removed Deprecated Packages and Methods

All deprecated packages, and many deprecated classes and methods have been removed with version 4.0. If you are upgrading from a previous release of Spring, you should ensure that you have fixed any deprecated calls that you were making to outdated APIs.

For a complete set of changes, check out the API Differences Report.

Note that optional third-party dependencies have been raised to a 2010/2011 minimum (i.e. Spring 4 generally only supports versions released in late 2010 or later now): notably, Hibernate 3.6+, EhCache 2.1+, Quartz 1.8+, Groovy 1.8+, and Joda-Time 2.0+. As an exception to the rule, Spring 4 requires the recent Hibernate Validator 4.3+, and support for Jackson has been focused on 2.0+ now (with Jackson 1.8/1.9 support retained for the time being where Spring 3.2 had it; now just in deprecated form).

3.3 Java 8 (as well as 6 and 7)

Spring Framework 4.0 provides support for several Java 8 features. You can make use of lambda expressions and method references with Spring’s callback interfaces. There is first-class support for java.time (JSR-310), and several existing annotations have been retrofitted as @Repeatable. You can also use Java 8’s parameter name discovery (based on the -parameters compiler flag) as an alternative to compiling your code with debug information enabled.

Spring remains compatible with older versions of Java and the JDK: concretely, Java SE 6 (specifically, a minimum level equivalent to JDK 6 update 18, as released in January 2010) and above are still fully supported. However, for newly started development projects based on Spring 4, we recommend the use of Java 7 or 8.

3.4 Java EE 6 and 7

Java EE version 6 or above is now considered the baseline for Spring Framework 4, with the JPA 2.0 and Servlet 3.0 specifications being of particular relevance. In order to remain compatible with Google App Engine and older application servers, it is possible to deploy a Spring 4 application into a Servlet 2.5 environment. However, Servlet 3.0+ is strongly recommended and a prerequisite in Spring’s test and mock packages for test setups in development environments.

Spring框架参考文档 Spring Framework Reference Documentation

If you are a WebSphere 7 user, be sure to install the JPA 2.0 feature pack. On WebLogic 10.3.4 or higher, install the JPA 2.0 patch that comes with it. This turns both of those server generations into Spring 4 compatible deployment environments.

On a more forward-looking note, Spring Framework 4.0 supports the Java EE 7 level of applicable specifications now: in particular, JMS 2.0, JTA 1.2, JPA 2.1, Bean Validation 1.1, and JSR-236 Concurrency Utilities. As usual, this support focuses on individual use of those specifications, e.g. on Tomcat or in standalone environments. However, it works equally well when a Spring application is deployed to a Java EE 7 server.

Note that Hibernate 4.3 is a JPA 2.1 provider and therefore only supported as of Spring Framework 4.0. The same applies to Hibernate Validator 5.0 as a Bean Validation 1.1 provider. Neither of the two are officially supported with Spring Framework 3.2.

3.5 Groovy Bean Definition DSL

Beginning with Spring Framework 4.0, it is possible to define external bean configuration using a Groovy DSL. This is similar in concept to using XML bean definitions but allows for a more concise syntax. Using Groovy also allows you to easily embed bean definitions directly in your bootstrap code. For example:

def reader = new GroovyBeanDefinitionReader(myApplicationContext)
reader.beans {
    dataSource(BasicDataSource) {
        driverClassName = "org.hsqldb.jdbcDriver"
        url = "jdbc:hsqldb:mem:grailsDB"
        username = "sa"
        password = ""
        settings = [mynew:"setting"]
    }
    sessionFactory(SessionFactory) {
        dataSource = dataSource
    }
    myService(MyService) {
        nestedBean = { AnotherBean bean ->
            dataSource = dataSource
        }
    }
}

For more information consult the GroovyBeanDefinitionReader javadocs.

3.6 Core Container Improvements

There have been several general improvements to the core container:

  • Spring now treats generic types as a form of qualifier when injecting Beans. For example, if you are using a Spring Data Repository you can now easily inject a specific implementation: @Autowired Repository<Customer> customerRepository.
  • If you use Spring’s meta-annotation support, you can now develop custom annotations that expose specific attributes from the source annotation.
  • Beans can now be ordered when they are autowired into lists and arrays. Both the @Order annotation and Ordered interface are supported.
  • The @Lazy annotation can now be used on injection points, as well as on @Bean definitions.
  • The @Description annotation has been introduced for developers using Java-based configuration.
  • A generalized model for conditionally filtering beans has been added via the @Conditional annotation. This is similar to @Profile support but allows for user-defined strategies to be developed programmatically.
  • CGLIB-based proxy classes no longer require a default constructor. Support is provided via the objenesis library which is repackaged inline and distributed as part of the Spring Framework. With this strategy, no constructor at all is being invoked for proxy instances anymore.
  • There is managed time zone support across the framework now, e.g. on LocaleContext.

3.7 General Web Improvements

Deployment to Servlet 2.5 servers remains an option, but Spring Framework 4.0 is now focused primarily on Servlet 3.0+ environments. If you are using the Spring MVC Test Framework you will need to ensure that a Servlet 3.0 compatible JAR is in your test classpath.

In addition to the WebSocket support mentioned later, the following general improvements have been made to Spring’s Web modules:

3.8 WebSocket, SockJS, and STOMP Messaging

A new spring-websocket module provides comprehensive support for WebSocket-based, two-way communication between client and server in web applications. It is compatible with JSR-356, the Java WebSocket API, and in addition provides SockJS-based fallback options (i.e. WebSocket emulation) for use in browsers that don’t yet support the WebSocket protocol (e.g. Internet Explorer < 10).

A new spring-messaging module adds support for STOMP as the WebSocket sub-protocol to use in applications along with an annotation programming model for routing and processing STOMP messages from WebSocket clients. As a result an @Controller can now contain both @RequestMapping and @MessageMappingmethods for handling HTTP requests and messages from WebSocket-connected clients. The new spring-messaging module also contains key abstractions formerly from the Spring Integration project such as Message, MessageChannel, MessageHandler, and others to serve as a foundation for messaging-based applications.

For further details, including a more thorough introduction, see the Chapter 26, WebSocket Support section.

3.9 Testing Improvements

In addition to pruning of deprecated code within the spring-test module, Spring Framework 4.0 introduces several new features for use in unit and integration testing.

  • Almost all annotations in the spring-test module (e.g., @ContextConfiguration, @WebAppConfiguration, @ContextHierarchy, @ActiveProfiles, etc.) can now be used as meta-annotations to create custom composed annotations and reduce configuration duplication across a test suite.
  • Active bean definition profiles can now be resolved programmatically, simply by implementing a custom ActiveProfilesResolver and registering it via theresolver attribute of @ActiveProfiles.
  • A new SocketUtils class has been introduced in the spring-core module which enables you to scan for free TCP and UDP server ports on localhost. This functionality is not specific to testing but can prove very useful when writing integration tests that require the use of sockets, for example tests that start an in-memory SMTP server, FTP server, Servlet container, etc.
  • As of Spring 4.0, the set of mocks in the org.springframework.mock.web package is now based on the Servlet 3.0 API. Furthermore, several of the Servlet API mocks (e.g., MockHttpServletRequest, MockServletContext, etc.) have been updated with minor enhancements and improved configurability.

4. New Features and Enhancements in Spring Framework 4.1

4.1 JMS Improvements

Spring 4.1 introduces a much simpler infrastructure to register JMS listener endpoints by annotating bean methods with @JmsListener. The XML namespace has been enhanced to support this new style (jms:annotation-driven), and it is also possible to fully configure the infrastructure using Java config (@EnableJms,JmsListenerContainerFactory). It is also possible to register listener endpoints programmatically using JmsListenerConfigurer.

Spring 4.1 also aligns its JMS support to allow you to benefit from the spring-messaging abstraction introduced in 4.0, that is:

  • Message listener endpoints can have a more flexible signature and benefit from standard messaging annotations such as @Payload, @Header, @Headers, and@SendTo. It is also possible to use a standard Message in lieu of javax.jms.Message as method argument.
  • A new JmsMessageOperations interface is available and permits JmsTemplate like operations using the Message abstraction.

Finally, Spring 4.1 provides additional miscellaneous improvements:

  • Synchronous request-reply operations support in JmsTemplate
  • Listener priority can be specified per <jms:listener/> element
  • Recovery options for the message listener container are configurable using a BackOff implementation
  • JMS 2.0 shared consumers are supported

4.2 Caching Improvements

Spring 4.1 supports JCache (JSR-107) annotations using Spring’s existing cache configuration and infrastructure abstraction; no changes are required to use the standard annotations.

Spring 4.1 also improves its own caching abstraction significantly:

  • Caches can be resolved at runtime using a CacheResolver. As a result the value argument defining the cache name(s) to use is no longer mandatory.
  • More operation-level customizations: cache resolver, cache manager, key generator
  • A new @CacheConfig class-level annotation allows common settings to be shared at the class level without enabling any cache operation.
  • Better exception handling of cached methods using CacheErrorHandler

Spring 4.1 also has a breaking change in the Cache interface as a new putIfAbsent method has been added.

4.3 Web Improvements

  • The existing support for resource handling based on the ResourceHttpRequestHandler has been expanded with new abstractions ResourceResolver,ResourceTransformer, and ResourceUrlProvider. A number of built-in implementations provide support for versioned resource URLs (for effective HTTP caching), locating gzipped resources, generating an HTML 5 AppCache manifests, and more. See Section 22.16.9, “Serving of Resources”.
  • JDK 1.8’s java.util.Optional is now supported for @RequestParam, @RequestHeader, and @MatrixVariable controller method arguments.
  • ListenableFuture is supported as a return value alternative to DeferredResult where an underlying service (or perhaps a call to AsyncRestTemplate) already returns ListenableFuture.
  • @ModelAttribute methods are now invoked in an order that respects inter-dependencies. See SPR-6299.
  • Jackson’s @JsonView is supported directly on @ResponseBody and ResponseEntity controller methods for serializing different amounts of detail for the same POJO (e.g. summary vs. detail page). This is also supported with View-based rendering by adding the serialization view type as a model attribute under a special key. See the section called “Jackson Serialization View Support” for details.
  • JSONP is now supported with Jackson. See the section called “Jackson JSONP Support”.
  • A new lifecycle option is available for intercepting @ResponseBody and ResponseEntity methods just after the controller method returns and before the response is written. To take advantage declare an @ControllerAdvice bean that implements ResponseBodyAdvice. The built-in support for @JsonView and JSONP take advantage of this. See Section 22.4.1, “Intercepting requests with a HandlerInterceptor”.

  • There are three new HttpMessageConverter options:

    • Gson — lighter footprint than Jackson; has already been in use in Spring Android.
    • Google Protocol Buffers — efficient and effective as an inter-service communication data protocol within an enterprise but can also be exposed as JSON and XML for browsers.
    • Jackson based XML serialization is now supported through the jackson-dataformat-xml extension. When using @EnableWebMvc or<mvc:annotation-driven/>, this is used by default instead of JAXB2 if jackson-dataformat-xml is in the classpath.
  • Views such as JSPs can now build links to controllers by referring to controller mappings by name. A default name is assigned to every @RequestMapping. For example FooController with method handleFoo is named "FC#handleFoo". The naming strategy is pluggable. It is also possible to name an @RequestMappingexplicitly through its name attribute. A new mvcUrl function in the Spring JSP tag library makes this easy to use in JSP pages. See Section 22.7.2, “Building URIs to Controllers and methods from views”.
  • ResponseEntity provides a builder-style API to guide controller methods towards the preparation of server-side responses, e.g. ResponseEntity.ok().
  • RequestEntity is a new type that provides a builder-style API to guide client-side REST code towards the preparation of HTTP requests.

  • MVC Java config and XML namespace:

    • View resolvers can now be configured including support for content negotiation, see Section 22.16.8, “View Resolvers”.
    • View controllers now have built-in support for redirects and for setting the response status. An application can use this to configure redirect URLs, render 404 responses with a view, send "no content" responses, etc. Some use cases are listed here.
    • Path matching customizations are frequently used and now built-in. See Section 22.16.11, “Path Matching”.
  • Groovy markup template support (based on Groovy 2.3). See the GroovyMarkupConfigurer and respecitve ViewResolver and `View' implementations.

4.4 WebSocket Messaging Improvements

  • SockJS (Java) client-side support. See SockJsClient and classes in same package.
  • New application context events SessionSubscribeEvent and SessionUnsubscribeEvent published when STOMP clients subscribe and unsubscribe.
  • New "websocket" scope. See Section 26.4.14, “WebSocket Scope”.
  • @SendToUser can target only a single session and does not require an authenticated user.
  • @MessageMapping methods can use dot "." instead of slash "/" as path separator. See SPR-11660.
  • STOMP/WebSocket monitoring info collected and logged. See Section 26.4.16, “Runtime Monitoring”.
  • Significantly optimized and improved logging that should remain very readable and compact even at DEBUG level.
  • Optimized message creation including support for temporary message mutability and avoiding automatic message id and timestamp creation. See Javadoc ofMessageHeaderAccessor.
  • Close STOMP/WebSocket connections that have no activity within 60 seconds after the WebSocket session is established. See SPR-11884.

4.5 Testing Improvements

5. New Features and Enhancements in Spring Framework 4.2

5.1 Core Container Improvements

  • Annotations such as @Bean get detected and processed on Java 8 default methods as well, allowing for composing a configuration class from interfaces with default@Bean methods.
  • Configuration classes may declare @Import with regular component classes now, allowing for a mix of imported configuration classes and component classes.
  • Configuration classes may declare an @Order value, getting processed in a corresponding order (e.g. for overriding beans by name) even when detected through classpath scanning.
  • @Resource injection points support an @Lazy declaration, analogous to @Autowired, receiving a lazy-initializing proxy for the requested target bean.

  • The application event infrastructure now offers an annotation-based model as well as the ability to publish any arbitrary event.

    • Any public method in a managed bean can be annotated with @EventListener to consume events.
    • @TransactionalEventListener provides transaction-bound event support.

  • Spring Framework 4.2 introduces first-class support for declaring and looking up aliases for annotation attributes. The new @AliasFor annotation can be used to declare a pair of aliased attributes within a single annotation or to declare an alias from one attribute in a custom composed annotation to an attribute in a meta-annotation.

    • The following annotations have been retrofitted with @AliasFor support in order to provide meaningful aliases for their value attributes: @Cacheable,@CacheEvict, @CachePut, @ComponentScan, @ComponentScan.Filter, @ImportResource, @Scope, @ManagedResource, @Header, @Payload,@SendToUser, @ActiveProfiles, @ContextConfiguration, @Sql, @TestExecutionListeners, @TestPropertySource, @Transactional,@ControllerAdvice, @CookieValue, @CrossOrigin, @MatrixVariable, @RequestHeader, @RequestMapping, @RequestParam, @RequestPart,@ResponseStatus, @SessionAttributes, @ActionMapping, @RenderMapping, @EventListener, @TransactionalEventListener.

    • For example, @ContextConfiguration from the spring-test module is now declared as follows:

      public @interface ContextConfiguration {

    @AliasFor("locations")
    String[] value() default {};

    @AliasFor("value")
    String[] locations() default {};

    // ...
}
    • Similarly, composed annotations that override attributes from meta-annotations can now use @AliasFor for fine-grained control over exactly which attributes are overridden within an annotation hierarchy. In fact, it is now possible to declare an alias for the value attribute of a meta-annotation.

    • For example, one can now develop a composed annotation with a custom attribute override as follows.

      @ContextConfiguration
public @interface MyTestConfig {

    @AliasFor(annotation = ContextConfiguration.class, attribute = "value")
    String[] xmlFiles();

    // ...
}
    • See Spring Annotation Programming Model.
  • Numerous improvements to Spring’s search algorithms used for finding meta-annotations. For example, locally declared composed annotations are now favored over inherited annotations.
  • Composed annotations that override attributes from meta-annotations can now be discovered on interfaces and on abstract, bridge, & interface methods as well as on classes, standard methods, constructors, and fields.
  • Maps representing annotation attributes (and AnnotationAttributes instances) can be synthesized (i.e., converted) into an annotation.
  • The features of field-based data binding (DirectFieldAccessor) have been aligned with the current property-based data binding (BeanWrapper). In particular, field-based binding now supports navigation for Collections, Arrays, and Maps.
  • DefaultConversionService now provides out-of-the-box converters for Stream, Charset, Currency, and TimeZone. Such converters can be added individually to any arbitrary ConversionService as well.
  • DefaultFormattingConversionService comes with out-of-the-box support for the value types in JSR-354 Money & Currency (if the 'javax.money' API is present on the classpath): namely, MonetaryAmount and CurrencyUnit. This includes support for applying @NumberFormat.
  • @NumberFormat can now be used as a meta-annotation.
  • JavaMailSenderImpl has a new testConnection() method for checking connectivity to the server.
  • ScheduledTaskRegistrar exposes scheduled tasks.
  • Apache commons-pool2 is now supported for a pooling AOP CommonsPool2TargetSource.
  • Introduced StandardScriptFactory as a JSR-223 based mechanism for scripted beans, exposed through the lang:std element in XML. Supports e.g. JavaScript and JRuby. (Note: JRubyScriptFactory and lang:jruby are deprecated now, in favor of using JSR-223.)

5.2 Data Access Improvements

  • javax.transaction.Transactional is now supported via AspectJ.
  • SimpleJdbcCallOperations now supports named binding.
  • Full support for Hibernate ORM 5.0: as a JPA provider (automatically adapted) as well as through its native API (covered by the neworg.springframework.orm.hibernate5 package).
  • Embedded databases can now be automatically assigned unique names, and <jdbc:embedded-database> supports a new database-name attribute. See "Testing Improvements" below for further details.

5.3 JMS Improvements

  • The autoStartup attribute can be controlled via JmsListenerContainerFactory.
  • The type of the reply Destination can now be configured per listener container.
  • The value of the @SendTo annotation can now use a SpEL expression.
  • The response destination can be computed at runtime using JmsResponse
  • @JmsListener is now a repeatable annotation to declare several JMS containers on the same method (use the newly introduced @JmsListeners if you’re not using Java8 yet).

5.4 Web Improvements

  • HTTP Streaming and Server-Sent Events support, see the section called “HTTP Streaming”.
  • Built-in support for CORS including global (MVC Java config and XML namespace) and local (e.g. @CrossOrigin) configuration. See Chapter 27, CORS Supportfor details.

  • HTTP caching updates:

    • new CacheControl builder; plugged into ResponseEntity, WebContentGenerator, ResourceHttpRequestHandler.
    • improved ETag/Last-Modified support in WebRequest.
  • Custom mapping annotations, using @RequestMapping as a meta-annotation.
  • Public methods in AbstractHandlerMethodMapping to register and unregister request mappings at runtime.
  • Protected createDispatcherServlet method in AbstractDispatcherServletInitializer to further customize the DispatcherServlet instance to use.
  • HandlerMethod as a method argument on @ExceptionHandler methods, especially handy in @ControllerAdvice components.
  • java.util.concurrent.CompletableFuture as an @Controller method return value type.
  • Byte-range request support in HttpHeaders and for serving static resources.
  • @ResponseStatus detected on nested exceptions.

  • UriTemplateHandler extension point in the RestTemplate.

    • DefaultUriTemplateHandler exposes baseUrl property and path segment encoding options.
    • the extension point can also be used to plug in any URI template library.
  • OkHTTP integration with the RestTemplate.
  • Custom baseUrl alternative for methods in MvcUriComponentsBuilder.
  • Serialization/deserialization exception messages are now logged at WARN level.
  • Default JSON prefix has been changed from "{} && " to the safer ")]}', " one.
  • New RequestBodyAdvice extension point and built-in implementation to support Jackson’s @JsonView on @RequestBody method arguments.
  • When using GSON or Jackson 2.6+, the handler method return type is used to improve serialization of parameterized types like List<Foo>.
  • Introduced ScriptTemplateView as a JSR-223 based mechanism for scripted web views, with a focus on JavaScript view templating on Nashorn (JDK 8).

5.5 WebSocket Messaging Improvements

  • Expose presence information about connected users and subscriptions:

    • new SimpUserRegistry exposed as a bean named "userRegistry".
    • sharing of presence information across cluster of servers (see broker relay config options).
  • Resolve user destinations across cluster of servers (see broker relay config options).
  • StompSubProtocolErrorHandler extension point to customize and control STOMP ERROR frames to clients.
  • Global @MessageExceptionHandler methods via @ControllerAdvice components.
  • Heart-beats and a SpEL expression 'selector' header for subscriptions with SimpleBrokerMessageHandler.
  • STOMP client for use over TCP and WebSocket; see Section 26.4.13, “STOMP Client”.
  • @SendTo and @SendToUser can contain destination variable placeholders.
  • Jackson’s @JsonView supported for return values on @MessageMapping and @SubscribeMapping methods.
  • ListenableFuture and CompletableFuture as return value types from @MessageMapping and @SubscribeMapping methods.
  • MarshallingMessageConverter for XML payloads.

5.6 Testing Improvements

  • JUnit-based integration tests can now be executed with JUnit rules instead of the SpringJUnit4ClassRunner. This allows Spring-based integration tests to be run with alternative runners like JUnit’s Parameterized or third-party runners such as the MockitoJUnitRunner.

  • The Spring MVC Test framework now provides first-class support for HtmlUnit, including integration with Selenium’s WebDriver, allowing for page-based web application testing without the need to deploy to a Servlet container.

  • AopTestUtils is a new testing utility that allows developers to obtain a reference to the underlying target object hidden behind one or more Spring proxies.

  • ReflectionTestUtils now supports setting and getting static fields, including constants.
  • The original ordering of bean definition profiles declared via @ActiveProfiles is now retained in order to support use cases such as Spring Boot’sConfigFileApplicationListener which loads configuration files based on the names of active profiles.
  • @DirtiesContext supports new BEFORE_METHOD, BEFORE_CLASS, and BEFORE_EACH_TEST_METHOD modes for closing the ApplicationContext before a test — for example, if some rogue (i.e., yet to be determined) test within a large test suite has corrupted the original configuration for the ApplicationContext.
  • @Commit is a new annotation that may be used as a direct replacement for @Rollback(false).

  • @Rollback may now be used to configure class-level default rollback semantics.

    • Consequently, @TransactionConfiguration is now deprecated and will be removed in a subsequent release.
  • @Sql now supports execution of inlined SQL statements via a new statements attribute.
  • The ContextCache that is used for caching ApplicationContexts between tests is now a public API with a default implementation that can be replaced for custom caching needs.
  • DefaultTestContext, DefaultBootstrapContext, and DefaultCacheAwareContextLoaderDelegate are now public classes in the support subpackage, allowing for custom extensions.
  • TestContextBootstrappers are now responsible for building the TestContext.
  • In the Spring MVC Test framework, MvcResult details can now be logged at DEBUG level or written to a custom OutputStream or Writer. See the new log(),print(OutputStream), and print(Writer) methods in MockMvcResultHandlers for details.
  • The JDBC XML namespace supports a new database-name attribute in <jdbc:embedded-database>, allowing developers to set unique names for embedded databases –- for example, via a SpEL expression or a property placeholder that is influenced by the current active bean definition profiles.

  • Embedded databases can now be automatically assigned a unique name, allowing common test database configuration to be reused in differentApplicationContexts within a test suite.

  • MockHttpServletRequest and MockHttpServletResponse now provide better support for date header formatting via the getDateHeader and setDateHeadermethods.

6. New Features and Enhancements in Spring Framework 4.3

6.1 Core Container Improvements

  • Core container exceptions provide richer metadata to evaluate programmatically.
  • Java 8 default methods get detected as bean property getters/setters.
  • It is no longer necessary to specify the @Autowired annotation if the target bean only defines one constructor.
  • @Configuration classes support constructor injection.
  • Any SpEL expression used to specify the condition of an @EventListener can now refer to beans (e.g. @beanName.method()).
  • Composed annotations can now override array attributes in meta-annotations with a single element of the component type of the array. For example, theString[] path attribute of @RequestMapping can be overridden with String path in a composed annotation.
  • @Scheduled and @Schedules may now be used as meta-annotations to create custom composed annotations with attribute overrides.
  • @Scheduled is properly supported on beans of any scope.

6.2 Data Access Improvements

  • jdbc:initialize-database and jdbc:embedded-database support a configurable separator to be applied to each script.

6.3 Caching Improvements

Spring 4.3 allows concurrent calls on a given key to be synchronized so that the value is only computed once. This is an opt-in feature that should be enabled via the newsync attribute on @Cacheable. This features introduces a breaking change in the Cache interface as a get(Object key, Callable<T> valueLoader) method has been added.

Spring 4.3 also improves the caching abstraction as follows:

  • SpEL expressions in caches-related annotations can now refer to beans (i.e. @beanName.method()).
  • ConcurrentMapCacheManager and ConcurrentMapCache now support the serialization of cache entries via a new storeByValue attribute.
  • @Cacheable, @CacheEvict, @CachePut, and @Caching may now be used as meta-annotations to create custom composed annotations with attribute overrides.

6.4 JMS Improvements

  • @SendTo can now be specified at the class level to share a common reply destination.
  • @JmsListener and @JmsListeners may now be used as meta-annotations to create custom composed annotations with attribute overrides.

6.5 Web Improvements

  • Built-in support for HTTP HEAD and HTTP OPTIONS.

  • New @GetMapping, @PostMapping, @PutMapping, @DeleteMapping, and @PatchMapping composed annotations for @RequestMapping.

  • New @RequestScope, @SessionScope, and @ApplicationScope composed annotations for web scopes.

  • New @RestControllerAdvice annotation with combined @ControllerAdvice with @ResponseBody semantics.
  • @ResponseStatus is now supported at the class level and inherited by all methods.
  • New @SessionAttribute annotation for access to session attributes (see example).
  • New @RequestAttribute annotation for access to request attributes (see example).
  • @ModelAttribute allows preventing data binding via binding=false attribute (see reference).
  • Consistent exposure of Errors and custom Throwables to MVC exception handlers.
  • Consistent charset handling in HTTP message converters, including a UTF-8 default for multipart text content.
  • Static resource handling uses the configured ContentNegotiationManager for media type determination.
  • RestTemplate and AsyncRestTemplate support strict URI variable encoding via DefaultUriTemplateHandler.
  • AsyncRestTemplate supports request interception.

6.6 WebSocket Messaging Improvements

  • @SendTo and @SendToUser can now be specified at class-level to share a common destination.

6.7 Testing Improvements

  • The JUnit support in the Spring TestContext Framework now requires JUnit 4.12 or higher.
  • New SpringRunner alias for the SpringJUnit4ClassRunner.
  • Test related annotations may now be declared on interfaces — for example, for use with test interfaces that make use of Java 8 based interface default methods.
  • An empty declaration of @ContextConfiguration can now be completely omitted if default XML files, Groovy scripts, or @Configuration classes are detected.
  • @Transactional test methods are no longer required to be public (e.g., in TestNG and JUnit 5).
  • @BeforeTransaction and @AfterTransaction methods are no longer required to be public and may now be declared on Java 8 based interface default methods.
  • The ApplicationContext cache in the Spring TestContext Framework is now bounded with a default maximum size of 32 and a least recently used eviction policy. The maximum size can be configured by setting a JVM system property or Spring property called spring.test.context.cache.maxSize.
  • New ContextCustomizer API for customizing a test ApplicationContext after bean definitions have been loaded into the context but before the context has been refreshed. Customizers can be registered globally by third parties, foregoing the need to implement a custom ContextLoader.
  • @Sql and @SqlGroup may now be used as meta-annotations to create custom composed annotations with attribute overrides.
  • ReflectionTestUtils now automatically unwraps proxies when setting or getting a field.
  • Server-side Spring MVC Test supports expectations on response headers with multiple values.
  • Server-side Spring MVC Test parses form data request content and populates request parameters.
  • Server-side Spring MVC Test supports mock-like assertions for invoked handler methods.
  • Client-side REST test support allows indicating how many times a request is expected and whether the order of declaration for expectations should be ignored (seeSection 15.6.3, “Client-Side REST Tests”).
  • Client-side REST Test supports expectations for form data in the request body.

6.8 Support for new library and server generations

  • Hibernate ORM 5.2 (still supporting 4.2/4.3 and 5.0/5.1 as well, with 3.6 deprecated now)
  • Jackson 2.8 (minimum raised to Jackson 2.6+ as of Spring 4.3)
  • OkHttp 3.x (still supporting OkHttp 2.x side by side)
  • Netty 4.1
  • Undertow 1.4
  • Tomcat 8.5.2 as well as 9.0 M6

Furthermore, Spring Framework 4.3 embeds the updated ASM 5.1 and Objenesis 2.4 in spring-core.jar.

Part III. Core Technologies

 

This part of the reference documentation covers all of those technologies that are absolutely integral to the Spring Framework.

Foremost amongst these is the Spring Framework’s Inversion of Control (IoC) container. A thorough treatment of the Spring Framework’s IoC container is closely followed by comprehensive coverage of Spring’s Aspect-Oriented Programming (AOP) technologies. The Spring Framework has its own AOP framework, which is conceptually easy to understand, and which successfully addresses the 80% sweet spot of AOP requirements in Java enterprise programming.

Coverage of Spring’s integration with AspectJ (currently the richest - in terms of features - and certainly most mature AOP implementation in the Java enterprise space) is also provided.

7. The IoC container

7.1 Introduction to the Spring IoC container and beans

This chapter covers the Spring Framework implementation of the Inversion of Control (IoC) [1] principle. IoC is also known as dependency injection (DI). It is a process whereby objects define their dependencies, that is, the other objects they work with, only through constructor arguments, arguments to a factory method, or properties that are set on the object instance after it is constructed or returned from a factory method. The container then injects those dependencies when it creates the bean. This process is fundamentally the inverse, hence the name Inversion of Control (IoC), of the bean itself controlling the instantiation or location of its dependencies by using direct construction of classes, or a mechanism such as the Service Locator pattern.

The org.springframework.beans and org.springframework.context packages are the basis for Spring Framework’s IoC container. The BeanFactory interface provides an advanced configuration mechanism capable of managing any type of object. ApplicationContext is a sub-interface of BeanFactory. It adds easier integration with Spring’s AOP features; message resource handling (for use in internationalization), event publication; and application-layer specific contexts such as theWebApplicationContext for use in web applications.

In short, the BeanFactory provides the configuration framework and basic functionality, and the ApplicationContext adds more enterprise-specific functionality. TheApplicationContext is a complete superset of the BeanFactory, and is used exclusively in this chapter in descriptions of Spring’s IoC container. For more information on using the BeanFactory instead of the ApplicationContext, refer to Section 7.16, “The BeanFactory”.

In Spring, the objects that form the backbone of your application and that are managed by the Spring IoC container are called beans. A bean is an object that is instantiated, assembled, and otherwise managed by a Spring IoC container. Otherwise, a bean is simply one of many objects in your application. Beans, and thedependencies among them, are reflected in the configuration metadata used by a container.

7.2 Container overview

The interface org.springframework.context.ApplicationContext represents the Spring IoC container and is responsible for instantiating, configuring, and assembling the aforementioned beans. The container gets its instructions on what objects to instantiate, configure, and assemble by reading configuration metadata. The configuration metadata is represented in XML, Java annotations, or Java code. It allows you to express the objects that compose your application and the rich interdependencies between such objects.

Several implementations of the ApplicationContext interface are supplied out-of-the-box with Spring. In standalone applications it is common to create an instance ofClassPathXmlApplicationContext or FileSystemXmlApplicationContext. While XML has been the traditional format for defining configuration metadata you can instruct the container to use Java annotations or code as the metadata format by providing a small amount of XML configuration to declaratively enable support for these additional metadata formats.

In most application scenarios, explicit user code is not required to instantiate one or more instances of a Spring IoC container. For example, in a web application scenario, a simple eight (or so) lines of boilerplate web descriptor XML in the web.xml file of the application will typically suffice (see Section 7.15.4, “Convenient ApplicationContext instantiation for web applications”). If you are using the Spring Tool Suite Eclipse-powered development environment this boilerplate configuration can be easily created with few mouse clicks or keystrokes.

The following diagram is a high-level view of how Spring works. Your application classes are combined with configuration metadata so that after theApplicationContext is created and initialized, you have a fully configured and executable system or application.

Figure 7.1. The Spring IoC container

Spring框架参考文档 Spring Framework Reference Documentation

7.2.1 Configuration metadata

As the preceding diagram shows, the Spring IoC container consumes a form of configuration metadata; this configuration metadata represents how you as an application developer tell the Spring container to instantiate, configure, and assemble the objects in your application.

Configuration metadata is traditionally supplied in a simple and intuitive XML format, which is what most of this chapter uses to convey key concepts and features of the Spring IoC container.

Spring框架参考文档 Spring Framework Reference Documentation

XML-based metadata is not the only allowed form of configuration metadata. The Spring IoC container itself is totally decoupled from the format in which this configuration metadata is actually written. These days many developers choose Java-based configuration for their Spring applications.

For information about using other forms of metadata with the Spring container, see:

  • Annotation-based configuration: Spring 2.5 introduced support for annotation-based configuration metadata.
  • Java-based configuration: Starting with Spring 3.0, many features provided by the Spring JavaConfig project became part of the core Spring Framework. Thus you can define beans external to your application classes by using Java rather than XML files. To use these new features, see the @Configuration, @Bean, @Importand @DependsOn annotations.

Spring configuration consists of at least one and typically more than one bean definition that the container must manage. XML-based configuration metadata shows these beans configured as <bean/> elements inside a top-level <beans/> element. Java configuration typically uses @Bean annotated methods within a @Configurationclass.

These bean definitions correspond to the actual objects that make up your application. Typically you define service layer objects, data access objects (DAOs), presentation objects such as Struts Action instances, infrastructure objects such as Hibernate SessionFactories, JMS Queues, and so forth. Typically one does not configure fine-grained domain objects in the container, because it is usually the responsibility of DAOs and business logic to create and load domain objects. However, you can use Spring’s integration with AspectJ to configure objects that have been created outside the control of an IoC container. See Using AspectJ to dependency-inject domain objects with Spring.

The following example shows the basic structure of XML-based configuration metadata:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd">

    <bean id="..." class="...">
        <!-- collaborators and configuration for this bean go here -->
    </bean>

    <bean id="..." class="...">
        <!-- collaborators and configuration for this bean go here -->
    </bean>

    <!-- more bean definitions go here -->

</beans>

The id attribute is a string that you use to identify the individual bean definition. The class attribute defines the type of the bean and uses the fully qualified classname. The value of the id attribute refers to collaborating objects. The XML for referring to collaborating objects is not shown in this example; see Dependencies for more information.

7.2.2 Instantiating a container

Instantiating a Spring IoC container is straightforward. The location path or paths supplied to an ApplicationContext constructor are actually resource strings that allow the container to load configuration metadata from a variety of external resources such as the local file system, from the Java CLASSPATH, and so on.

ApplicationContext context =
    new ClassPathXmlApplicationContext(new String[] {"services.xml", "daos.xml"});
Spring框架参考文档 Spring Framework Reference Documentation

After you learn about Spring’s IoC container, you may want to know more about Spring’s Resource abstraction, as described in Chapter 8, Resources, which provides a convenient mechanism for reading an InputStream from locations defined in a URI syntax. In particular, Resource paths are used to construct applications contexts as described in Section 8.7, “Application contexts and Resource paths”.

The following example shows the service layer objects (services.xml) configuration file:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd">

    <!-- services -->

    <bean id="petStore" class="org.springframework.samples.jpetstore.services.PetStoreServiceImpl">
        <property name="accountDao" ref="accountDao"/>
        <property name="itemDao" ref="itemDao"/>
        <!-- additional collaborators and configuration for this bean go here -->
    </bean>

    <!-- more bean definitions for services go here -->

</beans>

The following example shows the data access objects daos.xml file:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd">

    <bean id="accountDao"
        class="org.springframework.samples.jpetstore.dao.jpa.JpaAccountDao">
        <!-- additional collaborators and configuration for this bean go here -->
    </bean>

    <bean id="itemDao" class="org.springframework.samples.jpetstore.dao.jpa.JpaItemDao">
        <!-- additional collaborators and configuration for this bean go here -->
    </bean>

    <!-- more bean definitions for data access objects go here -->

</beans>

In the preceding example, the service layer consists of the class PetStoreServiceImpl, and two data access objects of the type JpaAccountDao and JpaItemDao(based on the JPA Object/Relational mapping standard). The property name element refers to the name of the JavaBean property, and the ref element refers to the name of another bean definition. This linkage between id and ref elements expresses the dependency between collaborating objects. For details of configuring an object’s dependencies, see Dependencies.

Composing XML-based configuration metadata

It can be useful to have bean definitions span multiple XML files. Often each individual XML configuration file represents a logical layer or module in your architecture.

You can use the application context constructor to load bean definitions from all these XML fragments. This constructor takes multiple Resource locations, as was shown in the previous section. Alternatively, use one or more occurrences of the <import/> element to load bean definitions from another file or files. For example:

<beans>
    <import resource="services.xml"/>
    <import resource="resources/messageSource.xml"/>
    <import resource="/resources/themeSource.xml"/>

    <bean id="bean1" class="..."/>
    <bean id="bean2" class="..."/>
</beans>

In the preceding example, external bean definitions are loaded from three files: services.xml, messageSource.xml, and themeSource.xml. All location paths are relative to the definition file doing the importing, so services.xml must be in the same directory or classpath location as the file doing the importing, whilemessageSource.xml and themeSource.xml must be in a resources location below the location of the importing file. As you can see, a leading slash is ignored, but given that these paths are relative, it is better form not to use the slash at all. The contents of the files being imported, including the top level <beans/> element, must be valid XML bean definitions according to the Spring Schema.

Spring框架参考文档 Spring Framework Reference Documentation

It is possible, but not recommended, to reference files in parent directories using a relative "../" path. Doing so creates a dependency on a file that is outside the current application. In particular, this reference is not recommended for "classpath:" URLs (for example, "classpath:../services.xml"), where the runtime resolution process chooses the "nearest" classpath root and then looks into its parent directory. Classpath configuration changes may lead to the choice of a different, incorrect directory.

You can always use fully qualified resource locations instead of relative paths: for example, "file:C:/config/services.xml" or "classpath:/config/services.xml". However, be aware that you are coupling your application’s configuration to specific absolute locations. It is generally preferable to keep an indirection for such absolute locations, for example, through "${…​}" placeholders that are resolved against JVM system properties at runtime.

7.2.3 Using the container

The ApplicationContext is the interface for an advanced factory capable of maintaining a registry of different beans and their dependencies. Using the methodT getBean(String name, Class<T> requiredType) you can retrieve instances of your beans.

The ApplicationContext enables you to read bean definitions and access them as follows:

// create and configure beans
ApplicationContext context =
    new ClassPathXmlApplicationContext(new String[] {"services.xml", "daos.xml"});

// retrieve configured instance
PetStoreService service = context.getBean("petStore", PetStoreService.class);

// use configured instance
List<String> userList = service.getUsernameList();

You use getBean() to retrieve instances of your beans. The ApplicationContext interface has a few other methods for retrieving beans, but ideally your application code should never use them. Indeed, your application code should have no calls to the getBean() method at all, and thus no dependency on Spring APIs at all. For example, Spring’s integration with web frameworks provides for dependency injection for various web framework classes such as controllers and JSF-managed beans.

7.3 Bean overview

A Spring IoC container manages one or more beans. These beans are created with the configuration metadata that you supply to the container, for example, in the form of XML <bean/> definitions.

Within the container itself, these bean definitions are represented as BeanDefinition objects, which contain (among other information) the following metadata:

  • A package-qualified class name: typically the actual implementation class of the bean being defined.
  • Bean behavioral configuration elements, which state how the bean should behave in the container (scope, lifecycle callbacks, and so forth).
  • References to other beans that are needed for the bean to do its work; these references are also called collaborators or dependencies.
  • Other configuration settings to set in the newly created object, for example, the number of connections to use in a bean that manages a connection pool, or the size limit of the pool.

This metadata translates to a set of properties that make up each bean definition.

Table 7.1. The bean definition

Property Explained in…​

class

Section 7.3.2, “Instantiating beans”

name

Section 7.3.1, “Naming beans”

scope

Section 7.5, “Bean scopes”

constructor arguments

Section 7.4.1, “Dependency Injection”

properties

Section 7.4.1, “Dependency Injection”

autowiring mode

Section 7.4.5, “Autowiring collaborators”

lazy-initialization mode

Section 7.4.4, “Lazy-initialized beans”

initialization method

the section called “Initialization callbacks”

destruction method

the section called “Destruction callbacks”

In addition to bean definitions that contain information on how to create a specific bean, the ApplicationContext implementations also permit the registration of existing objects that are created outside the container, by users. This is done by accessing the ApplicationContext’s BeanFactory via the method getBeanFactory()which returns the BeanFactory implementation DefaultListableBeanFactory. DefaultListableBeanFactory supports this registration through the methodsregisterSingleton(..) and registerBeanDefinition(..). However, typical applications work solely with beans defined through metadata bean definitions.

Spring框架参考文档 Spring Framework Reference Documentation

Bean metadata and manually supplied singleton instances need to be registered as early as possible, in order for the container to properly reason about them during autowiring and other introspection steps. While overriding of existing metadata and existing singleton instances is supported to some degree, the registration of new beans at runtime (concurrently with live access to factory) is not officially supported and may lead to concurrent access exceptions and/or inconsistent state in the bean container.

7.3.1 Naming beans

Every bean has one or more identifiers. These identifiers must be unique within the container that hosts the bean. A bean usually has only one identifier, but if it requires more than one, the extra ones can be considered aliases.

In XML-based configuration metadata, you use the id and/or name attributes to specify the bean identifier(s). The id attribute allows you to specify exactly one id. Conventionally these names are alphanumeric ('myBean', 'fooService', etc.), but may contain special characters as well. If you want to introduce other aliases to the bean, you can also specify them in the name attribute, separated by a comma (,), semicolon (;), or white space. As a historical note, in versions prior to Spring 3.1, the id attribute was defined as an xsd:ID type, which constrained possible characters. As of 3.1, it is defined as an xsd:string type. Note that bean id uniqueness is still enforced by the container, though no longer by XML parsers.

You are not required to supply a name or id for a bean. If no name or id is supplied explicitly, the container generates a unique name for that bean. However, if you want to refer to that bean by name, through the use of the ref element or Service Locator style lookup, you must provide a name. Motivations for not supplying a name are related to using inner beans and autowiring collaborators.

Bean Naming Conventions

The convention is to use the standard Java convention for instance field names when naming beans. That is, bean names start with a lowercase letter, and are camel-cased from then on. Examples of such names would be (without quotes) 'accountManager', 'accountService', 'userDao', 'loginController', and so forth.

Naming beans consistently makes your configuration easier to read and understand, and if you are using Spring AOP it helps a lot when applying advice to a set of beans related by name.

Spring框架参考文档 Spring Framework Reference Documentation

With component scanning in the classpath, Spring generates bean names for unnamed components, following the rules above: essentially, taking the simple class name and turning its initial character to lower-case. However, in the (unusual) special case when there is more than one character and both the first and second characters are upper case, the original casing gets preserved. These are the same rules as defined byjava.beans.Introspector.decapitalize (which Spring is using here).

Aliasing a bean outside the bean definition

In a bean definition itself, you can supply more than one name for the bean, by using a combination of up to one name specified by the id attribute, and any number of other names in the name attribute. These names can be equivalent aliases to the same bean, and are useful for some situations, such as allowing each component in an application to refer to a common dependency by using a bean name that is specific to that component itself.

Specifying all aliases where the bean is actually defined is not always adequate, however. It is sometimes desirable to introduce an alias for a bean that is defined elsewhere. This is commonly the case in large systems where configuration is split amongst each subsystem, each subsystem having its own set of object definitions. In XML-based configuration metadata, you can use the <alias/> element to accomplish this.

<alias name="fromName" alias="toName"/>

In this case, a bean in the same container which is named fromName, may also, after the use of this alias definition, be referred to as toName.

For example, the configuration metadata for subsystem A may refer to a DataSource via the name subsystemA-dataSource. The configuration metadata for subsystem B may refer to a DataSource via the name subsystemB-dataSource. When composing the main application that uses both these subsystems the main application refers to the DataSource via the name myApp-dataSource. To have all three names refer to the same object you add to the MyApp configuration metadata the following aliases definitions:

<alias name="subsystemA-dataSource" alias="subsystemB-dataSource"/>
<alias name="subsystemA-dataSource" alias="myApp-dataSource" />

Now each component and the main application can refer to the dataSource through a name that is unique and guaranteed not to clash with any other definition (effectively creating a namespace), yet they refer to the same bean.

Java-configuration

If you are using Java-configuration, the @Bean annotation can be used to provide aliases see Section 7.12.3, “Using the @Bean annotation” for details.

7.3.2 Instantiating beans

A bean definition essentially is a recipe for creating one or more objects. The container looks at the recipe for a named bean when asked, and uses the configuration metadata encapsulated by that bean definition to create (or acquire) an actual object.

If you use XML-based configuration metadata, you specify the type (or class) of object that is to be instantiated in the class attribute of the <bean/> element. Thisclass attribute, which internally is a Class property on a BeanDefinition instance, is usually mandatory. (For exceptions, see the section called “Instantiation using an instance factory method” and Section 7.7, “Bean definition inheritance”.) You use the Class property in one of two ways:

  • Typically, to specify the bean class to be constructed in the case where the container itself directly creates the bean by calling its constructor reflectively, somewhat equivalent to Java code using the new operator.
  • To specify the actual class containing the static factory method that will be invoked to create the object, in the less common case where the container invokes astatic factory method on a class to create the bean. The object type returned from the invocation of the static factory method may be the same class or another class entirely.
 

Inner class names.  If you want to configure a bean definition for a static nested class, you have to use the binary name of the nested class.

For example, if you have a class called Foo in the com.example package, and this Foo class has a static nested class called Bar, the value of the 'class'attribute on a bean definition would be…​

com.example.Foo$Bar

Notice the use of the $ character in the name to separate the nested class name from the outer class name.

Instantiation with a constructor

When you create a bean by the constructor approach, all normal classes are usable by and compatible with Spring. That is, the class being developed does not need to implement any specific interfaces or to be coded in a specific fashion. Simply specifying the bean class should suffice. However, depending on what type of IoC you use for that specific bean, you may need a default (empty) constructor.

The Spring IoC container can manage virtually any class you want it to manage; it is not limited to managing true JavaBeans. Most Spring users prefer actual JavaBeans with only a default (no-argument) constructor and appropriate setters and getters modeled after the properties in the container. You can also have more exotic non-bean-style classes in your container. If, for example, you need to use a legacy connection pool that absolutely does not adhere to the JavaBean specification, Spring can manage it as well.

With XML-based configuration metadata you can specify your bean class as follows:

<bean id="exampleBean" class="examples.ExampleBean"/>

<bean name="anotherExample" class="examples.ExampleBeanTwo"/>

For details about the mechanism for supplying arguments to the constructor (if required) and setting object instance properties after the object is constructed, seeInjecting Dependencies.

Instantiation with a static factory method

When defining a bean that you create with a static factory method, you use the class attribute to specify the class containing the static factory method and an attribute named factory-method to specify the name of the factory method itself. You should be able to call this method (with optional arguments as described later) and return a live object, which subsequently is treated as if it had been created through a constructor. One use for such a bean definition is to call static factories in legacy code.

The following bean definition specifies that the bean will be created by calling a factory-method. The definition does not specify the type (class) of the returned object, only the class containing the factory method. In this example, the createInstance() method must be a static method.

<bean id="clientService"
    class="examples.ClientService"
    factory-method="createInstance"/>
public class ClientService {
    private static ClientService clientService = new ClientService();
    private ClientService() {}

    public static ClientService createInstance() {
        return clientService;
    }
}

For details about the mechanism for supplying (optional) arguments to the factory method and setting object instance properties after the object is returned from the factory, see Dependencies and configuration in detail.

Instantiation using an instance factory method

Similar to instantiation through a static factory method, instantiation with an instance factory method invokes a non-static method of an existing bean from the container to create a new bean. To use this mechanism, leave the class attribute empty, and in the factory-bean attribute, specify the name of a bean in the current (or parent/ancestor) container that contains the instance method that is to be invoked to create the object. Set the name of the factory method itself with thefactory-method attribute.

<!-- the factory bean, which contains a method called createInstance() -->
<bean id="serviceLocator" class="examples.DefaultServiceLocator">
    <!-- inject any dependencies required by this locator bean -->
</bean>

<!-- the bean to be created via the factory bean -->
<bean id="clientService"
    factory-bean="serviceLocator"
    factory-method="createClientServiceInstance"/>
public class DefaultServiceLocator {

    private static ClientService clientService = new ClientServiceImpl();
    private DefaultServiceLocator() {}

    public ClientService createClientServiceInstance() {
        return clientService;
    }
}

One factory class can also hold more than one factory method as shown here:

<bean id="serviceLocator" class="examples.DefaultServiceLocator">
    <!-- inject any dependencies required by this locator bean -->
</bean>

<bean id="clientService"
    factory-bean="serviceLocator"
    factory-method="createClientServiceInstance"/>

<bean id="accountService"
    factory-bean="serviceLocator"
    factory-method="createAccountServiceInstance"/>
public class DefaultServiceLocator {

    private static ClientService clientService = new ClientServiceImpl();
    private static AccountService accountService = new AccountServiceImpl();

    private DefaultServiceLocator() {}

    public ClientService createClientServiceInstance() {
        return clientService;
    }

    public AccountService createAccountServiceInstance() {
        return accountService;
    }

}

This approach shows that the factory bean itself can be managed and configured through dependency injection (DI). See Dependencies and configuration in detail.

Spring框架参考文档 Spring Framework Reference Documentation

In Spring documentation, factory bean refers to a bean that is configured in the Spring container that will create objects through an instance or static factory method. By contrast, FactoryBean (notice the capitalization) refers to a Spring-specific FactoryBean.

7.4 Dependencies

A typical enterprise application does not consist of a single object (or bean in the Spring parlance). Even the simplest application has a few objects that work together to present what the end-user sees as a coherent application. This next section explains how you go from defining a number of bean definitions that stand alone to a fully realized application where objects collaborate to achieve a goal.

7.4.1 Dependency Injection

Dependency injection (DI) is a process whereby objects define their dependencies, that is, the other objects they work with, only through constructor arguments, arguments to a factory method, or properties that are set on the object instance after it is constructed or returned from a factory method. The container then injects those dependencies when it creates the bean. This process is fundamentally the inverse, hence the name Inversion of Control (IoC), of the bean itself controlling the instantiation or location of its dependencies on its own by using direct construction of classes, or the Service Locator pattern.

Code is cleaner with the DI principle and decoupling is more effective when objects are provided with their dependencies. The object does not look up its dependencies, and does not know the location or class of the dependencies. As such, your classes become easier to test, in particular when the dependencies are on interfaces or abstract base classes, which allow for stub or mock implementations to be used in unit tests.

DI exists in two major variants, Constructor-based dependency injection and Setter-based dependency injection.

Constructor-based dependency injection

Constructor-based DI is accomplished by the container invoking a constructor with a number of arguments, each representing a dependency. Calling a static factory method with specific arguments to construct the bean is nearly equivalent, and this discussion treats arguments to a constructor and to a static factory method similarly. The following example shows a class that can only be dependency-injected with constructor injection. Notice that there is nothing special about this class, it is a POJO that has no dependencies on container specific interfaces, base classes or annotations.

public class SimpleMovieLister {

    // the SimpleMovieLister has a dependency on a MovieFinder
    private MovieFinder movieFinder;

    // a constructor so that the Spring container can inject a MovieFinder
    public SimpleMovieLister(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // business logic that actually uses the injected MovieFinder is omitted...

}
Constructor argument resolution

Constructor argument resolution matching occurs using the argument’s type. If no potential ambiguity exists in the constructor arguments of a bean definition, then the order in which the constructor arguments are defined in a bean definition is the order in which those arguments are supplied to the appropriate constructor when the bean is being instantiated. Consider the following class:

package x.y;

public class Foo {

    public Foo(Bar bar, Baz baz) {
        // ...
    }

}

No potential ambiguity exists, assuming that Bar and Baz classes are not related by inheritance. Thus the following configuration works fine, and you do not need to specify the constructor argument indexes and/or types explicitly in the <constructor-arg/> element.

<beans>
    <bean id="foo" class="x.y.Foo">
        <constructor-arg ref="bar"/>
        <constructor-arg ref="baz"/>
    </bean>

    <bean id="bar" class="x.y.Bar"/>

    <bean id="baz" class="x.y.Baz"/>
</beans>

When another bean is referenced, the type is known, and matching can occur (as was the case with the preceding example). When a simple type is used, such as<value>true</value>, Spring cannot determine the type of the value, and so cannot match by type without help. Consider the following class:

package examples;

public class ExampleBean {

    // Number of years to calculate the Ultimate Answer
    private int years;

    // The Answer to Life, the Universe, and Everything
    private String ultimateAnswer;

    public ExampleBean(int years, String ultimateAnswer) {
        this.years = years;
        this.ultimateAnswer = ultimateAnswer;
    }

}

In the preceding scenario, the container can use type matching with simple types if you explicitly specify the type of the constructor argument using the type attribute. For example:

<bean id="exampleBean" class="examples.ExampleBean">
    <constructor-arg type="int" value="7500000"/>
    <constructor-arg type="java.lang.String" value="42"/>
</bean>

Use the index attribute to specify explicitly the index of constructor arguments. For example:

<bean id="exampleBean" class="examples.ExampleBean">
    <constructor-arg index="0" value="7500000"/>
    <constructor-arg index="1" value="42"/>
</bean>

In addition to resolving the ambiguity of multiple simple values, specifying an index resolves ambiguity where a constructor has two arguments of the same type. Note that the index is 0 based.

You can also use the constructor parameter name for value disambiguation:

<bean id="exampleBean" class="examples.ExampleBean">
    <constructor-arg name="years" value="7500000"/>
    <constructor-arg name="ultimateAnswer" value="42"/>
</bean>

Keep in mind that to make this work out of the box your code must be compiled with the debug flag enabled so that Spring can look up the parameter name from the constructor. If you can’t compile your code with debug flag (or don’t want to) you can use @ConstructorProperties JDK annotation to explicitly name your constructor arguments. The sample class would then have to look as follows:

package examples;

public class ExampleBean {

    // Fields omitted

    @ConstructorProperties({"years", "ultimateAnswer"})
    public ExampleBean(int years, String ultimateAnswer) {
        this.years = years;
        this.ultimateAnswer = ultimateAnswer;
    }

}

Setter-based dependency injection

Setter-based DI is accomplished by the container calling setter methods on your beans after invoking a no-argument constructor or no-argument static factory method to instantiate your bean.

The following example shows a class that can only be dependency-injected using pure setter injection. This class is conventional Java. It is a POJO that has no dependencies on container specific interfaces, base classes or annotations.

public class SimpleMovieLister {

    // the SimpleMovieLister has a dependency on the MovieFinder
    private MovieFinder movieFinder;

    // a setter method so that the Spring container can inject a MovieFinder
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // business logic that actually uses the injected MovieFinder is omitted...

}

The ApplicationContext supports constructor-based and setter-based DI for the beans it manages. It also supports setter-based DI after some dependencies have already been injected through the constructor approach. You configure the dependencies in the form of a BeanDefinition, which you use in conjunction withPropertyEditor instances to convert properties from one format to another. However, most Spring users do not work with these classes directly (i.e., programmatically) but rather with XML bean definitions, annotated components (i.e., classes annotated with @Component, @Controller, etc.), or @Bean methods in Java-based@Configuration classes. These sources are then converted internally into instances of BeanDefinition and used to load an entire Spring IoC container instance.

Constructor-based or setter-based DI?

Since you can mix constructor-based and setter-based DI, it is a good rule of thumb to use constructors for mandatory dependencies and setter methods or configuration methods for optional dependencies. Note that use of the @Required annotation on a setter method can be used to make the property a required dependency.

The Spring team generally advocates constructor injection as it enables one to implement application components as immutable objects and to ensure that required dependencies are not null. Furthermore constructor-injected components are always returned to client (calling) code in a fully initialized state. As a side note, a large number of constructor arguments is a bad code smell, implying that the class likely has too many responsibilities and should be refactored to better address proper separation of concerns.

Setter injection should primarily only be used for optional dependencies that can be assigned reasonable default values within the class. Otherwise, not-null checks must be performed everywhere the code uses the dependency. One benefit of setter injection is that setter methods make objects of that class amenable to reconfiguration or re-injection later. Management through JMX MBeans is therefore a compelling use case for setter injection.

Use the DI style that makes the most sense for a particular class. Sometimes, when dealing with third-party classes for which you do not have the source, the choice is made for you. For example, if a third-party class does not expose any setter methods, then constructor injection may be the only available form of DI.

Dependency resolution process

The container performs bean dependency resolution as follows:

  • The ApplicationContext is created and initialized with configuration metadata that describes all the beans. Configuration metadata can be specified via XML, Java code, or annotations.
  • For each bean, its dependencies are expressed in the form of properties, constructor arguments, or arguments to the static-factory method if you are using that instead of a normal constructor. These dependencies are provided to the bean, when the bean is actually created.
  • Each property or constructor argument is an actual definition of the value to set, or a reference to another bean in the container.
  • Each property or constructor argument which is a value is converted from its specified format to the actual type of that property or constructor argument. By default Spring can convert a value supplied in string format to all built-in types, such as int, long, String, boolean, etc.

The Spring container validates the configuration of each bean as the container is created. However, the bean properties themselves are not set until the bean is actually created. Beans that are singleton-scoped and set to be pre-instantiated (the default) are created when the container is created. Scopes are defined in Section 7.5, “Bean scopes”. Otherwise, the bean is created only when it is requested. Creation of a bean potentially causes a graph of beans to be created, as the bean’s dependencies and its dependencies' dependencies (and so on) are created and assigned. Note that resolution mismatches among those dependencies may show up late, i.e. on first creation of the affected bean.

Circular dependencies

If you use predominantly constructor injection, it is possible to create an unresolvable circular dependency scenario.

For example: Class A requires an instance of class B through constructor injection, and class B requires an instance of class A through constructor injection. If you configure beans for classes A and B to be injected into each other, the Spring IoC container detects this circular reference at runtime, and throws aBeanCurrentlyInCreationException.

One possible solution is to edit the source code of some classes to be configured by setters rather than constructors. Alternatively, avoid constructor injection and use setter injection only. In other words, although it is not recommended, you can configure circular dependencies with setter injection.

Unlike the typical case (with no circular dependencies), a circular dependency between bean A and bean B forces one of the beans to be injected into the other prior to being fully initialized itself (a classic chicken/egg scenario).

You can generally trust Spring to do the right thing. It detects configuration problems, such as references to non-existent beans and circular dependencies, at container load-time. Spring sets properties and resolves dependencies as late as possible, when the bean is actually created. This means that a Spring container which has loaded correctly can later generate an exception when you request an object if there is a problem creating that object or one of its dependencies. For example, the bean throws an exception as a result of a missing or invalid property. This potentially delayed visibility of some configuration issues is why ApplicationContext implementations by default pre-instantiate singleton beans. At the cost of some upfront time and memory to create these beans before they are actually needed, you discover configuration issues when the ApplicationContext is created, not later. You can still override this default behavior so that singleton beans will lazy-initialize, rather than be pre-instantiated.

If no circular dependencies exist, when one or more collaborating beans are being injected into a dependent bean, each collaborating bean is totally configured prior to being injected into the dependent bean. This means that if bean A has a dependency on bean B, the Spring IoC container completely configures bean B prior to invoking the setter method on bean A. In other words, the bean is instantiated (if not a pre-instantiated singleton), its dependencies are set, and the relevant lifecycle methods (such as a configured init method or the InitializingBean callback method) are invoked.

Examples of dependency injection

The following example uses XML-based configuration metadata for setter-based DI. A small part of a Spring XML configuration file specifies some bean definitions:

<bean id="exampleBean" class="examples.ExampleBean">
    <!-- setter injection using the nested ref element -->
    <property name="beanOne">
        <ref bean="anotherExampleBean"/>
    </property>

    <!-- setter injection using the neater ref attribute -->
    <property name="beanTwo" ref="yetAnotherBean"/>
    <property name="integerProperty" value="1"/>
</bean>

<bean id="anotherExampleBean" class="examples.AnotherBean"/>
<bean id="yetAnotherBean" class="examples.YetAnotherBean"/>
public class ExampleBean {

    private AnotherBean beanOne;
    private YetAnotherBean beanTwo;
    private int i;

    public void setBeanOne(AnotherBean beanOne) {
        this.beanOne = beanOne;
    }

    public void setBeanTwo(YetAnotherBean beanTwo) {
        this.beanTwo = beanTwo;
    }

    public void setIntegerProperty(int i) {
        this.i = i;
    }

}

In the preceding example, setters are declared to match against the properties specified in the XML file. The following example uses constructor-based DI:

<bean id="exampleBean" class="examples.ExampleBean">
    <!-- constructor injection using the nested ref element -->
    <constructor-arg>
        <ref bean="anotherExampleBean"/>
    </constructor-arg>

    <!-- constructor injection using the neater ref attribute -->
    <constructor-arg ref="yetAnotherBean"/>

    <constructor-arg type="int" value="1"/>
</bean>

<bean id="anotherExampleBean" class="examples.AnotherBean"/>
<bean id="yetAnotherBean" class="examples.YetAnotherBean"/>
public class ExampleBean {

    private AnotherBean beanOne;
    private YetAnotherBean beanTwo;
    private int i;

    public ExampleBean(
        AnotherBean anotherBean, YetAnotherBean yetAnotherBean, int i) {
        this.beanOne = anotherBean;
        this.beanTwo = yetAnotherBean;
        this.i = i;
    }

}

The constructor arguments specified in the bean definition will be used as arguments to the constructor of the ExampleBean.

Now consider a variant of this example, where instead of using a constructor, Spring is told to call a static factory method to return an instance of the object:

<bean id="exampleBean" class="examples.ExampleBean" factory-method="createInstance">
    <constructor-arg ref="anotherExampleBean"/>
    <constructor-arg ref="yetAnotherBean"/>
    <constructor-arg value="1"/>
</bean>

<bean id="anotherExampleBean" class="examples.AnotherBean"/>
<bean id="yetAnotherBean" class="examples.YetAnotherBean"/>
public class ExampleBean {

    // a private constructor
    private ExampleBean(...) {
        ...
    }

    // a static factory method; the arguments to this method can be
    // considered the dependencies of the bean that is returned,
    // regardless of how those arguments are actually used.
    public static ExampleBean createInstance (
        AnotherBean anotherBean, YetAnotherBean yetAnotherBean, int i) {

        ExampleBean eb = new ExampleBean (...);
        // some other operations...
        return eb;
    }

}

Arguments to the static factory method are supplied via <constructor-arg/> elements, exactly the same as if a constructor had actually been used. The type of the class being returned by the factory method does not have to be of the same type as the class that contains the static factory method, although in this example it is. An instance (non-static) factory method would be used in an essentially identical fashion (aside from the use of the factory-bean attribute instead of the class attribute), so details will not be discussed here.

7.4.2 Dependencies and configuration in detail

As mentioned in the previous section, you can define bean properties and constructor arguments as references to other managed beans (collaborators), or as values defined inline. Spring’s XML-based configuration metadata supports sub-element types within its <property/> and <constructor-arg/> elements for this purpose.

Straight values (primitives, Strings, and so on)

The value attribute of the <property/> element specifies a property or constructor argument as a human-readable string representation. Spring’s conversion serviceis used to convert these values from a String to the actual type of the property or argument.

<bean id="myDataSource" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
    <!-- results in a setDriverClassName(String) call -->
    <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
    <property name="url" value="jdbc:mysql://localhost:3306/mydb"/>
    <property name="username" value="root"/>
    <property name="password" value="masterkaoli"/>
</bean>

The following example uses the p-namespace for even more succinct XML configuration.

<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:p="http://www.springframework.org/schema/p"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
    http://www.springframework.org/schema/beans/spring-beans.xsd">

    <bean id="myDataSource" class="org.apache.commons.dbcp.BasicDataSource"
        destroy-method="close"
        p:driverClassName="com.mysql.jdbc.Driver"
        p:url="jdbc:mysql://localhost:3306/mydb"
        p:username="root"
        p:password="masterkaoli"/>

</beans>

The preceding XML is more succinct; however, typos are discovered at runtime rather than design time, unless you use an IDE such as IntelliJ IDEA or the Spring Tool Suite (STS) that support automatic property completion when you create bean definitions. Such IDE assistance is highly recommended.

You can also configure a java.util.Properties instance as:

<bean id="mappings"
    class="org.springframework.beans.factory.config.PropertyPlaceholderConfigurer">

    <!-- typed as a java.util.Properties -->
    <property name="properties">
        <value>
            jdbc.driver.className=com.mysql.jdbc.Driver
            jdbc.url=jdbc:mysql://localhost:3306/mydb
        </value>
    </property>
</bean>

The Spring container converts the text inside the <value/> element into a java.util.Properties instance by using the JavaBeans PropertyEditor mechanism. This is a nice shortcut, and is one of a few places where the Spring team do favor the use of the nested <value/> element over the value attribute style.

The idref element

The idref element is simply an error-proof way to pass the id (string value - not a reference) of another bean in the container to a <constructor-arg/> or<property/> element.

<bean id="theTargetBean" class="..."/>

<bean id="theClientBean" class="...">
    <property name="targetName">
        <idref bean="theTargetBean" />
    </property>
</bean>

The above bean definition snippet is exactly equivalent (at runtime) to the following snippet:

<bean id="theTargetBean" class="..." />

<bean id="client" class="...">
    <property name="targetName" value="theTargetBean" />
</bean>

The first form is preferable to the second, because using the idref tag allows the container to validate at deployment time that the referenced, named bean actually exists. In the second variation, no validation is performed on the value that is passed to the targetName property of the client bean. Typos are only discovered (with most likely fatal results) when the client bean is actually instantiated. If the client bean is a prototype bean, this typo and the resulting exception may only be discovered long after the container is deployed.

Spring框架参考文档 Spring Framework Reference Documentation

The local attribute on the idref element is no longer supported in the 4.0 beans xsd since it does not provide value over a regular bean reference anymore. Simply change your existing idref local references to idref bean when upgrading to the 4.0 schema.

A common place (at least in versions earlier than Spring 2.0) where the <idref/> element brings value is in the configuration of AOP interceptors in aProxyFactoryBean bean definition. Using <idref/> elements when you specify the interceptor names prevents you from misspelling an interceptor id.

References to other beans (collaborators)

The ref element is the final element inside a <constructor-arg/> or <property/> definition element. Here you set the value of the specified property of a bean to be a reference to another bean (a collaborator) managed by the container. The referenced bean is a dependency of the bean whose property will be set, and it is initialized on demand as needed before the property is set. (If the collaborator is a singleton bean, it may be initialized already by the container.) All references are ultimately a reference to another object. Scoping and validation depend on whether you specify the id/name of the other object through the bean, local, or parentattributes.

Specifying the target bean through the bean attribute of the <ref/> tag is the most general form, and allows creation of a reference to any bean in the same container or parent container, regardless of whether it is in the same XML file. The value of the bean attribute may be the same as the id attribute of the target bean, or as one of the values in the name attribute of the target bean.

<ref bean="someBean"/>

Specifying the target bean through the parent attribute creates a reference to a bean that is in a parent container of the current container. The value of the parentattribute may be the same as either the id attribute of the target bean, or one of the values in the name attribute of the target bean, and the target bean must be in a parent container of the current one. You use this bean reference variant mainly when you have a hierarchy of containers and you want to wrap an existing bean in a parent container with a proxy that will have the same name as the parent bean.

<!-- in the parent context -->
<bean id="accountService" class="com.foo.SimpleAccountService">
    <!-- insert dependencies as required as here -->
</bean>
<!-- in the child (descendant) context -->
<bean id="accountService" <!-- bean name is the same as the parent bean -->
    class="org.springframework.aop.framework.ProxyFactoryBean">
    <property name="target">
        <ref parent="accountService"/> <!-- notice how we refer to the parent bean -->
    </property>
    <!-- insert other configuration and dependencies as required here -->
</bean>
Spring框架参考文档 Spring Framework Reference Documentation

The local attribute on the ref element is no longer supported in the 4.0 beans xsd since it does not provide value over a regular bean reference anymore. Simply change your existing ref local references to ref bean when upgrading to the 4.0 schema.

Inner beans

A <bean/> element inside the <property/> or <constructor-arg/> elements defines a so-called inner bean.

<bean id="outer" class="...">
    <!-- instead of using a reference to a target bean, simply define the target bean inline -->
    <property name="target">
        <bean class="com.example.Person"> <!-- this is the inner bean -->
            <property name="name" value="Fiona Apple"/>
            <property name="age" value="25"/>
        </bean>
    </property>
</bean>

An inner bean definition does not require a defined id or name; if specified, the container does not use such a value as an identifier. The container also ignores thescope flag on creation: Inner beans are always anonymous and they are always created with the outer bean. It is not possible to inject inner beans into collaborating beans other than into the enclosing bean or to access them independently.

As a corner case, it is possible to receive destruction callbacks from a custom scope, e.g. for a request-scoped inner bean contained within a singleton bean: The creation of the inner bean instance will be tied to its containing bean, but destruction callbacks allow it to participate in the request scope’s lifecycle. This is not a common scenario; inner beans typically simply share their containing bean’s scope.

Collections

In the <list/>, <set/>, <map/>, and <props/> elements, you set the properties and arguments of the Java Collection types List, Set, Map, andProperties, respectively.

<bean id="moreComplexObject" class="example.ComplexObject">
    <!-- results in a setAdminEmails(java.util.Properties) call -->
    <property name="adminEmails">
        <props>
            <prop key="administrator">administrator@example.org</prop>
            <prop key="support">support@example.org</prop>
            <prop key="development">development@example.org</prop>
        </props>
    </property>
    <!-- results in a setSomeList(java.util.List) call -->
    <property name="someList">
        <list>
            <value>a list element followed by a reference</value>
            <ref bean="myDataSource" />
        </list>
    </property>
    <!-- results in a setSomeMap(java.util.Map) call -->
    <property name="someMap">
        <map>
            <entry key="an entry" value="just some string"/>
            <entry key ="a ref" value-ref="myDataSource"/>
        </map>
    </property>
    <!-- results in a setSomeSet(java.util.Set) call -->
    <property name="someSet">
        <set>
            <value>just some string</value>
            <ref bean="myDataSource" />
        </set>
    </property>
</bean>

The value of a map key or value, or a set value, can also again be any of the following elements:

bean | ref | idref | list | set | map | props | value | null
Collection merging

The Spring container also supports the merging of collections. An application developer can define a parent-style <list/>, <map/>, <set/> or <props/> element, and have child-style <list/>, <map/>, <set/> or <props/> elements inherit and override values from the parent collection. That is, the child collection’s values are the result of merging the elements of the parent and child collections, with the child’s collection elements overriding values specified in the parent collection.

This section on merging discusses the parent-child bean mechanism. Readers unfamiliar with parent and child bean definitions may wish to read the relevant sectionbefore continuing.

The following example demonstrates collection merging:

<beans>
    <bean id="parent" abstract="true" class="example.ComplexObject">
        <property name="adminEmails">
            <props>
                <prop key="administrator">administrator@example.com</prop>
                <prop key="support">support@example.com</prop>
            </props>
        </property>
    </bean>
    <bean id="child" parent="parent">
        <property name="adminEmails">
            <!-- the merge is specified on the child collection definition -->
            <props merge="true">
                <prop key="sales">sales@example.com</prop>
                <prop key="support">support@example.co.uk</prop>
            </props>
        </property>
    </bean>
<beans>

Notice the use of the merge=true attribute on the <props/> element of the adminEmails property of the child bean definition. When the child bean is resolved and instantiated by the container, the resulting instance has an adminEmails Properties collection that contains the result of the merging of the child’s adminEmailscollection with the parent’s adminEmails collection.

administrator=administrator@example.com
sales=sales@example.com
support=support@example.co.uk

The child Properties collection’s value set inherits all property elements from the parent <props/>, and the child’s value for the support value overrides the value in the parent collection.

This merging behavior applies similarly to the <list/>, <map/>, and <set/> collection types. In the specific case of the <list/> element, the semantics associated with the List collection type, that is, the notion of an ordered collection of values, is maintained; the parent’s values precede all of the child list’s values. In the case of the Map, Set, and Properties collection types, no ordering exists. Hence no ordering semantics are in effect for the collection types that underlie the associatedMap, Set, and Properties implementation types that the container uses internally.

Limitations of collection merging

You cannot merge different collection types (such as a Map and a List), and if you do attempt to do so an appropriate Exception is thrown. The merge attribute must be specified on the lower, inherited, child definition; specifying the merge attribute on a parent collection definition is redundant and will not result in the desired merging.

Strongly-typed collection

With the introduction of generic types in Java 5, you can use strongly typed collections. That is, it is possible to declare a Collection type such that it can only containString elements (for example). If you are using Spring to dependency-inject a strongly-typed Collection into a bean, you can take advantage of Spring’s type-conversion support such that the elements of your strongly-typed Collection instances are converted to the appropriate type prior to being added to the Collection.

public class Foo {

    private Map<String, Float> accounts;

    public void setAccounts(Map<String, Float> accounts) {
        this.accounts = accounts;
    }
}
<beans>
    <bean id="foo" class="x.y.Foo">
        <property name="accounts">
            <map>
                <entry key="one" value="9.99"/>
                <entry key="two" value="2.75"/>
                <entry key="six" value="3.99"/>
            </map>
        </property>
    </bean>
</beans>

When the accounts property of the foo bean is prepared for injection, the generics information about the element type of the strongly-typed Map<String, Float> is available by reflection. Thus Spring’s type conversion infrastructure recognizes the various value elements as being of type Float, and the string values 9.99, 2.75, and 3.99 are converted into an actual Float type.

Null and empty string values

Spring treats empty arguments for properties and the like as empty Strings. The following XML-based configuration metadata snippet sets the email property to the empty String value ("").

<bean class="ExampleBean">
    <property name="email" value=""/>
</bean>

The preceding example is equivalent to the following Java code:

exampleBean.setEmail("")

The <null/> element handles null values. For example:

<bean class="ExampleBean">
    <property name="email">
        <null/>
    </property>
</bean>

The above configuration is equivalent to the following Java code:

exampleBean.setEmail(null)

XML shortcut with the p-namespace

The p-namespace enables you to use the bean element’s attributes, instead of nested <property/> elements, to describe your property values and/or collaborating beans.

Spring supports extensible configuration formats with namespaces, which are based on an XML Schema definition. The beans configuration format discussed in this chapter is defined in an XML Schema document. However, the p-namespace is not defined in an XSD file and exists only in the core of Spring.

The following example shows two XML snippets that resolve to the same result: The first uses standard XML format and the second uses the p-namespace.

<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:p="http://www.springframework.org/schema/p"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd">

    <bean name="classic" class="com.example.ExampleBean">
        <property name="email" value="foo@bar.com"/>
    </bean>

    <bean name="p-namespace" class="com.example.ExampleBean"
        p:email="foo@bar.com"/>
</beans>

The example shows an attribute in the p-namespace called email in the bean definition. This tells Spring to include a property declaration. As previously mentioned, the p-namespace does not have a schema definition, so you can set the name of the attribute to the property name.

This next example includes two more bean definitions that both have a reference to another bean:

<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:p="http://www.springframework.org/schema/p"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd">

    <bean name="john-classic" class="com.example.Person">
        <property name="name" value="John Doe"/>
        <property name="spouse" ref="jane"/>
    </bean>

    <bean name="john-modern"
        class="com.example.Person"
        p:name="John Doe"
        p:spouse-ref="jane"/>

    <bean name="jane" class="com.example.Person">
        <property name="name" value="Jane Doe"/>
    </bean>
</beans>

As you can see, this example includes not only a property value using the p-namespace, but also uses a special format to declare property references. Whereas the first bean definition uses <property name="spouse" ref="jane"/> to create a reference from bean john to bean jane, the second bean definition usesp:spouse-ref="jane" as an attribute to do the exact same thing. In this case spouse is the property name, whereas the -ref part indicates that this is not a straight value but rather a reference to another bean.

Spring框架参考文档 Spring Framework Reference Documentation

The p-namespace is not as flexible as the standard XML format. For example, the format for declaring property references clashes with properties that end in Ref, whereas the standard XML format does not. We recommend that you choose your approach carefully and communicate this to your team members, to avoid producing XML documents that use all three approaches at the same time.

XML shortcut with the c-namespace

Similar to the the section called “XML shortcut with the p-namespace”, the c-namespace, newly introduced in Spring 3.1, allows usage of inlined attributes for configuring the constructor arguments rather then nested constructor-arg elements.

Let’s review the examples from the section called “Constructor-based dependency injection” with the c: namespace:

<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:c="http://www.springframework.org/schema/c"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd">

    <bean id="bar" class="x.y.Bar"/>
    <bean id="baz" class="x.y.Baz"/>

    <!-- traditional declaration -->
    <bean id="foo" class="x.y.Foo">
        <constructor-arg ref="bar"/>
        <constructor-arg ref="baz"/>
        <constructor-arg value="foo@bar.com"/>
    </bean>

    <!-- c-namespace declaration -->
    <bean id="foo" class="x.y.Foo" c:bar-ref="bar" c:baz-ref="baz" c:email="foo@bar.com"/>

</beans>

The c: namespace uses the same conventions as the p: one (trailing -ref for bean references) for setting the constructor arguments by their names. And just as well, it needs to be declared even though it is not defined in an XSD schema (but it exists inside the Spring core).

For the rare cases where the constructor argument names are not available (usually if the bytecode was compiled without debugging information), one can use fallback to the argument indexes:

<!-- c-namespace index declaration -->
<bean id="foo" class="x.y.Foo" c:_0-ref="bar" c:_1-ref="baz"/>
Spring框架参考文档 Spring Framework Reference Documentation

Due to the XML grammar, the index notation requires the presence of the leading _ as XML attribute names cannot start with a number (even though some IDE allow it).

In practice, the constructor resolution mechanism is quite efficient in matching arguments so unless one really needs to, we recommend using the name notation through-out your configuration.

Compound property names

You can use compound or nested property names when you set bean properties, as long as all components of the path except the final property name are not null. Consider the following bean definition.

<bean id="foo" class="foo.Bar">
    <property name="fred.bob.sammy" value="123" />
</bean>

The foo bean has a fred property, which has a bob property, which has a sammy property, and that final sammy property is being set to the value 123. In order for this to work, the fred property of foo, and the bob property of fred must not be null after the bean is constructed, or a NullPointerException is thrown.

7.4.3 Using depends-on

If a bean is a dependency of another that usually means that one bean is set as a property of another. Typically you accomplish this with the <ref/> element in XML-based configuration metadata. However, sometimes dependencies between beans are less direct; for example, a static initializer in a class needs to be triggered, such as database driver registration. The depends-on attribute can explicitly force one or more beans to be initialized before the bean using this element is initialized. The following example uses the depends-on attribute to express a dependency on a single bean:

<bean id="beanOne" class="ExampleBean" depends-on="manager"/>
<bean id="manager" class="ManagerBean" />

To express a dependency on multiple beans, supply a list of bean names as the value of the depends-on attribute, with commas, whitespace and semicolons, used as valid delimiters:

<bean id="beanOne" class="ExampleBean" depends-on="manager,accountDao">
    <property name="manager" ref="manager" />
</bean>

<bean id="manager" class="ManagerBean" />
<bean id="accountDao" class="x.y.jdbc.JdbcAccountDao" />
Spring框架参考文档 Spring Framework Reference Documentation

The depends-on attribute in the bean definition can specify both an initialization time dependency and, in the case of singleton beans only, a corresponding destroy time dependency. Dependent beans that define a depends-on relationship with a given bean are destroyed first, prior to the given bean itself being destroyed. Thus depends-on can also control shutdown order.

7.4.4 Lazy-initialized beans

By default, ApplicationContext implementations eagerly create and configure all singleton beans as part of the initialization process. Generally, this pre-instantiation is desirable, because errors in the configuration or surrounding environment are discovered immediately, as opposed to hours or even days later. When this behavior isnot desirable, you can prevent pre-instantiation of a singleton bean by marking the bean definition as lazy-initialized. A lazy-initialized bean tells the IoC container to create a bean instance when it is first requested, rather than at startup.

In XML, this behavior is controlled by the lazy-init attribute on the <bean/> element; for example:

<bean id="lazy" class="com.foo.ExpensiveToCreateBean" lazy-init="true"/>
<bean name="not.lazy" class="com.foo.AnotherBean"/>

When the preceding configuration is consumed by an ApplicationContext, the bean named lazy is not eagerly pre-instantiated when the ApplicationContext is starting up, whereas the not.lazy bean is eagerly pre-instantiated.

However, when a lazy-initialized bean is a dependency of a singleton bean that is not lazy-initialized, the ApplicationContext creates the lazy-initialized bean at startup, because it must satisfy the singleton’s dependencies. The lazy-initialized bean is injected into a singleton bean elsewhere that is not lazy-initialized.

You can also control lazy-initialization at the container level by using the default-lazy-init attribute on the <beans/> element; for example:

<beans default-lazy-init="true">
    <!-- no beans will be pre-instantiated... -->
</beans>

7.4.5 Autowiring collaborators

The Spring container can autowire relationships between collaborating beans. You can allow Spring to resolve collaborators (other beans) automatically for your bean by inspecting the contents of the ApplicationContext. Autowiring has the following advantages:

  • Autowiring can significantly reduce the need to specify properties or constructor arguments. (Other mechanisms such as a bean template discussed elsewhere in this chapter are also valuable in this regard.)
  • Autowiring can update a configuration as your objects evolve. For example, if you need to add a dependency to a class, that dependency can be satisfied automatically without you needing to modify the configuration. Thus autowiring can be especially useful during development, without negating the option of switching to explicit wiring when the code base becomes more stable.

When using XML-based configuration metadata [2], you specify autowire mode for a bean definition with the autowire attribute of the <bean/> element. The autowiring functionality has four modes. You specify autowiring per bean and thus can choose which ones to autowire.

Table 7.2. Autowiring modes

Mode Explanation

no

(Default) No autowiring. Bean references must be defined via a ref element. Changing the default setting is not recommended for larger deployments, because specifying collaborators explicitly gives greater control and clarity. To some extent, it documents the structure of a system.

byName

Autowiring by property name. Spring looks for a bean with the same name as the property that needs to be autowired. For example, if a bean definition is set to autowire by name, and it contains a master property (that is, it has a setMaster(..) method), Spring looks for a bean definition named master, and uses it to set the property.

byType

Allows a property to be autowired if exactly one bean of the property type exists in the container. If more than one exists, a fatal exception is thrown, which indicates that you may not use byType autowiring for that bean. If there are no matching beans, nothing happens; the property is not set.

constructor

Analogous to byType, but applies to constructor arguments. If there is not exactly one bean of the constructor argument type in the container, a fatal error is raised.

With byType or constructor autowiring mode, you can wire arrays and typed-collections. In such cases all autowire candidates within the container that match the expected type are provided to satisfy the dependency. You can autowire strongly-typed Maps if the expected key type is String. An autowired Maps values will consist of all bean instances that match the expected type, and the Maps keys will contain the corresponding bean names.

You can combine autowire behavior with dependency checking, which is performed after autowiring completes.

Limitations and disadvantages of autowiring

Autowiring works best when it is used consistently across a project. If autowiring is not used in general, it might be confusing to developers to use it to wire only one or two bean definitions.

Consider the limitations and disadvantages of autowiring:

  • Explicit dependencies in property and constructor-arg settings always override autowiring. You cannot autowire so-called simple properties such as primitives,Strings, and Classes (and arrays of such simple properties). This limitation is by-design.
  • Autowiring is less exact than explicit wiring. Although, as noted in the above table, Spring is careful to avoid guessing in case of ambiguity that might have unexpected results, the relationships between your Spring-managed objects are no longer documented explicitly.
  • Wiring information may not be available to tools that may generate documentation from a Spring container.
  • Multiple bean definitions within the container may match the type specified by the setter method or constructor argument to be autowired. For arrays, collections, or Maps, this is not necessarily a problem. However for dependencies that expect a single value, this ambiguity is not arbitrarily resolved. If no unique bean definition is available, an exception is thrown.

In the latter scenario, you have several options:

  • Abandon autowiring in favor of explicit wiring.
  • Avoid autowiring for a bean definition by setting its autowire-candidate attributes to false as described in the next section.
  • Designate a single bean definition as the primary candidate by setting the primary attribute of its <bean/> element to true.
  • Implement the more fine-grained control available with annotation-based configuration, as described in Section 7.9, “Annotation-based container configuration”.

Excluding a bean from autowiring

On a per-bean basis, you can exclude a bean from autowiring. In Spring’s XML format, set the autowire-candidate attribute of the <bean/> element to false; the container makes that specific bean definition unavailable to the autowiring infrastructure (including annotation style configurations such as @Autowired).

You can also limit autowire candidates based on pattern-matching against bean names. The top-level <beans/> element accepts one or more patterns within itsdefault-autowire-candidates attribute. For example, to limit autowire candidate status to any bean whose name ends with Repository, provide a value of *Repository. To provide multiple patterns, define them in a comma-separated list. An explicit value of true or false for a bean definitions autowire-candidateattribute always takes precedence, and for such beans, the pattern matching rules do not apply.

These techniques are useful for beans that you never want to be injected into other beans by autowiring. It does not mean that an excluded bean cannot itself be configured using autowiring. Rather, the bean itself is not a candidate for autowiring other beans.

7.4.6 Method injection

In most application scenarios, most beans in the container are singletons. When a singleton bean needs to collaborate with another singleton bean, or a non-singleton bean needs to collaborate with another non-singleton bean, you typically handle the dependency by defining one bean as a property of the other. A problem arises when the bean lifecycles are different. Suppose singleton bean A needs to use non-singleton (prototype) bean B, perhaps on each method invocation on A. The container only creates the singleton bean A once, and thus only gets one opportunity to set the properties. The container cannot provide bean A with a new instance of bean B every time one is needed.

A solution is to forego some inversion of control. You can make bean A aware of the container by implementing the ApplicationContextAware interface, and bymaking a getBean("B") call to the container ask for (a typically new) bean B instance every time bean A needs it. The following is an example of this approach:

// a class that uses a stateful Command-style class to perform some processing
package fiona.apple;

// Spring-API imports
import org.springframework.beans.BeansException;
import org.springframework.context.ApplicationContext;
import org.springframework.context.ApplicationContextAware;

public class CommandManager implements ApplicationContextAware {

    private ApplicationContext applicationContext;

    public Object process(Map commandState) {
        // grab a new instance of the appropriate Command
        Command command = createCommand();
        // set the state on the (hopefully brand new) Command instance
        command.setState(commandState);
        return command.execute();
    }

    protected Command createCommand() {
        // notice the Spring API dependency!
        return this.applicationContext.getBean("command", Command.class);
    }

    public void setApplicationContext(
            ApplicationContext applicationContext) throws BeansException {
        this.applicationContext = applicationContext;
    }
}

The preceding is not desirable, because the business code is aware of and coupled to the Spring Framework. Method Injection, a somewhat advanced feature of the Spring IoC container, allows this use case to be handled in a clean fashion.

 

You can read more about the motivation for Method Injection in this blog entry.

Lookup method injection

Lookup method injection is the ability of the container to override methods on container managed beans, to return the lookup result for another named bean in the container. The lookup typically involves a prototype bean as in the scenario described in the preceding section. The Spring Framework implements this method injection by using bytecode generation from the CGLIB library to generate dynamically a subclass that overrides the method.

Spring框架参考文档 Spring Framework Reference Documentation
  • For this dynamic subclassing to work, the class that the Spring bean container will subclass cannot be final, and the method to be overridden cannot be final either.
  • Unit-testing a class that has an abstract method requires you to subclass the class yourself and to supply a stub implementation of the abstractmethod.
  • Concrete methods are also necessary for component scanning which requires concrete classes to pick up.
  • A further key limitation is that lookup methods won’t work with factory methods and in particular not with @Bean methods in configuration classes, since the container is not in charge of creating the instance in that case and therefore cannot create a runtime-generated subclass on the fly.
  • Finally, objects that have been the target of method injection cannot be serialized.

Looking at the CommandManager class in the previous code snippet, you see that the Spring container will dynamically override the implementation of thecreateCommand() method. Your CommandManager class will not have any Spring dependencies, as can be seen in the reworked example:

package fiona.apple;

// no more Spring imports!

public abstract class CommandManager {

    public Object process(Object commandState) {
        // grab a new instance of the appropriate Command interface
        Command command = createCommand();
        // set the state on the (hopefully brand new) Command instance
        command.setState(commandState);
        return command.execute();
    }

    // okay... but where is the implementation of this method?
    protected abstract Command createCommand();
}

In the client class containing the method to be injected (the CommandManager in this case), the method to be injected requires a signature of the following form:

<public|protected> [abstract] <return-type> theMethodName(no-arguments);

If the method is abstract, the dynamically-generated subclass implements the method. Otherwise, the dynamically-generated subclass overrides the concrete method defined in the original class. For example:

<!-- a stateful bean deployed as a prototype (non-singleton) -->
<bean id="command" class="fiona.apple.AsyncCommand" scope="prototype">
    <!-- inject dependencies here as required -->
</bean>

<!-- commandProcessor uses statefulCommandHelper -->
<bean id="commandManager" class="fiona.apple.CommandManager">
    <lookup-method name="createCommand" bean="command"/>
</bean>

The bean identified as commandManager calls its own method createCommand() whenever it needs a new instance of the command bean. You must be careful to deploy the command bean as a prototype, if that is actually what is needed. If it is deployed as a singleton, the same instance of the command bean is returned each time.

Spring框架参考文档 Spring Framework Reference Documentation

The interested reader may also find the ServiceLocatorFactoryBean (in the org.springframework.beans.factory.config package) to be of use. The approach used in ServiceLocatorFactoryBean is similar to that of another utility class, ObjectFactoryCreatingFactoryBean, but it allows you to specify your own lookup interface as opposed to a Spring-specific lookup interface. Consult the javadocs of these classes for additional information.

Arbitrary method replacement

A less useful form of method injection than lookup method injection is the ability to replace arbitrary methods in a managed bean with another method implementation. Users may safely skip the rest of this section until the functionality is actually needed.

With XML-based configuration metadata, you can use the replaced-method element to replace an existing method implementation with another, for a deployed bean. Consider the following class, with a method computeValue, which we want to override:

public class MyValueCalculator {

    public String computeValue(String input) {
        // some real code...
    }

    // some other methods...

}

A class implementing the org.springframework.beans.factory.support.MethodReplacer interface provides the new method definition.

/**
 * meant to be used to override the existing computeValue(String)
 * implementation in MyValueCalculator
 */
public class ReplacementComputeValue implements MethodReplacer {

    public Object reimplement(Object o, Method m, Object[] args) throws Throwable {
        // get the input value, work with it, and return a computed result
        String input = (String) args[0];
        ...
        return ...;
    }
}

The bean definition to deploy the original class and specify the method override would look like this:

<bean id="myValueCalculator" class="x.y.z.MyValueCalculator">
    <!-- arbitrary method replacement -->
    <replaced-method name="computeValue" replacer="replacementComputeValue">
        <arg-type>String</arg-type>
    </replaced-method>
</bean>

<bean id="replacementComputeValue" class="a.b.c.ReplacementComputeValue"/>

You can use one or more contained <arg-type/> elements within the <replaced-method/> element to indicate the method signature of the method being overridden. The signature for the arguments is necessary only if the method is overloaded and multiple variants exist within the class. For convenience, the type string for an argument may be a substring of the fully qualified type name. For example, the following all match java.lang.String:

java.lang.String
String
Str

Because the number of arguments is often enough to distinguish between each possible choice, this shortcut can save a lot of typing, by allowing you to type only the shortest string that will match an argument type.

7.5 Bean scopes

When you create a bean definition, you create a recipe for creating actual instances of the class defined by that bean definition. The idea that a bean definition is a recipe is important, because it means that, as with a class, you can create many object instances from a single recipe.

You can control not only the various dependencies and configuration values that are to be plugged into an object that is created from a particular bean definition, but also the scope of the objects created from a particular bean definition. This approach is powerful and flexible in that you can choose the scope of the objects you create through configuration instead of having to bake in the scope of an object at the Java class level. Beans can be defined to be deployed in one of a number of scopes: out of the box, the Spring Framework supports seven scopes, five of which are available only if you use a web-aware ApplicationContext.

The following scopes are supported out of the box. You can also create a custom scope.

Table 7.3. Bean scopes

Scope Description

singleton

(Default) Scopes a single bean definition to a single object instance per Spring IoC container.

prototype

Scopes a single bean definition to any number of object instances.

request

Scopes a single bean definition to the lifecycle of a single HTTP request; that is, each HTTP request has its own instance of a bean created off the back of a single bean definition. Only valid in the context of a web-aware Spring ApplicationContext.

session

Scopes a single bean definition to the lifecycle of an HTTP Session. Only valid in the context of a web-aware Spring ApplicationContext.

globalSession

Scopes a single bean definition to the lifecycle of a global HTTP Session. Typically only valid when used in a Portlet context. Only valid in the context of a web-aware Spring ApplicationContext.

application

Scopes a single bean definition to the lifecycle of a ServletContext. Only valid in the context of a web-aware Spring ApplicationContext.

websocket

Scopes a single bean definition to the lifecycle of a WebSocket. Only valid in the context of a web-aware Spring ApplicationContext.

Spring框架参考文档 Spring Framework Reference Documentation

As of Spring 3.0, a thread scope is available, but is not registered by default. For more information, see the documentation for SimpleThreadScope. For instructions on how to register this or any other custom scope, see the section called “Using a custom scope”.

7.5.1 The singleton scope

Only one shared instance of a singleton bean is managed, and all requests for beans with an id or ids matching that bean definition result in that one specific bean instance being returned by the Spring container.

To put it another way, when you define a bean definition and it is scoped as a singleton, the Spring IoC container creates exactly one instance of the object defined by that bean definition. This single instance is stored in a cache of such singleton beans, and all subsequent requests and references for that named bean return the cached object.

Spring框架参考文档 Spring Framework Reference Documentation

Spring’s concept of a singleton bean differs from the Singleton pattern as defined in the Gang of Four (GoF) patterns book. The GoF Singleton hard-codes the scope of an object such that one and only one instance of a particular class is created per ClassLoader. The scope of the Spring singleton is best described as per container and per bean. This means that if you define one bean for a particular class in a single Spring container, then the Spring container creates one and only one instance of the class defined by that bean definition. The singleton scope is the default scope in Spring. To define a bean as a singleton in XML, you would write, for example:

<bean id="accountService" class="com.foo.DefaultAccountService"/>

<!-- the following is equivalent, though redundant (singleton scope is the default) -->
<bean id="accountService" class="com.foo.DefaultAccountService" scope="singleton"/>

7.5.2 The prototype scope

The non-singleton, prototype scope of bean deployment results in the creation of a new bean instance every time a request for that specific bean is made. That is, the bean is injected into another bean or you request it through a getBean() method call on the container. As a rule, use the prototype scope for all stateful beans and the singleton scope for stateless beans.

The following diagram illustrates the Spring prototype scope. A data access object (DAO) is not typically configured as a prototype, because a typical DAO does not hold any conversational state; it was just easier for this author to reuse the core of the singleton diagram.

Spring框架参考文档 Spring Framework Reference Documentation

The following example defines a bean as a prototype in XML:

<bean id="accountService" class="com.foo.DefaultAccountService" scope="prototype"/>

In contrast to the other scopes, Spring does not manage the complete lifecycle of a prototype bean: the container instantiates, configures, and otherwise assembles a prototype object, and hands it to the client, with no further record of that prototype instance. Thus, although initialization lifecycle callback methods are called on all objects regardless of scope, in the case of prototypes, configured destruction lifecycle callbacks are not called. The client code must clean up prototype-scoped objects and release expensive resources that the prototype bean(s) are holding. To get the Spring container to release resources held by prototype-scoped beans, try using a custom bean post-processor, which holds a reference to beans that need to be cleaned up.

In some respects, the Spring container’s role in regard to a prototype-scoped bean is a replacement for the Java new operator. All lifecycle management past that point must be handled by the client. (For details on the lifecycle of a bean in the Spring container, see Section 7.6.1, “Lifecycle callbacks”.)

7.5.3 Singleton beans with prototype-bean dependencies

When you use singleton-scoped beans with dependencies on prototype beans, be aware that dependencies are resolved at instantiation time. Thus if you dependency-inject a prototype-scoped bean into a singleton-scoped bean, a new prototype bean is instantiated and then dependency-injected into the singleton bean. The prototype instance is the sole instance that is ever supplied to the singleton-scoped bean.

However, suppose you want the singleton-scoped bean to acquire a new instance of the prototype-scoped bean repeatedly at runtime. You cannot dependency-inject a prototype-scoped bean into your singleton bean, because that injection occurs only once, when the Spring container is instantiating the singleton bean and resolving and injecting its dependencies. If you need a new instance of a prototype bean at runtime more than once, see Section 7.4.6, “Method injection”

7.5.4 Request, session, global session, application, and WebSocket scopes

The request, session, globalSession, application, and websocket scopes are only available if you use a web-aware Spring ApplicationContextimplementation (such as XmlWebApplicationContext). If you use these scopes with regular Spring IoC containers such as the ClassPathXmlApplicationContext, an IllegalStateException will be thrown complaining about an unknown bean scope.

Initial web configuration

To support the scoping of beans at the request, session, globalSession, application, and websocket levels (web-scoped beans), some minor initial configuration is required before you define your beans. (This initial setup is not required for the standard scopes, singleton and prototype.)

How you accomplish this initial setup depends on your particular Servlet environment.

If you access scoped beans within Spring Web MVC, in effect, within a request that is processed by the Spring DispatcherServlet or DispatcherPortlet, then no special setup is necessary: DispatcherServlet and DispatcherPortlet already expose all relevant state.

If you use a Servlet 2.5 web container, with requests processed outside of Spring’s DispatcherServlet (for example, when using JSF or Struts), you need to register the org.springframework.web.context.request.RequestContextListener ServletRequestListener. For Servlet 3.0+, this can be done programmatically via the WebApplicationInitializer interface. Alternatively, or for older containers, add the following declaration to your web application’s web.xml file:

<web-app>
    ...
    <listener>
        <listener-class>
            org.springframework.web.context.request.RequestContextListener
        </listener-class>
    </listener>
    ...
</web-app>

Alternatively, if there are issues with your listener setup, consider using Spring’s RequestContextFilter. The filter mapping depends on the surrounding web application configuration, so you have to change it as appropriate.

<web-app>
    ...
    <filter>
        <filter-name>requestContextFilter</filter-name>
        <filter-class>org.springframework.web.filter.RequestContextFilter</filter-class>
    </filter>
    <filter-mapping>
        <filter-name>requestContextFilter</filter-name>
        <url-pattern>/*</url-pattern>
    </filter-mapping>
    ...
</web-app>

DispatcherServlet, RequestContextListener, and RequestContextFilter all do exactly the same thing, namely bind the HTTP request object to the Threadthat is servicing that request. This makes beans that are request- and session-scoped available further down the call chain.

Request scope

Consider the following XML configuration for a bean definition:

<bean id="loginAction" class="com.foo.LoginAction" scope="request"/>

The Spring container creates a new instance of the LoginAction bean by using the loginAction bean definition for each and every HTTP request. That is, theloginAction bean is scoped at the HTTP request level. You can change the internal state of the instance that is created as much as you want, because other instances created from the same loginAction bean definition will not see these changes in state; they are particular to an individual request. When the request completes processing, the bean that is scoped to the request is discarded.

When using annotation-driven components or Java Config, the @RequestScope annotation can be used to assign a component to the request scope.

@RequestScope
@Component
public class LoginAction {
    // ...
}

Session scope

Consider the following XML configuration for a bean definition:

<bean id="userPreferences" class="com.foo.UserPreferences" scope="session"/>

The Spring container creates a new instance of the UserPreferences bean by using the userPreferences bean definition for the lifetime of a single HTTP Session. In other words, the userPreferences bean is effectively scoped at the HTTP Session level. As with request-scoped beans, you can change the internal state of the instance that is created as much as you want, knowing that other HTTP Session instances that are also using instances created from the same userPreferencesbean definition do not see these changes in state, because they are particular to an individual HTTP Session. When the HTTP Session is eventually discarded, the bean that is scoped to that particular HTTP Session is also discarded.

When using annotation-driven components or Java Config, the @SessionScope annotation can be used to assign a component to the session scope.

@SessionScope
@Component
public class UserPreferences {
    // ...
}

Global session scope

Consider the following bean definition:

<bean id="userPreferences" class="com.foo.UserPreferences" scope="globalSession"/>

The globalSession scope is similar to the standard HTTP Session scope (described above), and applies only in the context of portlet-based web applications. The portlet specification defines the notion of a global Session that is shared among all portlets that make up a single portlet web application. Beans defined at theglobalSession scope are scoped (or bound) to the lifetime of the global portlet Session.

If you write a standard Servlet-based web application and you define one or more beans as having globalSession scope, the standard HTTP Session scope is used, and no error is raised.

Application scope

Consider the following XML configuration for a bean definition:

<bean id="appPreferences" class="com.foo.AppPreferences" scope="application"/>

The Spring container creates a new instance of the AppPreferences bean by using the appPreferences bean definition once for the entire web application. That is, the appPreferences bean is scoped at the ServletContext level, stored as a regular ServletContext attribute. This is somewhat similar to a Spring singleton bean but differs in two important ways: It is a singleton per ServletContext, not per Spring 'ApplicationContext' (for which there may be several in any given web application), and it is actually exposed and therefore visible as a ServletContext attribute.

When using annotation-driven components or Java Config, the @ApplicationScope annotation can be used to assign a component to the application scope.

@ApplicationScope
@Component
public class AppPreferences {
    // ...
}

Scoped beans as dependencies

The Spring IoC container manages not only the instantiation of your objects (beans), but also the wiring up of collaborators (or dependencies). If you want to inject (for example) an HTTP request scoped bean into another bean of a longer-lived scope, you may choose to inject an AOP proxy in place of the scoped bean. That is, you need to inject a proxy object that exposes the same public interface as the scoped object but that can also retrieve the real target object from the relevant scope (such as an HTTP request) and delegate method calls onto the real object.

Spring框架参考文档 Spring Framework Reference Documentation

You may also use <aop:scoped-proxy/> between beans that are scoped as singleton, with the reference then going through an intermediate proxy that is serializable and therefore able to re-obtain the target singleton bean on deserialization.

When declaring <aop:scoped-proxy/> against a bean of scope prototype, every method call on the shared proxy will lead to the creation of a new target instance which the call is then being forwarded to.

Also, scoped proxies are not the only way to access beans from shorter scopes in a lifecycle-safe fashion. You may also simply declare your injection point (i.e. the constructor/setter argument or autowired field) as ObjectFactory<MyTargetBean>, allowing for a getObject() call to retrieve the current instance on demand every time it is needed - without holding on to the instance or storing it separately.

The JSR-330 variant of this is called Provider, used with a Provider<MyTargetBean> declaration and a corresponding get() call for every retrieval attempt. See here for more details on JSR-330 overall.

The configuration in the following example is only one line, but it is important to understand the "why" as well as the "how" behind it.

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:aop="http://www.springframework.org/schema/aop"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/aop
        http://www.springframework.org/schema/aop/spring-aop.xsd">

    <!-- an HTTP Session-scoped bean exposed as a proxy -->
    <bean id="userPreferences" class="com.foo.UserPreferences" scope="session">
        <!-- instructs the container to proxy the surrounding bean -->
        <aop:scoped-proxy/>
    </bean>

    <!-- a singleton-scoped bean injected with a proxy to the above bean -->
    <bean id="userService" class="com.foo.SimpleUserService">
        <!-- a reference to the proxied userPreferences bean -->
        <property name="userPreferences" ref="userPreferences"/>
    </bean>
</beans>

To create such a proxy, you insert a child <aop:scoped-proxy/> element into a scoped bean definition (see the section called “Choosing the type of proxy to create”and Chapter 41, XML Schema-based configuration). Why do definitions of beans scoped at the request, session, globalSession and custom-scope levels require the <aop:scoped-proxy/> element? Let’s examine the following singleton bean definition and contrast it with what you need to define for the aforementioned scopes (note that the following userPreferences bean definition as it stands is incomplete).

<bean id="userPreferences" class="com.foo.UserPreferences" scope="session"/>

<bean id="userManager" class="com.foo.UserManager">
    <property name="userPreferences" ref="userPreferences"/>
</bean>

In the preceding example, the singleton bean userManager is injected with a reference to the HTTP Session-scoped bean userPreferences. The salient point here is that the userManager bean is a singleton: it will be instantiated exactly once per container, and its dependencies (in this case only one, the userPreferences bean) are also injected only once. This means that the userManager bean will only operate on the exact same userPreferences object, that is, the one that it was originally injected with.

This is not the behavior you want when injecting a shorter-lived scoped bean into a longer-lived scoped bean, for example injecting an HTTP Session-scoped collaborating bean as a dependency into singleton bean. Rather, you need a single userManager object, and for the lifetime of an HTTP Session, you need auserPreferences object that is specific to said HTTP Session. Thus the container creates an object that exposes the exact same public interface as theUserPreferences class (ideally an object that is a UserPreferences instance) which can fetch the real UserPreferences object from the scoping mechanism (HTTP request, Session, etc.). The container injects this proxy object into the userManager bean, which is unaware that this UserPreferences reference is a proxy. In this example, when a UserManager instance invokes a method on the dependency-injected UserPreferences object, it actually is invoking a method on the proxy. The proxy then fetches the real UserPreferences object from (in this case) the HTTP Session, and delegates the method invocation onto the retrieved realUserPreferences object.

Thus you need the following, correct and complete, configuration when injecting request-, session-, and globalSession-scoped beans into collaborating objects:

<bean id="userPreferences" class="com.foo.UserPreferences" scope="session">
    <aop:scoped-proxy/>
</bean>

<bean id="userManager" class="com.foo.UserManager">
    <property name="userPreferences" ref="userPreferences"/>
</bean>
Choosing the type of proxy to create

By default, when the Spring container creates a proxy for a bean that is marked up with the <aop:scoped-proxy/> element, a CGLIB-based class proxy is created.

Spring框架参考文档 Spring Framework Reference Documentation

CGLIB proxies only intercept public method calls! Do not call non-public methods on such a proxy; they will not be delegated to the actual scoped target object.

Alternatively, you can configure the Spring container to create standard JDK interface-based proxies for such scoped beans, by specifying false for the value of theproxy-target-class attribute of the <aop:scoped-proxy/> element. Using JDK interface-based proxies means that you do not need additional libraries in your application classpath to effect such proxying. However, it also means that the class of the scoped bean must implement at least one interface, and that all collaborators into which the scoped bean is injected must reference the bean through one of its interfaces.

<!-- DefaultUserPreferences implements the UserPreferences interface -->
<bean id="userPreferences" class="com.foo.DefaultUserPreferences" scope="session">
    <aop:scoped-proxy proxy-target-class="false"/>
</bean>

<bean id="userManager" class="com.foo.UserManager">
    <property name="userPreferences" ref="userPreferences"/>
</bean>

For more detailed information about choosing class-based or interface-based proxying, see Section 11.6, “Proxying mechanisms”.

7.5.5 Custom scopes

The bean scoping mechanism is extensible; You can define your own scopes, or even redefine existing scopes, although the latter is considered bad practice and youcannot override the built-in singleton and prototype scopes.

Creating a custom scope

To integrate your custom scope(s) into the Spring container, you need to implement the org.springframework.beans.factory.config.Scope interface, which is described in this section. For an idea of how to implement your own scopes, see the Scope implementations that are supplied with the Spring Framework itself and theScope javadocs, which explains the methods you need to implement in more detail.

The Scope interface has four methods to get objects from the scope, remove them from the scope, and allow them to be destroyed.

The following method returns the object from the underlying scope. The session scope implementation, for example, returns the session-scoped bean (and if it does not exist, the method returns a new instance of the bean, after having bound it to the session for future reference).

Object get(String name, ObjectFactory objectFactory)

The following method removes the object from the underlying scope. The session scope implementation for example, removes the session-scoped bean from the underlying session. The object should be returned, but you can return null if the object with the specified name is not found.

Object remove(String name)

The following method registers the callbacks the scope should execute when it is destroyed or when the specified object in the scope is destroyed. Refer to the javadocs or a Spring scope implementation for more information on destruction callbacks.

void registerDestructionCallback(String name, Runnable destructionCallback)

The following method obtains the conversation identifier for the underlying scope. This identifier is different for each scope. For a session scoped implementation, this identifier can be the session identifier.

String getConversationId()

Using a custom scope

After you write and test one or more custom Scope implementations, you need to make the Spring container aware of your new scope(s). The following method is the central method to register a new Scope with the Spring container:

void registerScope(String scopeName, Scope scope);

This method is declared on the ConfigurableBeanFactory interface, which is available on most of the concrete ApplicationContext implementations that ship with Spring via the BeanFactory property.

The first argument to the registerScope(..) method is the unique name associated with a scope; examples of such names in the Spring container itself aresingleton and prototype. The second argument to the registerScope(..) method is an actual instance of the custom Scope implementation that you wish to register and use.

Suppose that you write your custom Scope implementation, and then register it as below.

Spring框架参考文档 Spring Framework Reference Documentation

The example below uses SimpleThreadScope which is included with Spring, but not registered by default. The instructions would be the same for your own custom Scope implementations.

 
Scope threadScope = new SimpleThreadScope();
beanFactory.registerScope("thread", threadScope);

You then create bean definitions that adhere to the scoping rules of your custom Scope:

<bean id="..." class="..." scope="thread">

With a custom Scope implementation, you are not limited to programmatic registration of the scope. You can also do the Scope registration declaratively, using theCustomScopeConfigurer class:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:aop="http://www.springframework.org/schema/aop"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/aop
        http://www.springframework.org/schema/aop/spring-aop.xsd">

    <bean class="org.springframework.beans.factory.config.CustomScopeConfigurer">
        <property name="scopes">
            <map>
                <entry key="thread">
                    <bean class="org.springframework.context.support.SimpleThreadScope"/>
                </entry>
            </map>
        </property>
    </bean>

    <bean id="bar" class="x.y.Bar" scope="thread">
        <property name="name" value="Rick"/>
        <aop:scoped-proxy/>
    </bean>

    <bean id="foo" class="x.y.Foo">
        <property name="bar" ref="bar"/>
    </bean>

</beans>
Spring框架参考文档 Spring Framework Reference Documentation

When you place <aop:scoped-proxy/> in a FactoryBean implementation, it is the factory bean itself that is scoped, not the object returned fromgetObject().

7.6 Customizing the nature of a bean

7.6.1 Lifecycle callbacks

To interact with the container’s management of the bean lifecycle, you can implement the Spring InitializingBean and DisposableBean interfaces. The container calls afterPropertiesSet() for the former and destroy() for the latter to allow the bean to perform certain actions upon initialization and destruction of your beans.

Spring框架参考文档 Spring Framework Reference Documentation

The JSR-250 @PostConstruct and @PreDestroy annotations are generally considered best practice for receiving lifecycle callbacks in a modern Spring application. Using these annotations means that your beans are not coupled to Spring specific interfaces. For details see Section 7.9.8, “@PostConstruct and @PreDestroy”.

If you don’t want to use the JSR-250 annotations but you are still looking to remove coupling consider the use of init-method and destroy-method object definition metadata.

Internally, the Spring Framework uses BeanPostProcessor implementations to process any callback interfaces it can find and call the appropriate methods. If you need custom features or other lifecycle behavior Spring does not offer out-of-the-box, you can implement a BeanPostProcessor yourself. For more information, seeSection 7.8, “Container Extension Points”.

In addition to the initialization and destruction callbacks, Spring-managed objects may also implement the Lifecycle interface so that those objects can participate in the startup and shutdown process as driven by the container’s own lifecycle.

The lifecycle callback interfaces are described in this section.

Initialization callbacks

The org.springframework.beans.factory.InitializingBean interface allows a bean to perform initialization work after all necessary properties on the bean have been set by the container. The InitializingBean interface specifies a single method:

void afterPropertiesSet() throws Exception;

It is recommended that you do not use the InitializingBean interface because it unnecessarily couples the code to Spring. Alternatively, use the @PostConstructannotation or specify a POJO initialization method. In the case of XML-based configuration metadata, you use the init-method attribute to specify the name of the method that has a void no-argument signature. With Java config, you use the initMethod attribute of @Bean, see the section called “Receiving lifecycle callbacks”. For example, the following:

<bean id="exampleInitBean" class="examples.ExampleBean" init-method="init"/>
public class ExampleBean {

    public void init() {
        // do some initialization work
    }

}

…​is exactly the same as…​

<bean id="exampleInitBean" class="examples.AnotherExampleBean"/>
public class AnotherExampleBean implements InitializingBean {

    public void afterPropertiesSet() {
        // do some initialization work
    }

}

but does not couple the code to Spring.

Destruction callbacks

Implementing the org.springframework.beans.factory.DisposableBean interface allows a bean to get a callback when the container containing it is destroyed. The DisposableBean interface specifies a single method:

void destroy() throws Exception;

It is recommended that you do not use the DisposableBean callback interface because it unnecessarily couples the code to Spring. Alternatively, use the@PreDestroy annotation or specify a generic method that is supported by bean definitions. With XML-based configuration metadata, you use the destroy-methodattribute on the <bean/>. With Java config, you use the destroyMethod attribute of @Bean, see the section called “Receiving lifecycle callbacks”. For example, the following definition:

<bean id="exampleInitBean" class="examples.ExampleBean" destroy-method="cleanup"/>
public class ExampleBean {

    public void cleanup() {
        // do some destruction work (like releasing pooled connections)
    }

}

is exactly the same as:

<bean id="exampleInitBean" class="examples.AnotherExampleBean"/>
public class AnotherExampleBean implements DisposableBean {

    public void destroy() {
        // do some destruction work (like releasing pooled connections)
    }

}

but does not couple the code to Spring.

Spring框架参考文档 Spring Framework Reference Documentation

The destroy-method attribute of a <bean> element can be assigned a special (inferred) value which instructs Spring to automatically detect a publicclose or shutdown method on the specific bean class (any class that implements java.lang.AutoCloseable or java.io.Closeable would therefore match). This special (inferred) value can also be set on the default-destroy-method attribute of a <beans> element to apply this behavior to an entire set of beans (see the section called “Default initialization and destroy methods”). Note that this is the default behavior with Java config.

Default initialization and destroy methods

When you write initialization and destroy method callbacks that do not use the Spring-specific InitializingBean and DisposableBean callback interfaces, you typically write methods with names such as init(), initialize(), dispose(), and so on. Ideally, the names of such lifecycle callback methods are standardized across a project so that all developers use the same method names and ensure consistency.

You can configure the Spring container to look for named initialization and destroy callback method names on every bean. This means that you, as an application developer, can write your application classes and use an initialization callback called init(), without having to configure an init-method="init" attribute with each bean definition. The Spring IoC container calls that method when the bean is created (and in accordance with the standard lifecycle callback contract described previously). This feature also enforces a consistent naming convention for initialization and destroy method callbacks.

Suppose that your initialization callback methods are named init() and destroy callback methods are named destroy(). Your class will resemble the class in the following example.

public class DefaultBlogService implements BlogService {

    private BlogDao blogDao;

    public void setBlogDao(BlogDao blogDao) {
        this.blogDao = blogDao;
    }

    // this is (unsurprisingly) the initialization callback method
    public void init() {
        if (this.blogDao == null) {
            throw new IllegalStateException("The [blogDao] property must be set.");
        }
    }

}
<beans default-init-method="init">

    <bean id="blogService" class="com.foo.DefaultBlogService">
        <property name="blogDao" ref="blogDao" />
    </bean>

</beans>

The presence of the default-init-method attribute on the top-level <beans/> element attribute causes the Spring IoC container to recognize a method called initon beans as the initialization method callback. When a bean is created and assembled, if the bean class has such a method, it is invoked at the appropriate time.

You configure destroy method callbacks similarly (in XML, that is) by using the default-destroy-method attribute on the top-level <beans/> element.

Where existing bean classes already have callback methods that are named at variance with the convention, you can override the default by specifying (in XML, that is) the method name using the init-method and destroy-method attributes of the <bean/> itself.

The Spring container guarantees that a configured initialization callback is called immediately after a bean is supplied with all dependencies. Thus the initialization callback is called on the raw bean reference, which means that AOP interceptors and so forth are not yet applied to the bean. A target bean is fully created first, then an AOP proxy (for example) with its interceptor chain is applied. If the target bean and the proxy are defined separately, your code can even interact with the raw target bean, bypassing the proxy. Hence, it would be inconsistent to apply the interceptors to the init method, because doing so would couple the lifecycle of the target bean with its proxy/interceptors and leave strange semantics when your code interacts directly to the raw target bean.

Combining lifecycle mechanisms

As of Spring 2.5, you have three options for controlling bean lifecycle behavior: the InitializingBean and DisposableBean callback interfaces; custom init() anddestroy() methods; and the @PostConstruct and @PreDestroy annotations. You can combine these mechanisms to control a given bean.

Spring框架参考文档 Spring Framework Reference Documentation

If multiple lifecycle mechanisms are configured for a bean, and each mechanism is configured with a different method name, then each configured method is executed in the order listed below. However, if the same method name is configured - for example, init() for an initialization method - for more than one of these lifecycle mechanisms, that method is executed once, as explained in the preceding section.

Multiple lifecycle mechanisms configured for the same bean, with different initialization methods, are called as follows:

  • Methods annotated with @PostConstruct
  • afterPropertiesSet() as defined by the InitializingBean callback interface
  • A custom configured init() method

Destroy methods are called in the same order:

  • Methods annotated with @PreDestroy
  • destroy() as defined by the DisposableBean callback interface
  • A custom configured destroy() method

Startup and shutdown callbacks

The Lifecycle interface defines the essential methods for any object that has its own lifecycle requirements (e.g. starts and stops some background process):

public interface Lifecycle {

    void start();

    void stop();

    boolean isRunning();

}

Any Spring-managed object may implement that interface. Then, when the ApplicationContext itself receives start and stop signals, e.g. for a stop/restart scenario at runtime, it will cascade those calls to all Lifecycle implementations defined within that context. It does this by delegating to a LifecycleProcessor:

public interface LifecycleProcessor extends Lifecycle {

    void onRefresh();

    void onClose();

}

Notice that the LifecycleProcessor is itself an extension of the Lifecycle interface. It also adds two other methods for reacting to the context being refreshed and closed.

Spring框架参考文档 Spring Framework Reference Documentation

Note that the regular org.springframework.context.Lifecycle interface is just a plain contract for explicit start/stop notifications and does NOT imply auto-startup at context refresh time. Consider implementing org.springframework.context.SmartLifecycle instead for fine-grained control over auto-startup of a specific bean (including startup phases). Also, please note that stop notifications are not guaranteed to come before destruction: On regular shutdown, all Lifecycle beans will first receive a stop notification before the general destruction callbacks are being propagated; however, on hot refresh during a context’s lifetime or on aborted refresh attempts, only destroy methods will be called.

The order of startup and shutdown invocations can be important. If a "depends-on" relationship exists between any two objects, the dependent side will start after its dependency, and it will stop before its dependency. However, at times the direct dependencies are unknown. You may only know that objects of a certain type should start prior to objects of another type. In those cases, the SmartLifecycle interface defines another option, namely the getPhase() method as defined on its super-interface, Phased.

public interface Phased {

    int getPhase();

}
public interface SmartLifecycle extends Lifecycle, Phased {

    boolean isAutoStartup();

    void stop(Runnable callback);

}

When starting, the objects with the lowest phase start first, and when stopping, the reverse order is followed. Therefore, an object that implements SmartLifecycle and whose getPhase() method returns Integer.MIN_VALUE would be among the first to start and the last to stop. At the other end of the spectrum, a phase value ofInteger.MAX_VALUE would indicate that the object should be started last and stopped first (likely because it depends on other processes to be running). When considering the phase value, it’s also important to know that the default phase for any "normal" Lifecycle object that does not implement SmartLifecycle would be 0. Therefore, any negative phase value would indicate that an object should start before those standard components (and stop after them), and vice versa for any positive phase value.

As you can see the stop method defined by SmartLifecycle accepts a callback. Any implementation must invoke that callback’s run() method after that implementation’s shutdown process is complete. That enables asynchronous shutdown where necessary since the default implementation of the LifecycleProcessorinterface, DefaultLifecycleProcessor, will wait up to its timeout value for the group of objects within each phase to invoke that callback. The default per-phase timeout is 30 seconds. You can override the default lifecycle processor instance by defining a bean named "lifecycleProcessor" within the context. If you only want to modify the timeout, then defining the following would be sufficient:

<bean id="lifecycleProcessor" class="org.springframework.context.support.DefaultLifecycleProcessor">
    <!-- timeout value in milliseconds -->
    <property name="timeoutPerShutdownPhase" value="10000"/>
</bean>

As mentioned, the LifecycleProcessor interface defines callback methods for the refreshing and closing of the context as well. The latter will simply drive the shutdown process as if stop() had been called explicitly, but it will happen when the context is closing. The 'refresh' callback on the other hand enables another feature of SmartLifecycle beans. When the context is refreshed (after all objects have been instantiated and initialized), that callback will be invoked, and at that point the default lifecycle processor will check the boolean value returned by each SmartLifecycle object’s isAutoStartup() method. If "true", then that object will be started at that point rather than waiting for an explicit invocation of the context’s or its own start() method (unlike the context refresh, the context start does not happen automatically for a standard context implementation). The "phase" value as well as any "depends-on" relationships will determine the startup order in the same way as described above.

Shutting down the Spring IoC container gracefully in non-web applications

Spring框架参考文档 Spring Framework Reference Documentation

This section applies only to non-web applications. Spring’s web-based ApplicationContext implementations already have code in place to shut down the Spring IoC container gracefully when the relevant web application is shut down.

If you are using Spring’s IoC container in a non-web application environment; for example, in a rich client desktop environment; you register a shutdown hook with the JVM. Doing so ensures a graceful shutdown and calls the relevant destroy methods on your singleton beans so that all resources are released. Of course, you must still configure and implement these destroy callbacks correctly.

To register a shutdown hook, you call the registerShutdownHook() method that is declared on the ConfigurableApplicationContext interface:

import org.springframework.context.ConfigurableApplicationContext;
import org.springframework.context.support.ClassPathXmlApplicationContext;

public final class Boot {

    public static void main(final String[] args) throws Exception {

        ConfigurableApplicationContext ctx = new ClassPathXmlApplicationContext(
                new String []{"beans.xml"});

        // add a shutdown hook for the above context...
        ctx.registerShutdownHook();

        // app runs here...

        // main method exits, hook is called prior to the app shutting down...

    }
}

7.6.2 ApplicationContextAware and BeanNameAware

When an ApplicationContext creates an object instance that implements the org.springframework.context.ApplicationContextAware interface, the instance is provided with a reference to that ApplicationContext.

public interface ApplicationContextAware {

    void setApplicationContext(ApplicationContext applicationContext) throws BeansException;

}

Thus beans can manipulate programmatically the ApplicationContext that created them, through the ApplicationContext interface, or by casting the reference to a known subclass of this interface, such as ConfigurableApplicationContext, which exposes additional functionality. One use would be the programmatic retrieval of other beans. Sometimes this capability is useful; however, in general you should avoid it, because it couples the code to Spring and does not follow the Inversion of Control style, where collaborators are provided to beans as properties. Other methods of the ApplicationContext provide access to file resources, publishing application events, and accessing a MessageSource. These additional features are described in Section 7.15, “Additional Capabilities of the ApplicationContext”

As of Spring 2.5, autowiring is another alternative to obtain reference to the ApplicationContext. The "traditional" constructor and byType autowiring modes (as described in Section 7.4.5, “Autowiring collaborators”) can provide a dependency of type ApplicationContext for a constructor argument or setter method parameter, respectively. For more flexibility, including the ability to autowire fields and multiple parameter methods, use the new annotation-based autowiring features. If you do, theApplicationContext is autowired into a field, constructor argument, or method parameter that is expecting the ApplicationContext type if the field, constructor, or method in question carries the @Autowired annotation. For more information, see Section 7.9.2, “@Autowired”.

When an ApplicationContext creates a class that implements the org.springframework.beans.factory.BeanNameAware interface, the class is provided with a reference to the name defined in its associated object definition.

public interface BeanNameAware {

    void setBeanName(String name) throws BeansException;

}

The callback is invoked after population of normal bean properties but before an initialization callback such as InitializingBean afterPropertiesSet or a custom init-method.

7.6.3 Other Aware interfaces

Besides ApplicationContextAware and BeanNameAware discussed above, Spring offers a range of Aware interfaces that allow beans to indicate to the container that they require a certain infrastructure dependency. The most important Aware interfaces are summarized below - as a general rule, the name is a good indication of the dependency type:

Table 7.4. Aware interfaces

Name Injected Dependency Explained in…​

ApplicationContextAware

Declaring ApplicationContext

Section 7.6.2, “ApplicationContextAware and BeanNameAware”

ApplicationEventPublisherAware

Event publisher of the enclosing ApplicationContext

Section 7.15, “Additional Capabilities of the ApplicationContext”

BeanClassLoaderAware

Class loader used to load the bean classes.

Section 7.3.2, “Instantiating beans”

BeanFactoryAware

Declaring BeanFactory

Section 7.6.2, “ApplicationContextAware and BeanNameAware”

BeanNameAware

Name of the declaring bean

Section 7.6.2, “ApplicationContextAware and BeanNameAware”

BootstrapContextAware

Resource adapter BootstrapContext the container runs in. Typically available only in JCA aware ApplicationContexts

Chapter 32, JCA CCI

LoadTimeWeaverAware

Defined weaver for processing class definition at load time

Section 11.8.4, “Load-time weaving with AspectJ in the Spring Framework”

MessageSourceAware

Configured strategy for resolving messages (with support for parametrization and internationalization)

Section 7.15, “Additional Capabilities of the ApplicationContext”

NotificationPublisherAware

Spring JMX notification publisher

Section 31.7, “Notifications”

PortletConfigAware

Current PortletConfig the container runs in. Valid only in a web-aware Spring ApplicationContext

Chapter 25, Portlet MVC Framework

PortletContextAware

Current PortletContext the container runs in. Valid only in a web-aware Spring ApplicationContext

Chapter 25, Portlet MVC Framework

ResourceLoaderAware

Configured loader for low-level access to resources

Chapter 8, Resources

ServletConfigAware

Current ServletConfig the container runs in. Valid only in a web-aware Spring ApplicationContext

Chapter 22, Web MVC framework

ServletContextAware

Current ServletContext the container runs in. Valid only in a web-aware Spring ApplicationContext

Chapter 22, Web MVC framework

Note again that usage of these interfaces ties your code to the Spring API and does not follow the Inversion of Control style. As such, they are recommended for infrastructure beans that require programmatic access to the container.

7.7 Bean definition inheritance

A bean definition can contain a lot of configuration information, including constructor arguments, property values, and container-specific information such as initialization method, static factory method name, and so on. A child bean definition inherits configuration data from a parent definition. The child definition can override some values, or add others, as needed. Using parent and child bean definitions can save a lot of typing. Effectively, this is a form of templating.

If you work with an ApplicationContext interface programmatically, child bean definitions are represented by the ChildBeanDefinition class. Most users do not work with them on this level, instead configuring bean definitions declaratively in something like the ClassPathXmlApplicationContext. When you use XML-based configuration metadata, you indicate a child bean definition by using the parent attribute, specifying the parent bean as the value of this attribute.

<bean id="inheritedTestBean" abstract="true"
        class="org.springframework.beans.TestBean">
    <property name="name" value="parent"/>
    <property name="age" value="1"/>
</bean>

<bean id="inheritsWithDifferentClass"
        class="org.springframework.beans.DerivedTestBean"
        parent="inheritedTestBean" init-method="initialize">
    <property name="name" value="override"/>
    <!-- the age property value of 1 will be inherited from parent -->
</bean>

A child bean definition uses the bean class from the parent definition if none is specified, but can also override it. In the latter case, the child bean class must be compatible with the parent, that is, it must accept the parent’s property values.

A child bean definition inherits scope, constructor argument values, property values, and method overrides from the parent, with the option to add new values. Any scope, initialization method, destroy method, and/or static factory method settings that you specify will override the corresponding parent settings.

The remaining settings are always taken from the child definition: depends on, autowire mode, dependency check, singleton, lazy init.

The preceding example explicitly marks the parent bean definition as abstract by using the abstract attribute. If the parent definition does not specify a class, explicitly marking the parent bean definition as abstract is required, as follows:

<bean id="inheritedTestBeanWithoutClass" abstract="true">
    <property name="name" value="parent"/>
    <property name="age" value="1"/>
</bean>

<bean id="inheritsWithClass" class="org.springframework.beans.DerivedTestBean"
        parent="inheritedTestBeanWithoutClass" init-method="initialize">
    <property name="name" value="override"/>
    <!-- age will inherit the value of 1 from the parent bean definition-->
</bean>

The parent bean cannot be instantiated on its own because it is incomplete, and it is also explicitly marked as abstract. When a definition is abstract like this, it is usable only as a pure template bean definition that serves as a parent definition for child definitions. Trying to use such an abstract parent bean on its own, by referring to it as a ref property of another bean or doing an explicit getBean() call with the parent bean id, returns an error. Similarly, the container’s internalpreInstantiateSingletons() method ignores bean definitions that are defined as abstract.

Spring框架参考文档 Spring Framework Reference Documentation

ApplicationContext pre-instantiates all singletons by default. Therefore, it is important (at least for singleton beans) that if you have a (parent) bean definition which you intend to use only as a template, and this definition specifies a class, you must make sure to set the abstract attribute to true, otherwise the application context will actually (attempt to) pre-instantiate the abstract bean.

7.8 Container Extension Points

Typically, an application developer does not need to subclass ApplicationContext implementation classes. Instead, the Spring IoC container can be extended by plugging in implementations of special integration interfaces. The next few sections describe these integration interfaces.

7.8.1 Customizing beans using a BeanPostProcessor

The BeanPostProcessor interface defines callback methods that you can implement to provide your own (or override the container’s default) instantiation logic, dependency-resolution logic, and so forth. If you want to implement some custom logic after the Spring container finishes instantiating, configuring, and initializing a bean, you can plug in one or more BeanPostProcessor implementations.

You can configure multiple BeanPostProcessor instances, and you can control the order in which these BeanPostProcessors execute by setting the order property. You can set this property only if the BeanPostProcessor implements the Ordered interface; if you write your own BeanPostProcessor you should consider implementing the Ordered interface too. For further details, consult the javadocs of the BeanPostProcessor and Ordered interfaces. See also the note below onprogrammatic registration of BeanPostProcessors.

Spring框架参考文档 Spring Framework Reference Documentation

BeanPostProcessors operate on bean (or object) instances; that is to say, the Spring IoC container instantiates a bean instance and thenBeanPostProcessors do their work.

BeanPostProcessors are scoped per-container. This is only relevant if you are using container hierarchies. If you define a BeanPostProcessor in one container, it will only post-process the beans in that container. In other words, beans that are defined in one container are not post-processed by aBeanPostProcessor defined in another container, even if both containers are part of the same hierarchy.

To change the actual bean definition (i.e., the blueprint that defines the bean), you instead need to use a BeanFactoryPostProcessor as described inSection 7.8.2, “Customizing configuration metadata with a BeanFactoryPostProcessor”.

The org.springframework.beans.factory.config.BeanPostProcessor interface consists of exactly two callback methods. When such a class is registered as a post-processor with the container, for each bean instance that is created by the container, the post-processor gets a callback from the container both before container initialization methods (such as InitializingBean’s afterPropertiesSet() and any declared init method) are called as well as after any bean initialization callbacks. The post-processor can take any action with the bean instance, including ignoring the callback completely. A bean post-processor typically checks for callback interfaces or may wrap a bean with a proxy. Some Spring AOP infrastructure classes are implemented as bean post-processors in order to provide proxy-wrapping logic.

An ApplicationContext automatically detects any beans that are defined in the configuration metadata which implement the BeanPostProcessor interface. TheApplicationContext registers these beans as post-processors so that they can be called later upon bean creation. Bean post-processors can be deployed in the container just like any other beans.

Note that when declaring a BeanPostProcessor using an @Bean factory method on a configuration class, the return type of the factory method should be the implementation class itself or at least the org.springframework.beans.factory.config.BeanPostProcessor interface, clearly indicating the post-processor nature of that bean. Otherwise, the ApplicationContext won’t be able to autodetect it by type before fully creating it. Since a BeanPostProcessor needs to be instantiated early in order to apply to the initialization of other beans in the context, this early type detection is critical.

Spring框架参考文档 Spring Framework Reference Documentation

While the recommended approach for BeanPostProcessor registration is through ApplicationContext auto-detection (as described above), it is also possible to register them programmatically against a ConfigurableBeanFactory using the addBeanPostProcessor method. This can be useful when needing to evaluate conditional logic before registration, or even for copying bean post processors across contexts in a hierarchy. Note however thatBeanPostProcessors added programmatically do not respect the Ordered interface. Here it is the order of registration that dictates the order of execution. Note also that BeanPostProcessors registered programmatically are always processed before those registered through auto-detection, regardless of any explicit ordering.
Spring框架参考文档 Spring Framework Reference Documentation

Classes that implement the BeanPostProcessor interface are special and are treated differently by the container. All BeanPostProcessors and beans that they reference directly are instantiated on startup, as part of the special startup phase of the ApplicationContext. Next, all BeanPostProcessors are registered in a sorted fashion and applied to all further beans in the container. Because AOP auto-proxying is implemented as a BeanPostProcessoritself, neither BeanPostProcessors nor the beans they reference directly are eligible for auto-proxying, and thus do not have aspects woven into them.

For any such bean, you should see an informational log message: "Bean foo is not eligible for getting processed by all BeanPostProcessor interfaces (for example: not eligible for auto-proxying)".

Note that if you have beans wired into your BeanPostProcessor using autowiring or @Resource (which may fall back to autowiring), Spring might access unexpected beans when searching for type-matching dependency candidates, and therefore make them ineligible for auto-proxying or other kinds of bean post-processing. For example, if you have a dependency annotated with @Resource where the field/setter name does not directly correspond to the declared name of a bean and no name attribute is used, then Spring will access other beans for matching them by type.

The following examples show how to write, register, and use BeanPostProcessors in an ApplicationContext.

Example: Hello World, BeanPostProcessor-style

This first example illustrates basic usage. The example shows a custom BeanPostProcessor implementation that invokes the toString() method of each bean as it is created by the container and prints the resulting string to the system console.

Find below the custom BeanPostProcessor implementation class definition:

package scripting;

import org.springframework.beans.factory.config.BeanPostProcessor;
import org.springframework.beans.BeansException;

public class InstantiationTracingBeanPostProcessor implements BeanPostProcessor {

    // simply return the instantiated bean as-is
    public Object postProcessBeforeInitialization(Object bean,
            String beanName) throws BeansException {
        return bean; // we could potentially return any object reference here...
    }

    public Object postProcessAfterInitialization(Object bean,
            String beanName) throws BeansException {
        System.out.println("Bean '" + beanName + "' created : " + bean.toString());
        return bean;
    }

}
<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:lang="http://www.springframework.org/schema/lang"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/lang
        http://www.springframework.org/schema/lang/spring-lang.xsd">

    <lang:groovy id="messenger"
            script-source="classpath:org/springframework/scripting/groovy/Messenger.groovy">
        <lang:property name="message" value="Fiona Apple Is Just So Dreamy."/>
    </lang:groovy>

    <!--
    when the above bean (messenger) is instantiated, this custom
    BeanPostProcessor implementation will output the fact to the system console
    -->
    <bean class="scripting.InstantiationTracingBeanPostProcessor"/>

</beans>

Notice how the InstantiationTracingBeanPostProcessor is simply defined. It does not even have a name, and because it is a bean it can be dependency-injected just like any other bean. (The preceding configuration also defines a bean that is backed by a Groovy script. The Spring dynamic language support is detailed in the chapter entitled Chapter 35, Dynamic language support.)

The following simple Java application executes the preceding code and configuration:

import org.springframework.context.ApplicationContext;
import org.springframework.context.support.ClassPathXmlApplicationContext;
import org.springframework.scripting.Messenger;

public final class Boot {

    public static void main(final String[] args) throws Exception {
        ApplicationContext ctx = new ClassPathXmlApplicationContext("scripting/beans.xml");
        Messenger messenger = (Messenger) ctx.getBean("messenger");
        System.out.println(messenger);
    }

}

The output of the preceding application resembles the following:

Bean 'messenger' created : org.springframework.scripting.groovy.GroovyMessenger@272961
org.springframework.scripting.groovy.GroovyMessenger@272961

Example: The RequiredAnnotationBeanPostProcessor

Using callback interfaces or annotations in conjunction with a custom BeanPostProcessor implementation is a common means of extending the Spring IoC container. An example is Spring’s RequiredAnnotationBeanPostProcessor - a BeanPostProcessor implementation that ships with the Spring distribution which ensures that JavaBean properties on beans that are marked with an (arbitrary) annotation are actually (configured to be) dependency-injected with a value.

7.8.2 Customizing configuration metadata with a BeanFactoryPostProcessor

The next extension point that we will look at is the org.springframework.beans.factory.config.BeanFactoryPostProcessor. The semantics of this interface are similar to those of the BeanPostProcessor, with one major difference: BeanFactoryPostProcessor operates on the bean configuration metadata; that is, the Spring IoC container allows a BeanFactoryPostProcessor to read the configuration metadata and potentially change it before the container instantiates any beans other thanBeanFactoryPostProcessors.

You can configure multiple BeanFactoryPostProcessors, and you can control the order in which these BeanFactoryPostProcessors execute by setting the orderproperty. However, you can only set this property if the BeanFactoryPostProcessor implements the Ordered interface. If you write your ownBeanFactoryPostProcessor, you should consider implementing the Ordered interface too. Consult the javadocs of the BeanFactoryPostProcessor and Orderedinterfaces for more details.

Spring框架参考文档 Spring Framework Reference Documentation

If you want to change the actual bean instances (i.e., the objects that are created from the configuration metadata), then you instead need to use aBeanPostProcessor (described above in Section 7.8.1, “Customizing beans using a BeanPostProcessor”). While it is technically possible to work with bean instances within a BeanFactoryPostProcessor (e.g., using BeanFactory.getBean()), doing so causes premature bean instantiation, violating the standard container lifecycle. This may cause negative side effects such as bypassing bean post processing.

Also, BeanFactoryPostProcessors are scoped per-container. This is only relevant if you are using container hierarchies. If you define aBeanFactoryPostProcessor in one container, it will only be applied to the bean definitions in that container. Bean definitions in one container will not be post-processed by BeanFactoryPostProcessors in another container, even if both containers are part of the same hierarchy.

A bean factory post-processor is executed automatically when it is declared inside an ApplicationContext, in order to apply changes to the configuration metadata that define the container. Spring includes a number of predefined bean factory post-processors, such as PropertyOverrideConfigurer andPropertyPlaceholderConfigurer. A custom BeanFactoryPostProcessor can also be used, for example, to register custom property editors.

An ApplicationContext automatically detects any beans that are deployed into it that implement the BeanFactoryPostProcessor interface. It uses these beans as bean factory post-processors, at the appropriate time. You can deploy these post-processor beans as you would any other bean.

Spring框架参考文档 Spring Framework Reference Documentation

As with BeanPostProcessors , you typically do not want to configure BeanFactoryPostProcessors for lazy initialization. If no other bean references aBean(Factory)PostProcessor, that post-processor will not get instantiated at all. Thus, marking it for lazy initialization will be ignored, and theBean(Factory)PostProcessor will be instantiated eagerly even if you set the default-lazy-init attribute to true on the declaration of your<beans /> element.

Example: the Class name substitution PropertyPlaceholderConfigurer

You use the PropertyPlaceholderConfigurer to externalize property values from a bean definition in a separate file using the standard Java Properties format. Doing so enables the person deploying an application to customize environment-specific properties such as database URLs and passwords, without the complexity or risk of modifying the main XML definition file or files for the container.

Consider the following XML-based configuration metadata fragment, where a DataSource with placeholder values is defined. The example shows properties configured from an external Properties file. At runtime, a PropertyPlaceholderConfigurer is applied to the metadata that will replace some properties of the DataSource. The values to replace are specified as placeholders of the form ${property-name} which follows the Ant / log4j / JSP EL style.

<bean class="org.springframework.beans.factory.config.PropertyPlaceholderConfigurer">
    <property name="locations" value="classpath:com/foo/jdbc.properties"/>
</bean>

<bean id="dataSource" destroy-method="close"
        class="org.apache.commons.dbcp.BasicDataSource">
    <property name="driverClassName" value="${jdbc.driverClassName}"/>
    <property name="url" value="${jdbc.url}"/>
    <property name="username" value="${jdbc.username}"/>
    <property name="password" value="${jdbc.password}"/>
</bean>

The actual values come from another file in the standard Java Properties format:

jdbc.driverClassName=org.hsqldb.jdbcDriver
jdbc.url=jdbc:hsqldb:hsql://production:9002
jdbc.username=sa
jdbc.password=root

Therefore, the string ${jdbc.username} is replaced at runtime with the value 'sa', and the same applies for other placeholder values that match keys in the properties file. The PropertyPlaceholderConfigurer checks for placeholders in most properties and attributes of a bean definition. Furthermore, the placeholder prefix and suffix can be customized.

With the context namespace introduced in Spring 2.5, it is possible to configure property placeholders with a dedicated configuration element. One or more locations can be provided as a comma-separated list in the location attribute.

<context:property-placeholder location="classpath:com/foo/jdbc.properties"/>

The PropertyPlaceholderConfigurer not only looks for properties in the Properties file you specify. By default it also checks against the Java System properties if it cannot find a property in the specified properties files. You can customize this behavior by setting the systemPropertiesMode property of the configurer with one of the following three supported integer values:

  • never (0): Never check system properties
  • fallback (1): Check system properties if not resolvable in the specified properties files. This is the default.
  • override (2): Check system properties first, before trying the specified properties files. This allows system properties to override any other property source.

Consult the PropertyPlaceholderConfigurer javadocs for more information.

Spring框架参考文档 Spring Framework Reference Documentation

You can use the PropertyPlaceholderConfigurer to substitute class names, which is sometimes useful when you have to pick a particular implementation class at runtime. For example:

<bean class="org.springframework.beans.factory.config.PropertyPlaceholderConfigurer">
    <property name="locations">
        <value>classpath:com/foo/strategy.properties</value>
    </property>
    <property name="properties">
        <value>custom.strategy.class=com.foo.DefaultStrategy</value>
    </property>
</bean>

<bean id="serviceStrategy" class="${custom.strategy.class}"/>

If the class cannot be resolved at runtime to a valid class, resolution of the bean fails when it is about to be created, which is during thepreInstantiateSingletons() phase of an ApplicationContext for a non-lazy-init bean.

Example: the PropertyOverrideConfigurer

The PropertyOverrideConfigurer, another bean factory post-processor, resembles the PropertyPlaceholderConfigurer, but unlike the latter, the original definitions can have default values or no values at all for bean properties. If an overriding Properties file does not have an entry for a certain bean property, the default context definition is used.

Note that the bean definition is not aware of being overridden, so it is not immediately obvious from the XML definition file that the override configurer is being used. In case of multiple PropertyOverrideConfigurer instances that define different values for the same bean property, the last one wins, due to the overriding mechanism.

Properties file configuration lines take this format:

beanName.property=value

For example:

dataSource.driverClassName=com.mysql.jdbc.Driver
dataSource.url=jdbc:mysql:mydb

This example file can be used with a container definition that contains a bean called dataSource, which has driver and url properties.

Compound property names are also supported, as long as every component of the path except the final property being overridden is already non-null (presumably initialized by the constructors). In this example…​

foo.fred.bob.sammy=123
  1. the sammy property of the bob property of the fred property of the foo bean is set to the scalar value 123.
Spring框架参考文档 Spring Framework Reference Documentation

Specified override values are always literal values; they are not translated into bean references. This convention also applies when the original value in the XML bean definition specifies a bean reference.

With the context namespace introduced in Spring 2.5, it is possible to configure property overriding with a dedicated configuration element:

<context:property-override location="classpath:override.properties"/>

7.8.3 Customizing instantiation logic with a FactoryBean

Implement the org.springframework.beans.factory.FactoryBean interface for objects that are themselves factories.

The FactoryBean interface is a point of pluggability into the Spring IoC container’s instantiation logic. If you have complex initialization code that is better expressed in Java as opposed to a (potentially) verbose amount of XML, you can create your own FactoryBean, write the complex initialization inside that class, and then plug your custom FactoryBean into the container.

The FactoryBean interface provides three methods:

  • Object getObject(): returns an instance of the object this factory creates. The instance can possibly be shared, depending on whether this factory returns singletons or prototypes.
  • boolean isSingleton(): returns true if this FactoryBean returns singletons, false otherwise.
  • Class getObjectType(): returns the object type returned by the getObject() method or null if the type is not known in advance.

The FactoryBean concept and interface is used in a number of places within the Spring Framework; more than 50 implementations of the FactoryBean interface ship with Spring itself.

When you need to ask a container for an actual FactoryBean instance itself instead of the bean it produces, preface the bean’s id with the ampersand symbol ( &) when calling the getBean() method of the ApplicationContext. So for a given FactoryBean with an id of myBean, invoking getBean("myBean") on the container returns the product of the FactoryBean; whereas, invoking getBean("&myBean") returns the FactoryBean instance itself.

7.9 Annotation-based container configuration

Are annotations better than XML for configuring Spring?

The introduction of annotation-based configurations raised the question of whether this approach is 'better' than XML. The short answer is it depends. The long answer is that each approach has its pros and cons, and usually it is up to the developer to decide which strategy suits them better. Due to the way they are defined, annotations provide a lot of context in their declaration, leading to shorter and more concise configuration. However, XML excels at wiring up components without touching their source code or recompiling them. Some developers prefer having the wiring close to the source while others argue that annotated classes are no longer POJOs and, furthermore, that the configuration becomes decentralized and harder to control.

No matter the choice, Spring can accommodate both styles and even mix them together. It’s worth pointing out that through its JavaConfig option, Spring allows annotations to be used in a non-invasive way, without touching the target components source code and that in terms of tooling, all configuration styles are supported by the Spring Tool Suite.

An alternative to XML setups is provided by annotation-based configuration which rely on the bytecode metadata for wiring up components instead of angle-bracket declarations. Instead of using XML to describe a bean wiring, the developer moves the configuration into the component class itself by using annotations on the relevant class, method, or field declaration. As mentioned in the section called “Example: The RequiredAnnotationBeanPostProcessor”, using a BeanPostProcessor in conjunction with annotations is a common means of extending the Spring IoC container. For example, Spring 2.0 introduced the possibility of enforcing required properties with the @Required annotation. Spring 2.5 made it possible to follow that same general approach to drive Spring’s dependency injection. Essentially, the@Autowired annotation provides the same capabilities as described in Section 7.4.5, “Autowiring collaborators” but with more fine-grained control and wider applicability. Spring 2.5 also added support for JSR-250 annotations such as @PostConstruct, and @PreDestroy. Spring 3.0 added support for JSR-330 (Dependency Injection for Java) annotations contained in the javax.inject package such as @Inject and @Named. Details about those annotations can be found in the relevant section.

Spring框架参考文档 Spring Framework Reference Documentation

Annotation injection is performed before XML injection, thus the latter configuration will override the former for properties wired through both approaches.

As always, you can register them as individual bean definitions, but they can also be implicitly registered by including the following tag in an XML-based Spring configuration (notice the inclusion of the context namespace):

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:annotation-config/>

</beans>

(The implicitly registered post-processors include AutowiredAnnotationBeanPostProcessor, CommonAnnotationBeanPostProcessor,PersistenceAnnotationBeanPostProcessor, as well as the aforementioned RequiredAnnotationBeanPostProcessor.)

Spring框架参考文档 Spring Framework Reference Documentation

<context:annotation-config/> only looks for annotations on beans in the same application context in which it is defined. This means that, if you put<context:annotation-config/> in a WebApplicationContext for a DispatcherServlet, it only checks for @Autowired beans in your controllers, and not your services. See Section 22.2, “The DispatcherServlet” for more information.

7.9.1 @Required

The @Required annotation applies to bean property setter methods, as in the following example:

public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Required
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // ...

}

This annotation simply indicates that the affected bean property must be populated at configuration time, through an explicit property value in a bean definition or through autowiring. The container throws an exception if the affected bean property has not been populated; this allows for eager and explicit failure, avoidingNullPointerExceptions or the like later on. It is still recommended that you put assertions into the bean class itself, for example, into an init method. Doing so enforces those required references and values even when you use the class outside of a container.

7.9.2 @Autowired

Spring框架参考文档 Spring Framework Reference Documentation

JSR 330’s @Inject annotation can be used in place of Spring’s @Autowired annotation in the examples below. See here for more details.

You can apply the @Autowired annotation to constructors:

public class MovieRecommender {

    private final CustomerPreferenceDao customerPreferenceDao;

    @Autowired
    public MovieRecommender(CustomerPreferenceDao customerPreferenceDao) {
        this.customerPreferenceDao = customerPreferenceDao;
    }

    // ...

}
Spring框架参考文档 Spring Framework Reference Documentation

As of Spring Framework 4.3, the @Autowired constructor is no longer necessary if the target bean only defines one constructor. If several constructors are available, at least one must be annotated to teach the container which one it has to use.

As expected, you can also apply the @Autowired annotation to "traditional" setter methods:

public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Autowired
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // ...

}

You can also apply the annotation to methods with arbitrary names and/or multiple arguments:

public class MovieRecommender {

    private MovieCatalog movieCatalog;

    private CustomerPreferenceDao customerPreferenceDao;

    @Autowired
    public void prepare(MovieCatalog movieCatalog,
            CustomerPreferenceDao customerPreferenceDao) {
        this.movieCatalog = movieCatalog;
        this.customerPreferenceDao = customerPreferenceDao;
    }

    // ...

}

You can apply @Autowired to fields as well and even mix it with constructors:

public class MovieRecommender {

    private final CustomerPreferenceDao customerPreferenceDao;

    @Autowired
    private MovieCatalog movieCatalog;

    @Autowired
    public MovieRecommender(CustomerPreferenceDao customerPreferenceDao) {
        this.customerPreferenceDao = customerPreferenceDao;
    }

    // ...

}

It is also possible to provide all beans of a particular type from the ApplicationContext by adding the annotation to a field or method that expects an array of that type:

public class MovieRecommender {

    @Autowired
    private MovieCatalog[] movieCatalogs;

    // ...

}

The same applies for typed collections:

public class MovieRecommender {

    private Set<MovieCatalog> movieCatalogs;

    @Autowired
    public void setMovieCatalogs(Set<MovieCatalog> movieCatalogs) {
        this.movieCatalogs = movieCatalogs;
    }

    // ...

}
Spring框架参考文档 Spring Framework Reference Documentation

Your beans can implement the org.springframework.core.Ordered interface or either use the @Order or standard @Priority annotation if you want items in the array or list to be sorted into a specific order.

Even typed Maps can be autowired as long as the expected key type is String. The Map values will contain all beans of the expected type, and the keys will contain the corresponding bean names:

public class MovieRecommender {

    private Map<String, MovieCatalog> movieCatalogs;

    @Autowired
    public void setMovieCatalogs(Map<String, MovieCatalog> movieCatalogs) {
        this.movieCatalogs = movieCatalogs;
    }

    // ...

}

By default, the autowiring fails whenever zero candidate beans are available; the default behavior is to treat annotated methods, constructors, and fields as indicatingrequired dependencies. This behavior can be changed as demonstrated below.

public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Autowired(required=false)
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // ...

}
Spring框架参考文档 Spring Framework Reference Documentation

Only one annotated constructor per-class can be marked as required, but multiple non-required constructors can be annotated. In that case, each is considered among the candidates and Spring uses the greediest constructor whose dependencies can be satisfied, that is the constructor that has the largest number of arguments.

@Autowired’s required attribute is recommended over the `@Required annotation. The required attribute indicates that the property is not required for autowiring purposes, the property is ignored if it cannot be autowired. @Required, on the other hand, is stronger in that it enforces the property that was set by any means supported by the container. If no value is injected, a corresponding exception is raised.

You can also use @Autowired for interfaces that are well-known resolvable dependencies: BeanFactory, ApplicationContext, Environment, ResourceLoader,ApplicationEventPublisher, and MessageSource. These interfaces and their extended interfaces, such as ConfigurableApplicationContext orResourcePatternResolver, are automatically resolved, with no special setup necessary.

public class MovieRecommender {

    @Autowired
    private ApplicationContext context;

    public MovieRecommender() {
    }

    // ...

}
Spring框架参考文档 Spring Framework Reference Documentation

@Autowired, @Inject, @Resource, and @Value annotations are handled by Spring BeanPostProcessor implementations which in turn means that you cannot apply these annotations within your own BeanPostProcessor or BeanFactoryPostProcessor types (if any). These types must be 'wired up' explicitly via XML or using a Spring @Bean method.

7.9.3 Fine-tuning annotation-based autowiring with @Primary

Because autowiring by type may lead to multiple candidates, it is often necessary to have more control over the selection process. One way to accomplish this is with Spring’s @Primary annotation. @Primary indicates that a particular bean should be given preference when multiple beans are candidates to be autowired to a single-valued dependency. If exactly one 'primary' bean exists among the candidates, it will be the autowired value.

Let’s assume we have the following configuration that defines firstMovieCatalog as the primary MovieCatalog.

@Configuration
public class MovieConfiguration {

    @Bean
    @Primary
    public MovieCatalog firstMovieCatalog() { ... }

    @Bean
    public MovieCatalog secondMovieCatalog() { ... }

    // ...

}

With such configuration, the following MovieRecommender will be autowired with the firstMovieCatalog.

public class MovieRecommender {

    @Autowired
    private MovieCatalog movieCatalog;

    // ...

}

The corresponding bean definitions appear as follows.

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:annotation-config/>

    <bean class="example.SimpleMovieCatalog" primary="true">
        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean class="example.SimpleMovieCatalog">
        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean id="movieRecommender" class="example.MovieRecommender"/>

</beans>

7.9.4 Fine-tuning annotation-based autowiring with qualifiers

@Primary is an effective way to use autowiring by type with several instances when one primary candidate can be determined. When more control over the selection process is required, Spring’s @Qualifier annotation can be used. You can associate qualifier values with specific arguments, narrowing the set of type matches so that a specific bean is chosen for each argument. In the simplest case, this can be a plain descriptive value:

public class MovieRecommender {

    @Autowired
    @Qualifier("main")
    private MovieCatalog movieCatalog;

    // ...

}

The @Qualifier annotation can also be specified on individual constructor arguments or method parameters:

public class MovieRecommender {

    private MovieCatalog movieCatalog;

    private CustomerPreferenceDao customerPreferenceDao;

    @Autowired
    public void prepare(@Qualifier("main")MovieCatalog movieCatalog,
            CustomerPreferenceDao customerPreferenceDao) {
        this.movieCatalog = movieCatalog;
        this.customerPreferenceDao = customerPreferenceDao;
    }

    // ...

}

The corresponding bean definitions appear as follows. The bean with qualifier value "main" is wired with the constructor argument that is qualified with the same value.

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:annotation-config/>

    <bean class="example.SimpleMovieCatalog">
        <qualifier value="main"/>

        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean class="example.SimpleMovieCatalog">
        <qualifier value="action"/>

        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean id="movieRecommender" class="example.MovieRecommender"/>

</beans>

For a fallback match, the bean name is considered a default qualifier value. Thus you can define the bean with an id "main" instead of the nested qualifier element, leading to the same matching result. However, although you can use this convention to refer to specific beans by name, @Autowired is fundamentally about type-driven injection with optional semantic qualifiers. This means that qualifier values, even with the bean name fallback, always have narrowing semantics within the set of type matches; they do not semantically express a reference to a unique bean id. Good qualifier values are "main" or "EMEA" or "persistent", expressing characteristics of a specific component that are independent from the bean id, which may be auto-generated in case of an anonymous bean definition like the one in the preceding example.

Qualifiers also apply to typed collections, as discussed above, for example, to Set<MovieCatalog>. In this case, all matching beans according to the declared qualifiers are injected as a collection. This implies that qualifiers do not have to be unique; they rather simply constitute filtering criteria. For example, you can define multipleMovieCatalog beans with the same qualifier value "action", all of which would be injected into a Set<MovieCatalog> annotated with @Qualifier("action").

Spring框架参考文档 Spring Framework Reference Documentation

If you intend to express annotation-driven injection by name, do not primarily use @Autowired, even if is technically capable of referring to a bean name through @Qualifier values. Instead, use the JSR-250 @Resource annotation, which is semantically defined to identify a specific target component by its unique name, with the declared type being irrelevant for the matching process. @Autowired has rather different semantics: After selecting candidate beans by type, the specified String qualifier value will be considered within those type-selected candidates only, e.g. matching an "account" qualifier against beans marked with the same qualifier label.

For beans that are themselves defined as a collection/map or array type, @Resource is a fine solution, referring to the specific collection or array bean by unique name. That said, as of 4.3, collection/map and array types can be matched through Spring’s @Autowired type matching algorithm as well, as long as the element type information is preserved in @Bean return type signatures or collection inheritance hierarchies. In this case, qualifier values can be used to select among same-typed collections, as outlined in the previous paragraph.

As of 4.3, @Autowired also considers self references for injection, i.e. references back to the bean that is currently injected. Note that self injection is a fallback; regular dependencies on other components always have precedence. In that sense, self references do not participate in regular candidate selection and are therefore in particular never primary; on the contrary, they always end up as lowest precedence. In practice, use self references as a last resort only, e.g. for calling other methods on the same instance through the bean’s transactional proxy: Consider factoring out the affected methods to a separate delegate bean in such a scenario. Alternatively, use @Resource which may obtain a proxy back to the current bean by its unique name.

@Autowired applies to fields, constructors, and multi-argument methods, allowing for narrowing through qualifier annotations at the parameter level. By contrast, @Resource is supported only for fields and bean property setter methods with a single argument. As a consequence, stick with qualifiers if your injection target is a constructor or a multi-argument method.

You can create your own custom qualifier annotations. Simply define an annotation and provide the @Qualifier annotation within your definition:

@Target({ElementType.FIELD, ElementType.PARAMETER})
@Retention(RetentionPolicy.RUNTIME)
@Qualifier
public @interface Genre {

    String value();
}

Then you can provide the custom qualifier on autowired fields and parameters:

public class MovieRecommender {

    @Autowired
    @Genre("Action")
    private MovieCatalog actionCatalog;
    private MovieCatalog comedyCatalog;

    @Autowired
    public void setComedyCatalog(@Genre("Comedy") MovieCatalog comedyCatalog) {
        this.comedyCatalog = comedyCatalog;
    }

    // ...

}

Next, provide the information for the candidate bean definitions. You can add <qualifier/> tags as sub-elements of the <bean/> tag and then specify the type andvalue to match your custom qualifier annotations. The type is matched against the fully-qualified class name of the annotation. Or, as a convenience if no risk of conflicting names exists, you can use the short class name. Both approaches are demonstrated in the following example.

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:annotation-config/>

    <bean class="example.SimpleMovieCatalog">
        <qualifier type="Genre" value="Action"/>
        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean class="example.SimpleMovieCatalog">
        <qualifier type="example.Genre" value="Comedy"/>
        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean id="movieRecommender" class="example.MovieRecommender"/>

</beans>

In Section 7.10, “Classpath scanning and managed components”, you will see an annotation-based alternative to providing the qualifier metadata in XML. Specifically, see Section 7.10.8, “Providing qualifier metadata with annotations”.

In some cases, it may be sufficient to use an annotation without a value. This may be useful when the annotation serves a more generic purpose and can be applied across several different types of dependencies. For example, you may provide an offline catalog that would be searched when no Internet connection is available. First define the simple annotation:

@Target({ElementType.FIELD, ElementType.PARAMETER})
@Retention(RetentionPolicy.RUNTIME)
@Qualifier
public @interface Offline {

}

Then add the annotation to the field or property to be autowired:

public class MovieRecommender {

    @Autowired
    @Offline
    private MovieCatalog offlineCatalog;

    // ...

}

Now the bean definition only needs a qualifier type:

<bean class="example.SimpleMovieCatalog">
    <qualifier type="Offline"/>
    <!-- inject any dependencies required by this bean -->
</bean>

You can also define custom qualifier annotations that accept named attributes in addition to or instead of the simple value attribute. If multiple attribute values are then specified on a field or parameter to be autowired, a bean definition must match all such attribute values to be considered an autowire candidate. As an example, consider the following annotation definition:

@Target({ElementType.FIELD, ElementType.PARAMETER})
@Retention(RetentionPolicy.RUNTIME)
@Qualifier
public @interface MovieQualifier {

    String genre();

    Format format();

}

In this case Format is an enum:

public enum Format {
    VHS, DVD, BLURAY
}

The fields to be autowired are annotated with the custom qualifier and include values for both attributes: genre and format.

public class MovieRecommender {

    @Autowired
    @MovieQualifier(format=Format.VHS, genre="Action")
    private MovieCatalog actionVhsCatalog;

    @Autowired
    @MovieQualifier(format=Format.VHS, genre="Comedy")
    private MovieCatalog comedyVhsCatalog;

    @Autowired
    @MovieQualifier(format=Format.DVD, genre="Action")
    private MovieCatalog actionDvdCatalog;

    @Autowired
    @MovieQualifier(format=Format.BLURAY, genre="Comedy")
    private MovieCatalog comedyBluRayCatalog;

    // ...

}

Finally, the bean definitions should contain matching qualifier values. This example also demonstrates that bean meta attributes may be used instead of the<qualifier/> sub-elements. If available, the <qualifier/> and its attributes take precedence, but the autowiring mechanism falls back on the values provided within the <meta/> tags if no such qualifier is present, as in the last two bean definitions in the following example.

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:annotation-config/>

    <bean class="example.SimpleMovieCatalog">
        <qualifier type="MovieQualifier">
            <attribute key="format" value="VHS"/>
            <attribute key="genre" value="Action"/>
        </qualifier>
        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean class="example.SimpleMovieCatalog">
        <qualifier type="MovieQualifier">
            <attribute key="format" value="VHS"/>
            <attribute key="genre" value="Comedy"/>
        </qualifier>
        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean class="example.SimpleMovieCatalog">
        <meta key="format" value="DVD"/>
        <meta key="genre" value="Action"/>
        <!-- inject any dependencies required by this bean -->
    </bean>

    <bean class="example.SimpleMovieCatalog">
        <meta key="format" value="BLURAY"/>
        <meta key="genre" value="Comedy"/>
        <!-- inject any dependencies required by this bean -->
    </bean>

</beans>

7.9.5 Using generics as autowiring qualifiers

In addition to the @Qualifier annotation, it is also possible to use Java generic types as an implicit form of qualification. For example, suppose you have the following configuration:

@Configuration
public class MyConfiguration {

    @Bean
    public StringStore stringStore() {
        return new StringStore();
    }

    @Bean
    public IntegerStore integerStore() {
        return new IntegerStore();
    }

}

Assuming that beans above implement a generic interface, i.e. Store<String> and Store<Integer>, you can @Autowire the Store interface and the generic will be used as a qualifier:

@Autowired
private Store<String> s1; // <String> qualifier, injects the stringStore bean

@Autowired
private Store<Integer> s2; // <Integer> qualifier, injects the integerStore bean

Generic qualifiers also apply when autowiring Lists, Maps and Arrays:

// Inject all Store beans as long as they have an <Integer> generic
// Store<String> beans will not appear in this list
@Autowired
private List<Store<Integer>> s;

7.9.6 CustomAutowireConfigurer

The CustomAutowireConfigurer is a BeanFactoryPostProcessor that enables you to register your own custom qualifier annotation types even if they are not annotated with Spring’s @Qualifier annotation.

<bean id="customAutowireConfigurer"
        class="org.springframework.beans.factory.annotation.CustomAutowireConfigurer">
    <property name="customQualifierTypes">
        <set>
            <value>example.CustomQualifier</value>
        </set>
    </property>
</bean>

The AutowireCandidateResolver determines autowire candidates by:

  • the autowire-candidate value of each bean definition
  • any default-autowire-candidates pattern(s) available on the <beans/> element
  • the presence of @Qualifier annotations and any custom annotations registered with the CustomAutowireConfigurer

When multiple beans qualify as autowire candidates, the determination of a "primary" is the following: if exactly one bean definition among the candidates has aprimary attribute set to true, it will be selected.

7.9.7 @Resource

Spring also supports injection using the JSR-250 @Resource annotation on fields or bean property setter methods. This is a common pattern in Java EE 5 and 6, for example in JSF 1.2 managed beans or JAX-WS 2.0 endpoints. Spring supports this pattern for Spring-managed objects as well.

@Resource takes a name attribute, and by default Spring interprets that value as the bean name to be injected. In other words, it follows by-name semantics, as demonstrated in this example:

public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Resource(name="myMovieFinder")
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

}

If no name is specified explicitly, the default name is derived from the field name or setter method. In case of a field, it takes the field name; in case of a setter method, it takes the bean property name. So the following example is going to have the bean with name "movieFinder" injected into its setter method:

public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Resource
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

}
Spring框架参考文档 Spring Framework Reference Documentation

The name provided with the annotation is resolved as a bean name by the ApplicationContext of which the CommonAnnotationBeanPostProcessoris aware. The names can be resolved through JNDI if you configure Spring’s SimpleJndiBeanFactory explicitly. However, it is recommended that you rely on the default behavior and simply use Spring’s JNDI lookup capabilities to preserve the level of indirection.

In the exclusive case of @Resource usage with no explicit name specified, and similar to @Autowired, @Resource finds a primary type match instead of a specific named bean and resolves well-known resolvable dependencies: the BeanFactory, ApplicationContext, ResourceLoader, ApplicationEventPublisher, andMessageSource interfaces.

Thus in the following example, the customerPreferenceDao field first looks for a bean named customerPreferenceDao, then falls back to a primary type match for the type CustomerPreferenceDao. The "context" field is injected based on the known resolvable dependency type ApplicationContext.

public class MovieRecommender {

    @Resource
    private CustomerPreferenceDao customerPreferenceDao;

    @Resource
    private ApplicationContext context;

    public MovieRecommender() {
    }

    // ...

}

7.9.8 @PostConstruct and @PreDestroy

The CommonAnnotationBeanPostProcessor not only recognizes the @Resource annotation but also the JSR-250 lifecycle annotations. Introduced in Spring 2.5, the support for these annotations offers yet another alternative to those described in initialization callbacks and destruction callbacks. Provided that theCommonAnnotationBeanPostProcessor is registered within the Spring ApplicationContext, a method carrying one of these annotations is invoked at the same point in the lifecycle as the corresponding Spring lifecycle interface method or explicitly declared callback method. In the example below, the cache will be pre-populated upon initialization and cleared upon destruction.

public class CachingMovieLister {

    @PostConstruct
    public void populateMovieCache() {
        // populates the movie cache upon initialization...
    }

    @PreDestroy
    public void clearMovieCache() {
        // clears the movie cache upon destruction...
    }

}
Spring框架参考文档 Spring Framework Reference Documentation

For details about the effects of combining various lifecycle mechanisms, see the section called “Combining lifecycle mechanisms”.

7.10 Classpath scanning and managed components

Most examples in this chapter use XML to specify the configuration metadata that produces each BeanDefinition within the Spring container. The previous section (Section 7.9, “Annotation-based container configuration”) demonstrates how to provide a lot of the configuration metadata through source-level annotations. Even in those examples, however, the "base" bean definitions are explicitly defined in the XML file, while the annotations only drive the dependency injection. This section describes an option for implicitly detecting the candidate components by scanning the classpath. Candidate components are classes that match against a filter criteria and have a corresponding bean definition registered with the container. This removes the need to use XML to perform bean registration; instead you can use annotations (for example @Component), AspectJ type expressions, or your own custom filter criteria to select which classes will have bean definitions registered with the container.

Spring框架参考文档 Spring Framework Reference Documentation

Starting with Spring 3.0, many features provided by the Spring JavaConfig project are part of the core Spring Framework. This allows you to define beans using Java rather than using the traditional XML files. Take a look at the @Configuration, @Bean, @Import, and @DependsOn annotations for examples of how to use these new features.

7.10.1 @Component and further stereotype annotations

The @Repository annotation is a marker for any class that fulfills the role or stereotype of a repository (also known as Data Access Object or DAO). Among the uses of this marker is the automatic translation of exceptions as described in Section 20.2.2, “Exception translation”.

Spring provides further stereotype annotations: @Component, @Service, and @Controller. @Component is a generic stereotype for any Spring-managed component.@Repository, @Service, and @Controller are specializations of @Component for more specific use cases, for example, in the persistence, service, and presentation layers, respectively. Therefore, you can annotate your component classes with @Component, but by annotating them with @Repository, @Service, or@Controller instead, your classes are more properly suited for processing by tools or associating with aspects. For example, these stereotype annotations make ideal targets for pointcuts. It is also possible that @Repository, @Service, and @Controller may carry additional semantics in future releases of the Spring Framework. Thus, if you are choosing between using @Component or @Service for your service layer, @Service is clearly the better choice. Similarly, as stated above,@Repository is already supported as a marker for automatic exception translation in your persistence layer.

7.10.2 Meta-annotations

Many of the annotations provided by Spring can be used as meta-annotations in your own code. A meta-annotation is simply an annotation that can be applied to another annotation. For example, the @Service annotation mentioned above is meta-annotated with @Component:

@Target(ElementType.TYPE)
@Retention(RetentionPolicy.RUNTIME)
@Documented
@Component // Spring will see this and treat @Service in the same way as @Component
public @interface Service {

    // ....
}

Meta-annotations can also be combined to create composed annotations. For example, the @RestController annotation from Spring MVC is composed of@Controller and @ResponseBody.

In addition, composed annotations may optionally redeclare attributes from meta-annotations to allow user customization. This can be particularly useful when you want to only expose a subset of the meta-annotation’s attributes. For example, Spring’s @SessionScope annotation hardcodes the scope name to session but still allows customization of the proxyMode.

@Target({ElementType.TYPE, ElementType.METHOD})
@Retention(RetentionPolicy.RUNTIME)
@Documented
@Scope(WebApplicationContext.SCOPE_SESSION)
public @interface SessionScope {

    /**
     * Alias for {@link Scope#proxyMode}.
     * <p>Defaults to {@link ScopedProxyMode#TARGET_CLASS}.
     */
    @AliasFor(annotation = Scope.class)
    ScopedProxyMode proxyMode() default ScopedProxyMode.TARGET_CLASS;

}

@SessionScope can then be used without declaring the proxyMode as follows:

@Service
@SessionScope
public class SessionScopedService {
    // ...
}

Or with an overridden value for the proxyMode as follows:

@Service
@SessionScope(proxyMode = ScopedProxyMode.INTERFACES)
public class SessionScopedUserService implements UserService {
    // ...
}

For further details, consult the Spring Annotation Programming Model.

7.10.3 Automatically detecting classes and registering bean definitions

Spring can automatically detect stereotyped classes and register corresponding BeanDefinitions with the ApplicationContext. For example, the following two classes are eligible for such autodetection:

@Service
public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Autowired
    public SimpleMovieLister(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

}
@Repository
public class JpaMovieFinder implements MovieFinder {
    // implementation elided for clarity
}

To autodetect these classes and register the corresponding beans, you need to add @ComponentScan to your @Configuration class, where the basePackagesattribute is a common parent package for the two classes. (Alternatively, you can specify a comma/semicolon/space-separated list that includes the parent package of each class.)

@Configuration
@ComponentScan(basePackages = "org.example")
public class AppConfig  {
    ...
}
Spring框架参考文档 Spring Framework Reference Documentation

for concision, the above may have used the value attribute of the annotation, i.e. ComponentScan("org.example")

The following is an alternative using XML

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:component-scan base-package="org.example"/>

</beans>
Spring框架参考文档 Spring Framework Reference Documentation

The use of <context:component-scan> implicitly enables the functionality of <context:annotation-config>. There is usually no need to include the<context:annotation-config> element when using <context:component-scan>.
Spring框架参考文档 Spring Framework Reference Documentation

The scanning of classpath packages requires the presence of corresponding directory entries in the classpath. When you build JARs with Ant, make sure that you do not activate the files-only switch of the JAR task. Also, classpath directories may not get exposed based on security policies in some environments, e.g. standalone apps on JDK 1.7.0_45 and higher (which requires 'Trusted-Library' setup in your manifests; seehttp://stackoverflow.com/questions/19394570/java-jre-7u45-breaks-classloader-getresources).

Furthermore, the AutowiredAnnotationBeanPostProcessor and CommonAnnotationBeanPostProcessor are both included implicitly when you use the component-scan element. That means that the two components are autodetected and wired together - all without any bean configuration metadata provided in XML.

Spring框架参考文档 Spring Framework Reference Documentation

You can disable the registration of AutowiredAnnotationBeanPostProcessor and CommonAnnotationBeanPostProcessor by including theannotation-config attribute with a value of false.

7.10.4 Using filters to customize scanning

By default, classes annotated with @Component, @Repository, @Service, @Controller, or a custom annotation that itself is annotated with @Component are the only detected candidate components. However, you can modify and extend this behavior simply by applying custom filters. Add them as includeFilters or excludeFiltersparameters of the @ComponentScan annotation (or as include-filter or exclude-filter sub-elements of the component-scan element). Each filter element requires thetype and expression attributes. The following table describes the filtering options.

Table 7.5. Filter Types

Filter Type Example Expression Description

annotation (default)

org.example.SomeAnnotation

An annotation to be present at the type level in target components.

assignable

org.example.SomeClass

A class (or interface) that the target components are assignable to (extend/implement).

aspectj

org.example..*Service+

An AspectJ type expression to be matched by the target components.

regex

org\.example\.Default.*

A regex expression to be matched by the target components class names.

custom

org.example.MyTypeFilter

A custom implementation of the org.springframework.core.type .TypeFilter interface.

The following example shows the configuration ignoring all @Repository annotations and using "stub" repositories instead.

@Configuration
@ComponentScan(basePackages = "org.example",
        includeFilters = @Filter(type = FilterType.REGEX, pattern = ".*Stub.*Repository"),
        excludeFilters = @Filter(Repository.class))
public class AppConfig {
    ...
}

and the equivalent using XML

<beans>
    <context:component-scan base-package="org.example">
        <context:include-filter type="regex"
                expression=".*Stub.*Repository"/>
        <context:exclude-filter type="annotation"
                expression="org.springframework.stereotype.Repository"/>
    </context:component-scan>
</beans>
Spring框架参考文档 Spring Framework Reference Documentation

You can also disable the default filters by setting useDefaultFilters=false on the annotation or providing use-default-filters="false" as an attribute of the <component-scan/> element. This will in effect disable automatic detection of classes annotated with @Component, @Repository,@Service, @Controller, or @Configuration.

7.10.5 Defining bean metadata within components

Spring components can also contribute bean definition metadata to the container. You do this with the same @Bean annotation used to define bean metadata within@Configuration annotated classes. Here is a simple example:

@Component
public class FactoryMethodComponent {

    @Bean
    @Qualifier("public")
    public TestBean publicInstance() {
        return new TestBean("publicInstance");
    }

    public void doWork() {
        // Component method implementation omitted
    }

}

This class is a Spring component that has application-specific code contained in its doWork() method. However, it also contributes a bean definition that has a factory method referring to the method publicInstance(). The @Bean annotation identifies the factory method and other bean definition properties, such as a qualifier value through the @Qualifier annotation. Other method level annotations that can be specified are @Scope, @Lazy, and custom qualifier annotations.

Spring框架参考文档 Spring Framework Reference Documentation

In addition to its role for component initialization, the @Lazy annotation may also be placed on injection points marked with @Autowired or @Inject. In this context, it leads to the injection of a lazy-resolution proxy.

Autowired fields and methods are supported as previously discussed, with additional support for autowiring of @Bean methods:

@Component
public class FactoryMethodComponent {

    private static int i;

    @Bean
    @Qualifier("public")
    public TestBean publicInstance() {
        return new TestBean("publicInstance");
    }

    // use of a custom qualifier and autowiring of method parameters

    @Bean
    protected TestBean protectedInstance(
            @Qualifier("public") TestBean spouse,
            @Value("#{privateInstance.age}") String country) {
        TestBean tb = new TestBean("protectedInstance", 1);
        tb.setSpouse(spouse);
        tb.setCountry(country);
        return tb;
    }

    @Bean
    private TestBean privateInstance() {
        return new TestBean("privateInstance", i++);
    }

    @Bean
    @RequestScope
    public TestBean requestScopedInstance() {
        return new TestBean("requestScopedInstance", 3);
    }

}

The example autowires the String method parameter country to the value of the Age property on another bean named privateInstance. A Spring Expression Language element defines the value of the property through the notation #{ <expression> }. For @Value annotations, an expression resolver is preconfigured to look for bean names when resolving expression text.

The @Bean methods in a Spring component are processed differently than their counterparts inside a Spring @Configuration class. The difference is that@Component classes are not enhanced with CGLIB to intercept the invocation of methods and fields. CGLIB proxying is the means by which invoking methods or fields within @Bean methods in @Configuration classes creates bean metadata references to collaborating objects; such methods are not invoked with normal Java semantics but rather go through the container in order to provide the usual lifecycle management and proxying of Spring beans even when referring to other beans via programmatic calls to @Bean methods. In contrast, invoking a method or field in an @Bean method within a plain @Component class has standard Java semantics, with no special CGLIB processing or other constraints applying.

Spring框架参考文档 Spring Framework Reference Documentation

You may declare @Bean methods as static, allowing for them to be called without creating their containing configuration class as an instance. This makes particular sense when defining post-processor beans, e.g. of type BeanFactoryPostProcessor or BeanPostProcessor, since such beans will get initialized early in the container lifecycle and should avoid triggering other parts of the configuration at that point.

Note that calls to static @Bean methods will never get intercepted by the container, not even within @Configuration classes (see above). This is due to technical limitations: CGLIB subclassing can only override non-static methods. As a consequence, a direct call to another @Bean method will have standard Java semantics, resulting in an independent instance being returned straight from the factory method itself.

The Java language visibility of @Bean methods does not have an immediate impact on the resulting bean definition in Spring’s container. You may freely declare your factory methods as you see fit in non-@Configuration classes and also for static methods anywhere. However, regular @Bean methods in@Configuration classes need to be overridable, i.e. they must not be declared as private or final.

@Bean methods will also be discovered on base classes of a given component or configuration class, as well as on Java 8 default methods declared in interfaces implemented by the component or configuration class. This allows for a lot of flexibility in composing complex configuration arrangements, with even multiple inheritance being possible through Java 8 default methods as of Spring 4.2.

Finally, note that a single class may hold multiple @Bean methods for the same bean, as an arrangement of multiple factory methods to use depending on available dependencies at runtime. This is the same algorithm as for choosing the "greediest" constructor or factory method in other configuration scenarios: The variant with the largest number of satisfiable dependencies will be picked at construction time, analogous to how the container selects between multiple @Autowired constructors.

7.10.6 Naming autodetected components

When a component is autodetected as part of the scanning process, its bean name is generated by the BeanNameGenerator strategy known to that scanner. By default, any Spring stereotype annotation (@Component, @Repository, @Service, and @Controller) that contains a name value will thereby provide that name to the corresponding bean definition.

If such an annotation contains no name value or for any other detected component (such as those discovered by custom filters), the default bean name generator returns the uncapitalized non-qualified class name. For example, if the following two components were detected, the names would be myMovieLister andmovieFinderImpl:

@Service("myMovieLister")
public class SimpleMovieLister {
    // ...
}
@Repository
public class MovieFinderImpl implements MovieFinder {
    // ...
}
Spring框架参考文档 Spring Framework Reference Documentation

If you do not want to rely on the default bean-naming strategy, you can provide a custom bean-naming strategy. First, implement the BeanNameGeneratorinterface, and be sure to include a default no-arg constructor. Then, provide the fully-qualified class name when configuring the scanner:

 
@Configuration
@ComponentScan(basePackages = "org.example", nameGenerator = MyNameGenerator.class)
public class AppConfig {
    ...
}
<beans>
    <context:component-scan base-package="org.example"
        name-generator="org.example.MyNameGenerator" />
</beans>

As a general rule, consider specifying the name with the annotation whenever other components may be making explicit references to it. On the other hand, the auto-generated names are adequate whenever the container is responsible for wiring.

7.10.7 Providing a scope for autodetected components

As with Spring-managed components in general, the default and most common scope for autodetected components is singleton. However, sometimes you need a different scope which can be specified via the @Scope annotation. Simply provide the name of the scope within the annotation:

@Scope("prototype")
@Repository
public class MovieFinderImpl implements MovieFinder {
    // ...
}

For details on web-specific scopes, see Section 7.5.4, “Request, session, global session, application, and WebSocket scopes”.

Spring框架参考文档 Spring Framework Reference Documentation

To provide a custom strategy for scope resolution rather than relying on the annotation-based approach, implement the ScopeMetadataResolverinterface, and be sure to include a default no-arg constructor. Then, provide the fully-qualified class name when configuring the scanner:

 
@Configuration
@ComponentScan(basePackages = "org.example", scopeResolver = MyScopeResolver.class)
public class AppConfig {
    ...
}
<beans>
    <context:component-scan base-package="org.example"
            scope-resolver="org.example.MyScopeResolver" />
</beans>

When using certain non-singleton scopes, it may be necessary to generate proxies for the scoped objects. The reasoning is described in the section called “Scoped beans as dependencies”. For this purpose, a scoped-proxy attribute is available on the component-scan element. The three possible values are: no, interfaces, and targetClass. For example, the following configuration will result in standard JDK dynamic proxies:

@Configuration
@ComponentScan(basePackages = "org.example", scopedProxy = ScopedProxyMode.INTERFACES)
public class AppConfig {
    ...
}
<beans>
    <context:component-scan base-package="org.example"
        scoped-proxy="interfaces" />
</beans>

7.10.8 Providing qualifier metadata with annotations

The @Qualifier annotation is discussed in Section 7.9.4, “Fine-tuning annotation-based autowiring with qualifiers”. The examples in that section demonstrate the use of the @Qualifier annotation and custom qualifier annotations to provide fine-grained control when you resolve autowire candidates. Because those examples were based on XML bean definitions, the qualifier metadata was provided on the candidate bean definitions using the qualifier or meta sub-elements of the beanelement in the XML. When relying upon classpath scanning for autodetection of components, you provide the qualifier metadata with type-level annotations on the candidate class. The following three examples demonstrate this technique:

@Component
@Qualifier("Action")
public class ActionMovieCatalog implements MovieCatalog {
    // ...
}
@Component
@Genre("Action")
public class ActionMovieCatalog implements MovieCatalog {
    // ...
}
@Component
@Offline
public class CachingMovieCatalog implements MovieCatalog {
    // ...
}
Spring框架参考文档 Spring Framework Reference Documentation

As with most annotation-based alternatives, keep in mind that the annotation metadata is bound to the class definition itself, while the use of XML allows for multiple beans of the same type to provide variations in their qualifier metadata, because that metadata is provided per-instance rather than per-class.

7.11 Using JSR 330 Standard Annotations

Starting with Spring 3.0, Spring offers support for JSR-330 standard annotations (Dependency Injection). Those annotations are scanned in the same way as the Spring annotations. You just need to have the relevant jars in your classpath.

Spring框架参考文档 Spring Framework Reference Documentation

If you are using Maven, the javax.inject artifact is available in the standard Maven repository (http://repo1.maven.org/maven2/javax/inject/javax.inject/1/). You can add the following dependency to your file pom.xml:

<dependency>
    <groupId>javax.inject</groupId>
    <artifactId>javax.inject</artifactId>
    <version>1</version>
</dependency>

7.11.1 Dependency Injection with @Inject and @Named

Instead of @Autowired, @javax.inject.Inject may be used as follows:

import javax.inject.Inject;

public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Inject
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    public void listMovies() {
        this.movieFinder.findMovies(...);
        ...
    }
}

As with @Autowired, it is possible to use @Inject at the field level, method level and constructor-argument level. Furthermore, you may declare your injection point as a Provider, allowing for on-demand access to beans of shorter scopes or lazy access to other beans through a Provider.get() call. As a variant of the example above:

import javax.inject.Inject;
import javax.inject.Provider;

public class SimpleMovieLister {

    private Provider<MovieFinder> movieFinder;

    public void listMovies() {
        this.movieFinder.get().findMovies(...);
        ...
    }
}

If you would like to use a qualified name for the dependency that should be injected, you should use the @Named annotation as follows:

import javax.inject.Inject;
import javax.inject.Named;

public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Inject
    public void setMovieFinder(@Named("main") MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // ...
}

7.11.2 @Named: a standard equivalent to the @Component annotation

Instead of @Component, @javax.inject.Named may be used as follows:

import javax.inject.Inject;
import javax.inject.Named;

@Named("movieListener")
public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Inject
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // ...
}

It is very common to use @Component without specifying a name for the component. @Named can be used in a similar fashion:

import javax.inject.Inject;
import javax.inject.Named;

@Named
public class SimpleMovieLister {

    private MovieFinder movieFinder;

    @Inject
    public void setMovieFinder(MovieFinder movieFinder) {
        this.movieFinder = movieFinder;
    }

    // ...
}

When using @Named, it is possible to use component scanning in the exact same way as when using Spring annotations:

@Configuration
@ComponentScan(basePackages = "org.example")
public class AppConfig  {
    ...
}
Spring框架参考文档 Spring Framework Reference Documentation

In contrast to @Component, the JSR-330 @Named annotation is not composable. Please use Spring’s stereotype model for building custom component annotations.

7.11.3 Limitations of JSR-330 standard annotations

When working with standard annotations, it is important to know that some significant features are not available as shown in the table below:

Table 7.6. Spring component model elements vs. JSR-330 variants

Spring javax.inject.* javax.inject restrictions / comments

@Autowired

@Inject

@Inject has no 'required' attribute; can be used with Java 8’s Optional instead.

@Component

@Named

JSR-330 does not provide a composable model, just a way to identify named components.

@Scope("singleton")

@Singleton

The JSR-330 default scope is like Spring’s prototype. However, in order to keep it consistent with Spring’s general defaults, a JSR-330 bean declared in the Spring container is a singleton by default. In order to use a scope other than singleton, you should use Spring’s @Scope annotation. javax.inject also provides a @Scope annotation. Nevertheless, this one is only intended to be used for creating your own annotations.

@Qualifier

@Qualifier / @Named

javax.inject.Qualifier is just a meta-annotation for building custom qualifiers. Concrete String qualifiers (like Spring’s @Qualifier with a value) can be associated through javax.inject.Named.

@Value

-

no equivalent

@Required

-

no equivalent

@Lazy

-

no equivalent

ObjectFactory

Provider

javax.inject.Provider is a direct alternative to Spring’s ObjectFactory, just with a shorter get() method name. It can also be used in combination with Spring’s @Autowired or with non-annotated constructors and setter methods.

7.12 Java-based container configuration

7.12.1 Basic concepts: @Bean and @Configuration

The central artifacts in Spring’s new Java-configuration support are @Configuration-annotated classes and @Bean-annotated methods.

The @Bean annotation is used to indicate that a method instantiates, configures and initializes a new object to be managed by the Spring IoC container. For those familiar with Spring’s <beans/> XML configuration the @Bean annotation plays the same role as the <bean/> element. You can use @Bean annotated methods with any Spring @Component, however, they are most often used with @Configuration beans.

Annotating a class with @Configuration indicates that its primary purpose is as a source of bean definitions. Furthermore, @Configuration classes allow inter-bean dependencies to be defined by simply calling other @Bean methods in the same class. The simplest possible @Configuration class would read as follows:

@Configuration
public class AppConfig {

    @Bean
    public MyService myService() {
        return new MyServiceImpl();
    }

}

The AppConfig class above would be equivalent to the following Spring <beans/> XML:

<beans>
    <bean id="myService" class="com.acme.services.MyServiceImpl"/>
</beans>

Full @Configuration vs 'lite' @Beans mode?

When @Bean methods are declared within classes that are not annotated with @Configuration they are referred to as being processed in a 'lite' mode. For example, bean methods declared in a @Component or even in a plain old class will be considered 'lite'.

Unlike full @Configuration, lite @Bean methods cannot easily declare inter-bean dependencies. Usually one @Bean method should not invoke another @Beanmethod when operating in 'lite' mode.

Only using @Bean methods within @Configuration classes is a recommended approach of ensuring that 'full' mode is always used. This will prevent the same@Bean method from accidentally being invoked multiple times and helps to reduce subtle bugs that can be hard to track down when operating in 'lite' mode.

The @Bean and @Configuration annotations will be discussed in depth in the sections below. First, however, we’ll cover the various ways of creating a spring container using Java-based configuration.

7.12.2 Instantiating the Spring container using AnnotationConfigApplicationContext

The sections below document Spring’s AnnotationConfigApplicationContext, new in Spring 3.0. This versatile ApplicationContext implementation is capable of accepting not only @Configuration classes as input, but also plain @Component classes and classes annotated with JSR-330 metadata.

When @Configuration classes are provided as input, the @Configuration class itself is registered as a bean definition, and all declared @Bean methods within the class are also registered as bean definitions.

When @Component and JSR-330 classes are provided, they are registered as bean definitions, and it is assumed that DI metadata such as @Autowired or @Injectare used within those classes where necessary.

Simple construction

In much the same way that Spring XML files are used as input when instantiating a ClassPathXmlApplicationContext, @Configuration classes may be used as input when instantiating an AnnotationConfigApplicationContext. This allows for completely XML-free usage of the Spring container:

public static void main(String[] args) {
    ApplicationContext ctx = new AnnotationConfigApplicationContext(AppConfig.class);
    MyService myService = ctx.getBean(MyService.class);
    myService.doStuff();
}

As mentioned above, AnnotationConfigApplicationContext is not limited to working only with @Configuration classes. Any @Component or JSR-330 annotated class may be supplied as input to the constructor. For example:

public static void main(String[] args) {
    ApplicationContext ctx = new AnnotationConfigApplicationContext(MyServiceImpl.class, Dependency1.class, Dependency2.class);
    MyService myService = ctx.getBean(MyService.class);
    myService.doStuff();
}

The above assumes that MyServiceImpl, Dependency1 and Dependency2 use Spring dependency injection annotations such as @Autowired.

Building the container programmatically using register(Class<?>…​)

An AnnotationConfigApplicationContext may be instantiated using a no-arg constructor and then configured using the register() method. This approach is particularly useful when programmatically building an AnnotationConfigApplicationContext.

public static void main(String[] args) {
    AnnotationConfigApplicationContext ctx = new AnnotationConfigApplicationContext();
    ctx.register(AppConfig.class, OtherConfig.class);
    ctx.register(AdditionalConfig.class);
    ctx.refresh();
    MyService myService = ctx.getBean(MyService.class);
    myService.doStuff();
}

Enabling component scanning with scan(String…​)

To enable component scanning, just annotate your @Configuration class as follows:

@Configuration
@ComponentScan(basePackages = "com.acme")
public class AppConfig  {
    ...
}
Spring框架参考文档 Spring Framework Reference Documentation

Experienced Spring users will be familiar with the XML declaration equivalent from Spring’s context: namespace

<beans>
    <context:component-scan base-package="com.acme"/>
</beans>

In the example above, the com.acme package will be scanned, looking for any @Component-annotated classes, and those classes will be registered as Spring bean definitions within the container. AnnotationConfigApplicationContext exposes the scan(String…​) method to allow for the same component-scanning functionality:

public static void main(String[] args) {
    AnnotationConfigApplicationContext ctx = new AnnotationConfigApplicationContext();
    ctx.scan("com.acme");
    ctx.refresh();
    MyService myService = ctx.getBean(MyService.class);
}
Spring框架参考文档 Spring Framework Reference Documentation

Remember that @Configuration classes are meta-annotated with @Component, so they are candidates for component-scanning! In the example above, assuming that AppConfig is declared within the com.acme package (or any package underneath), it will be picked up during the call to scan(), and upon refresh() all its @Bean methods will be processed and registered as bean definitions within the container.

Support for web applications with AnnotationConfigWebApplicationContext

A WebApplicationContext variant of AnnotationConfigApplicationContext is available with AnnotationConfigWebApplicationContext. This implementation may be used when configuring the Spring ContextLoaderListener servlet listener, Spring MVC DispatcherServlet, etc. What follows is a web.xml snippet that configures a typical Spring MVC web application. Note the use of the contextClass context-param and init-param:

<web-app>
    <!-- Configure ContextLoaderListener to use AnnotationConfigWebApplicationContext
        instead of the default XmlWebApplicationContext -->
    <context-param>
        <param-name>contextClass</param-name>
        <param-value>
            org.springframework.web.context.support.AnnotationConfigWebApplicationContext
        </param-value>
    </context-param>

    <!-- Configuration locations must consist of one or more comma- or space-delimited
        fully-qualified @Configuration classes. Fully-qualified packages may also be
        specified for component-scanning -->
    <context-param>
        <param-name>contextConfigLocation</param-name>
        <param-value>com.acme.AppConfig</param-value>
    </context-param>

    <!-- Bootstrap the root application context as usual using ContextLoaderListener -->
    <listener>
        <listener-class>org.springframework.web.context.ContextLoaderListener</listener-class>
    </listener>

    <!-- Declare a Spring MVC DispatcherServlet as usual -->
    <servlet>
        <servlet-name>dispatcher</servlet-name>
        <servlet-class>org.springframework.web.servlet.DispatcherServlet</servlet-class>
        <!-- Configure DispatcherServlet to use AnnotationConfigWebApplicationContext
            instead of the default XmlWebApplicationContext -->
        <init-param>
            <param-name>contextClass</param-name>
            <param-value>
                org.springframework.web.context.support.AnnotationConfigWebApplicationContext
            </param-value>
        </init-param>
        <!-- Again, config locations must consist of one or more comma- or space-delimited
            and fully-qualified @Configuration classes -->
        <init-param>
            <param-name>contextConfigLocation</param-name>
            <param-value>com.acme.web.MvcConfig</param-value>
        </init-param>
    </servlet>

    <!-- map all requests for /app/* to the dispatcher servlet -->
    <servlet-mapping>
        <servlet-name>dispatcher</servlet-name>
        <url-pattern>/app/*</url-pattern>
    </servlet-mapping>
</web-app>

7.12.3 Using the @Bean annotation

@Bean is a method-level annotation and a direct analog of the XML <bean/> element. The annotation supports some of the attributes offered by <bean/>, such as:init-method, destroy-method, autowiring and name.

You can use the @Bean annotation in a @Configuration-annotated or in a @Component-annotated class.

Declaring a bean

To declare a bean, simply annotate a method with the @Bean annotation. You use this method to register a bean definition within an ApplicationContext of the type specified as the method’s return value. By default, the bean name will be the same as the method name. The following is a simple example of a @Bean method declaration:

@Configuration
public class AppConfig {

    @Bean
    public TransferService transferService() {
        return new TransferServiceImpl();
    }

}

The preceding configuration is exactly equivalent to the following Spring XML:

<beans>
    <bean id="transferService" class="com.acme.TransferServiceImpl"/>
</beans>

Both declarations make a bean named transferService available in the ApplicationContext, bound to an object instance of type TransferServiceImpl:

transferService -> com.acme.TransferServiceImpl

Bean dependencies

A @Bean annotated method can have an arbitrary number of parameters describing the dependencies required to build that bean. For instance if our TransferServicerequires an AccountRepository we can materialize that dependency via a method parameter:

@Configuration
public class AppConfig {

    @Bean
    public TransferService transferService(AccountRepository accountRepository) {
        return new TransferServiceImpl(accountRepository);
    }

}

The resolution mechanism is pretty much identical to constructor-based dependency injection, see the relevant section for more details.

Receiving lifecycle callbacks

Any classes defined with the @Bean annotation support the regular lifecycle callbacks and can use the @PostConstruct and @PreDestroy annotations from JSR-250, see JSR-250 annotations for further details.

The regular Spring lifecycle callbacks are fully supported as well. If a bean implements InitializingBean, DisposableBean, or Lifecycle, their respective methods are called by the container.

The standard set of *Aware interfaces such as BeanFactoryAware, BeanNameAware, MessageSourceAware, ApplicationContextAware, and so on are also fully supported.

The @Bean annotation supports specifying arbitrary initialization and destruction callback methods, much like Spring XML’s init-method and destroy-methodattributes on the bean element:

public class Foo {
    public void init() {
        // initialization logic
    }
}

public class Bar {
    public void cleanup() {
        // destruction logic
    }
}

@Configuration
public class AppConfig {

    @Bean(initMethod = "init")
    public Foo foo() {
        return new Foo();
    }

    @Bean(destroyMethod = "cleanup")
    public Bar bar() {
        return new Bar();
    }

}
Spring框架参考文档 Spring Framework Reference Documentation

By default, beans defined using Java config that have a public close or shutdown method are automatically enlisted with a destruction callback. If you have a public close or shutdown method and you do not wish for it to be called when the container shuts down, simply add@Bean(destroyMethod="") to your bean definition to disable the default (inferred) mode.

You may want to do that by default for a resource that you acquire via JNDI as its lifecycle is managed outside the application. In particular, make sure to always do it for a DataSource as it is known to be problematic on Java EE application servers.

@Bean(destroyMethod="")
public DataSource dataSource() throws NamingException {
    return (DataSource) jndiTemplate.lookup("MyDS");
}

Also, with @Bean methods, you will typically choose to use programmatic JNDI lookups: either using Spring’s JndiTemplate/JndiLocatorDelegatehelpers or straight JNDI InitialContext usage, but not the JndiObjectFactoryBean variant which would force you to declare the return type as theFactoryBean type instead of the actual target type, making it harder to use for cross-reference calls in other @Bean methods that intend to refer to the provided resource here.

Of course, in the case of Foo above, it would be equally as valid to call the init() method directly during construction:

@Configuration
public class AppConfig {
    @Bean
    public Foo foo() {
        Foo foo = new Foo();
        foo.init();
        return foo;
    }

    // ...

}
Spring框架参考文档 Spring Framework Reference Documentation

When you work directly in Java, you can do anything you like with your objects and do not always need to rely on the container lifecycle!

Specifying bean scope

Using the @Scope annotation

You can specify that your beans defined with the @Bean annotation should have a specific scope. You can use any of the standard scopes specified in the Bean Scopessection.

The default scope is singleton, but you can override this with the @Scope annotation:

@Configuration
public class MyConfiguration {

    @Bean
    @Scope("prototype")
    public Encryptor encryptor() {
        // ...
    }

}
@Scope and scoped-proxy

Spring offers a convenient way of working with scoped dependencies through scoped proxies. The easiest way to create such a proxy when using the XML configuration is the <aop:scoped-proxy/> element. Configuring your beans in Java with a @Scope annotation offers equivalent support with the proxyMode attribute. The default is no proxy ( ScopedProxyMode.NO), but you can specify ScopedProxyMode.TARGET_CLASS or ScopedProxyMode.INTERFACES.

If you port the scoped proxy example from the XML reference documentation (see preceding link) to our @Bean using Java, it would look like the following:

// an HTTP Session-scoped bean exposed as a proxy
@Bean
@SessionScope
public UserPreferences userPreferences() {
    return new UserPreferences();
}

@Bean
public Service userService() {
    UserService service = new SimpleUserService();
    // a reference to the proxied userPreferences bean
    service.setUserPreferences(userPreferences());
    return service;
}

Customizing bean naming

By default, configuration classes use a @Bean method’s name as the name of the resulting bean. This functionality can be overridden, however, with the name attribute.

@Configuration
public class AppConfig {

    @Bean(name = "myFoo")
    public Foo foo() {
        return new Foo();
    }

}

Bean aliasing

As discussed in Section 7.3.1, “Naming beans”, it is sometimes desirable to give a single bean multiple names, otherwise known as bean aliasing. The name attribute of the @Bean annotation accepts a String array for this purpose.

@Configuration
public class AppConfig {

    @Bean(name = { "dataSource", "subsystemA-dataSource", "subsystemB-dataSource" })
    public DataSource dataSource() {
        // instantiate, configure and return DataSource bean...
    }

}

Bean description

Sometimes it is helpful to provide a more detailed textual description of a bean. This can be particularly useful when beans are exposed (perhaps via JMX) for monitoring purposes.

To add a description to a @Bean the @Description annotation can be used:

@Configuration
public class AppConfig {

    @Bean
    @Description("Provides a basic example of a bean")
    public Foo foo() {
        return new Foo();
    }

}

7.12.4 Using the @Configuration annotation

@Configuration is a class-level annotation indicating that an object is a source of bean definitions. @Configuration classes declare beans via public @Beanannotated methods. Calls to @Bean methods on @Configuration classes can also be used to define inter-bean dependencies. See Section 7.12.1, “Basic concepts: @Bean and @Configuration” for a general introduction.

Injecting inter-bean dependencies

When @Beans have dependencies on one another, expressing that dependency is as simple as having one bean method call another:

@Configuration
public class AppConfig {

    @Bean
    public Foo foo() {
        return new Foo(bar());
    }

    @Bean
    public Bar bar() {
        return new Bar();
    }

}

In the example above, the foo bean receives a reference to bar via constructor injection.

Spring框架参考文档 Spring Framework Reference Documentation

This method of declaring inter-bean dependencies only works when the @Bean method is declared within a @Configuration class. You cannot declare inter-bean dependencies using plain @Component classes.

Lookup method injection

As noted earlier, lookup method injection is an advanced feature that you should use rarely. It is useful in cases where a singleton-scoped bean has a dependency on a prototype-scoped bean. Using Java for this type of configuration provides a natural means for implementing this pattern.

public abstract class CommandManager {
    public Object process(Object commandState) {
        // grab a new instance of the appropriate Command interface
        Command command = createCommand();

        // set the state on the (hopefully brand new) Command instance
        command.setState(commandState);
    return command.execute();
    }

    // okay... but where is the implementation of this method?
    protected abstract Command createCommand();
}

Using Java-configuration support , you can create a subclass of CommandManager where the abstract createCommand() method is overridden in such a way that it looks up a new (prototype) command object:

@Bean
@Scope("prototype")
public AsyncCommand asyncCommand() {
    AsyncCommand command = new AsyncCommand();
    // inject dependencies here as required
    return command;
}

@Bean
public CommandManager commandManager() {
    // return new anonymous implementation of CommandManager with command() overridden
    // to return a new prototype Command object
    return new CommandManager() {
        protected Command createCommand() {
            return asyncCommand();
        }
    }
}

Further information about how Java-based configuration works internally

The following example shows a @Bean annotated method being called twice:

@Configuration
public class AppConfig {

    @Bean
    public ClientService clientService1() {
        ClientServiceImpl clientService = new ClientServiceImpl();
        clientService.setClientDao(clientDao());
        return clientService;
    }

    @Bean
    public ClientService clientService2() {
        ClientServiceImpl clientService = new ClientServiceImpl();
        clientService.setClientDao(clientDao());
        return clientService;
    }

    @Bean
    public ClientDao clientDao() {
        return new ClientDaoImpl();
    }

}

clientDao() has been called once in clientService1() and once in clientService2(). Since this method creates a new instance of ClientDaoImpl and returns it, you would normally expect having 2 instances (one for each service). That definitely would be problematic: in Spring, instantiated beans have a singletonscope by default. This is where the magic comes in: All @Configuration classes are subclassed at startup-time with CGLIB. In the subclass, the child method checks the container first for any cached (scoped) beans before it calls the parent method and creates a new instance. Note that as of Spring 3.2, it is no longer necessary to add CGLIB to your classpath because CGLIB classes have been repackaged under org.springframework.cglib and included directly within the spring-core JAR.

Spring框架参考文档 Spring Framework Reference Documentation

The behavior could be different according to the scope of your bean. We are talking about singletons here.
Spring框架参考文档 Spring Framework Reference Documentation

There are a few restrictions due to the fact that CGLIB dynamically adds features at startup-time:

  • Configuration classes should not be final
  • They should have a constructor with no arguments

7.12.5 Composing Java-based configurations

Using the @Import annotation

Much as the <import/> element is used within Spring XML files to aid in modularizing configurations, the @Import annotation allows for loading @Bean definitions from another configuration class:

@Configuration
public class ConfigA {

     @Bean
    public A a() {
        return new A();
    }

}

@Configuration
@Import(ConfigA.class)
public class ConfigB {

    @Bean
    public B b() {
        return new B();
    }

}

Now, rather than needing to specify both ConfigA.class and ConfigB.class when instantiating the context, only ConfigB needs to be supplied explicitly:

public static void main(String[] args) {
    ApplicationContext ctx = new AnnotationConfigApplicationContext(ConfigB.class);

    // now both beans A and B will be available...
    A a = ctx.getBean(A.class);
    B b = ctx.getBean(B.class);
}

This approach simplifies container instantiation, as only one class needs to be dealt with, rather than requiring the developer to remember a potentially large number of@Configuration classes during construction.

Injecting dependencies on imported @Bean definitions

The example above works, but is simplistic. In most practical scenarios, beans will have dependencies on one another across configuration classes. When using XML, this is not an issue, per se, because there is no compiler involved, and one can simply declare ref="someBean" and trust that Spring will work it out during container initialization. Of course, when using @Configuration classes, the Java compiler places constraints on the configuration model, in that references to other beans must be valid Java syntax.

Fortunately, solving this problem is simple. As we already discussed, @Bean method can have an arbitrary number of parameters describing the bean dependencies. Let’s consider a more real-world scenario with several @Configuration classes, each depending on beans declared in the others:

@Configuration
public class ServiceConfig {

    @Bean
    public TransferService transferService(AccountRepository accountRepository) {
        return new TransferServiceImpl(accountRepository);
    }

}

@Configuration
public class RepositoryConfig {

    @Bean
    public AccountRepository accountRepository(DataSource dataSource) {
        return new JdbcAccountRepository(dataSource);
    }

}

@Configuration
@Import({ServiceConfig.class, RepositoryConfig.class})
public class SystemTestConfig {

    @Bean
    public DataSource dataSource() {
        // return new DataSource
    }

}

public static void main(String[] args) {
    ApplicationContext ctx = new AnnotationConfigApplicationContext(SystemTestConfig.class);
    // everything wires up across configuration classes...
    TransferService transferService = ctx.getBean(TransferService.class);
    transferService.transfer(100.00, "A123", "C456");
}

There is another way to achieve the same result. Remember that @Configuration classes are ultimately just another bean in the container: This means that they can take advantage of @Autowired and @Value injection etc just like any other bean!

Spring框架参考文档 Spring Framework Reference Documentation

Make sure that the dependencies you inject that way are of the simplest kind only. @Configuration classes are processed quite early during the initialization of the context and forcing a dependency to be injected this way may lead to unexpected early initialization. Whenever possible, resort to parameter-based injection as in the example above.

Also, be particularly careful with BeanPostProcessor and BeanFactoryPostProcessor definitions via @Bean. Those should usually be declared asstatic @Bean methods, not triggering the instantiation of their containing configuration class. Otherwise, @Autowired and @Value won’t work on the configuration class itself since it is being created as a bean instance too early.

 
@Configuration
public class ServiceConfig {

    @Autowired
    private AccountRepository accountRepository;

    @Bean
    public TransferService transferService() {
        return new TransferServiceImpl(accountRepository);
    }

}

@Configuration
public class RepositoryConfig {

    private final DataSource dataSource;

    @Autowired
    public RepositoryConfig(DataSource dataSource) {
        this.dataSource = dataSource;
    }

    @Bean
    public AccountRepository accountRepository() {
        return new JdbcAccountRepository(dataSource);
    }

}

@Configuration
@Import({ServiceConfig.class, RepositoryConfig.class})
public class SystemTestConfig {

    @Bean
    public DataSource dataSource() {
        // return new DataSource
    }

}

public static void main(String[] args) {
    ApplicationContext ctx = new AnnotationConfigApplicationContext(SystemTestConfig.class);
    // everything wires up across configuration classes...
    TransferService transferService = ctx.getBean(TransferService.class);
    transferService.transfer(100.00, "A123", "C456");
}
Spring框架参考文档 Spring Framework Reference Documentation

Constructor injection in @Configuration classes is only supported as of Spring Framework 4.3. Note also that there is no need to specify @Autowired if the target bean defines only one constructor; in the example above, @Autowired is not necessary on the RepositoryConfig constructor.

In the scenario above, using @Autowired works well and provides the desired modularity, but determining exactly where the autowired bean definitions are declared is still somewhat ambiguous. For example, as a developer looking at ServiceConfig, how do you know exactly where the @Autowired AccountRepository bean is declared? It’s not explicit in the code, and this may be just fine. Remember that the Spring Tool Suite provides tooling that can render graphs showing how everything is wired up - that may be all you need. Also, your Java IDE can easily find all declarations and uses of the AccountRepository type, and will quickly show you the location of @Bean methods that return that type.

In cases where this ambiguity is not acceptable and you wish to have direct navigation from within your IDE from one @Configuration class to another, consider autowiring the configuration classes themselves:

@Configuration
public class ServiceConfig {

    @Autowired
    private RepositoryConfig repositoryConfig;

    @Bean
    public TransferService transferService() {
        // navigate 'through' the config class to the @Bean method!
        return new TransferServiceImpl(repositoryConfig.accountRepository());
    }

}

In the situation above, it is completely explicit where AccountRepository is defined. However, ServiceConfig is now tightly coupled to RepositoryConfig; that’s the tradeoff. This tight coupling can be somewhat mitigated by using interface-based or abstract class-based @Configuration classes. Consider the following:

@Configuration
public class ServiceConfig {

    @Autowired
    private RepositoryConfig repositoryConfig;

    @Bean
    public TransferService transferService() {
        return new TransferServiceImpl(repositoryConfig.accountRepository());
    }
}

@Configuration
public interface RepositoryConfig {

    @Bean
    AccountRepository accountRepository();

}

@Configuration
public class DefaultRepositoryConfig implements RepositoryConfig {

    @Bean
    public AccountRepository accountRepository() {
        return new JdbcAccountRepository(...);
    }

}

@Configuration
@Import({ServiceConfig.class, DefaultRepositoryConfig.class}) // import the concrete config!
public class SystemTestConfig {

    @Bean
    public DataSource dataSource() {
        // return DataSource
    }

}

public static void main(String[] args) {
    ApplicationContext ctx = new AnnotationConfigApplicationContext(SystemTestConfig.class);
    TransferService transferService = ctx.getBean(TransferService.class);
    transferService.transfer(100.00, "A123", "C456");
}

Now ServiceConfig is loosely coupled with respect to the concrete DefaultRepositoryConfig, and built-in IDE tooling is still useful: it will be easy for the developer to get a type hierarchy of RepositoryConfig implementations. In this way, navigating @Configuration classes and their dependencies becomes no different than the usual process of navigating interface-based code.

Conditionally include @Configuration classes or @Bean methods

It is often useful to conditionally enable or disable a complete @Configuration class, or even individual @Bean methods, based on some arbitrary system state. One common example of this is to use the @Profile annotation to activate beans only when a specific profile has been enabled in the Spring Environment (seeSection 7.13.1, “Bean definition profiles” for details).

The @Profile annotation is actually implemented using a much more flexible annotation called @Conditional. The @Conditional annotation indicates specificorg.springframework.context.annotation.Condition implementations that should be consulted before a @Bean is registered.

Implementations of the Condition interface simply provide a matches(…​) method that returns true or false. For example, here is the actual Conditionimplementation used for @Profile:

@Override
public boolean matches(ConditionContext context, AnnotatedTypeMetadata metadata) {
    if (context.getEnvironment() != null) {
        // Read the @Profile annotation attributes
        MultiValueMap<String, Object> attrs = metadata.getAllAnnotationAttributes(Profile.class.getName());
        if (attrs != null) {
            for (Object value : attrs.get("value")) {
                if (context.getEnvironment().acceptsProfiles(((String[]) value))) {
                    return true;
                }
            }
            return false;
        }
    }
    return true;
}

See the @Conditional javadocs for more detail.

Combining Java and XML configuration

Spring’s @Configuration class support does not aim to be a 100% complete replacement for Spring XML. Some facilities such as Spring XML namespaces remain an ideal way to configure the container. In cases where XML is convenient or necessary, you have a choice: either instantiate the container in an "XML-centric" way using, for example, ClassPathXmlApplicationContext, or in a "Java-centric" fashion using AnnotationConfigApplicationContext and the @ImportResourceannotation to import XML as needed.

XML-centric use of @Configuration classes

It may be preferable to bootstrap the Spring container from XML and include @Configuration classes in an ad-hoc fashion. For example, in a large existing codebase that uses Spring XML, it will be easier to create @Configuration classes on an as-needed basis and include them from the existing XML files. Below you’ll find the options for using @Configuration classes in this kind of "XML-centric" situation.

Remember that @Configuration classes are ultimately just bean definitions in the container. In this example, we create a @Configuration class named AppConfigand include it within system-test-config.xml as a <bean/> definition. Because <context:annotation-config/> is switched on, the container will recognize the@Configuration annotation and process the @Bean methods declared in AppConfig properly.

@Configuration
public class AppConfig {

    @Autowired
    private DataSource dataSource;

    @Bean
    public AccountRepository accountRepository() {
        return new JdbcAccountRepository(dataSource);
    }

    @Bean
    public TransferService transferService() {
        return new TransferService(accountRepository());
    }

}

system-test-config.xml:

<beans>
    <!-- enable processing of annotations such as @Autowired and @Configuration -->
    <context:annotation-config/>
    <context:property-placeholder location="classpath:/com/acme/jdbc.properties"/>

    <bean class="com.acme.AppConfig"/>

    <bean class="org.springframework.jdbc.datasource.DriverManagerDataSource">
        <property name="url" value="${jdbc.url}"/>
        <property name="username" value="${jdbc.username}"/>
        <property name="password" value="${jdbc.password}"/>
    </bean>
</beans>

jdbc.properties:

jdbc.url=jdbc:hsqldb:hsql://localhost/xdb
jdbc.username=sa
jdbc.password=
public static void main(String[] args) {
    ApplicationContext ctx = new ClassPathXmlApplicationContext("classpath:/com/acme/system-test-config.xml");
    TransferService transferService = ctx.getBean(TransferService.class);
    // ...
}
Spring框架参考文档 Spring Framework Reference Documentation

In system-test-config.xml above, the AppConfig <bean/> does not declare an id element. While it would be acceptable to do so, it is unnecessary given that no other bean will ever refer to it, and it is unlikely that it will be explicitly fetched from the container by name. Likewise with the DataSourcebean - it is only ever autowired by type, so an explicit bean id is not strictly required.

Because @Configuration is meta-annotated with @Component, @Configuration-annotated classes are automatically candidates for component scanning. Using the same scenario as above, we can redefine system-test-config.xml to take advantage of component-scanning. Note that in this case, we don’t need to explicitly declare <context:annotation-config/>, because <context:component-scan/> enables the same functionality.

system-test-config.xml:

<beans>
    <!-- picks up and registers AppConfig as a bean definition -->
    <context:component-scan base-package="com.acme"/>
    <context:property-placeholder location="classpath:/com/acme/jdbc.properties"/>

    <bean class="org.springframework.jdbc.datasource.DriverManagerDataSource">
        <property name="url" value="${jdbc.url}"/>
        <property name="username" value="${jdbc.username}"/>
        <property name="password" value="${jdbc.password}"/>
    </bean>
</beans>
@Configuration class-centric use of XML with @ImportResource

In applications where @Configuration classes are the primary mechanism for configuring the container, it will still likely be necessary to use at least some XML. In these scenarios, simply use @ImportResource and define only as much XML as is needed. Doing so achieves a "Java-centric" approach to configuring the container and keeps XML to a bare minimum.

@Configuration
@ImportResource("classpath:/com/acme/properties-config.xml")
public class AppConfig {

    @Value("${jdbc.url}")
    private String url;

    @Value("${jdbc.username}")
    private String username;

    @Value("${jdbc.password}")
    private String password;

    @Bean
    public DataSource dataSource() {
        return new DriverManagerDataSource(url, username, password);
    }

}
properties-config.xml
<beans>
    <context:property-placeholder location="classpath:/com/acme/jdbc.properties"/>
</beans>
jdbc.properties
jdbc.url=jdbc:hsqldb:hsql://localhost/xdb
jdbc.username=sa
jdbc.password=
public static void main(String[] args) {
    ApplicationContext ctx = new AnnotationConfigApplicationContext(AppConfig.class);
    TransferService transferService = ctx.getBean(TransferService.class);
    // ...
}

7.13 Environment abstraction

The Environment is an abstraction integrated in the container that models two key aspects of the application environment: profiles and properties.

A profile is a named, logical group of bean definitions to be registered with the container only if the given profile is active. Beans may be assigned to a profile whether defined in XML or via annotations. The role of the Environment object with relation to profiles is in determining which profiles (if any) are currently active, and which profiles (if any) should be active by default.

Properties play an important role in almost all applications, and may originate from a variety of sources: properties files, JVM system properties, system environment variables, JNDI, servlet context parameters, ad-hoc Properties objects, Maps, and so on. The role of the Environment object with relation to properties is to provide the user with a convenient service interface for configuring property sources and resolving properties from them.

7.13.1 Bean definition profiles

Bean definition profiles is a mechanism in the core container that allows for registration of different beans in different environments. The word environment can mean different things to different users and this feature can help with many use cases, including:

  • working against an in-memory datasource in development vs looking up that same datasource from JNDI when in QA or production
  • registering monitoring infrastructure only when deploying an application into a performance environment
  • registering customized implementations of beans for customer A vs. customer B deployments

Let’s consider the first use case in a practical application that requires a DataSource. In a test environment, the configuration may look like this:

@Bean
public DataSource dataSource() {
    return new EmbeddedDatabaseBuilder()
        .setType(EmbeddedDatabaseType.HSQL)
        .addScript("my-schema.sql")
        .addScript("my-test-data.sql")
        .build();
}

Let’s now consider how this application will be deployed into a QA or production environment, assuming that the datasource for the application will be registered with the production application server’s JNDI directory. Our dataSource bean now looks like this:

@Bean(destroyMethod="")
public DataSource dataSource() throws Exception {
    Context ctx = new InitialContext();
    return (DataSource) ctx.lookup("java:comp/env/jdbc/datasource");
}

The problem is how to switch between using these two variations based on the current environment. Over time, Spring users have devised a number of ways to get this done, usually relying on a combination of system environment variables and XML <import/> statements containing ${placeholder} tokens that resolve to the correct configuration file path depending on the value of an environment variable. Bean definition profiles is a core container feature that provides a solution to this problem.

If we generalize the example use case above of environment-specific bean definitions, we end up with the need to register certain bean definitions in certain contexts, while not in others. You could say that you want to register a certain profile of bean definitions in situation A, and a different profile in situation B. Let’s first see how we can update our configuration to reflect this need.

@Profile

The @Profile annotation allows you to indicate that a component is eligible for registration when one or more specified profiles are active. Using our example above, we can rewrite the dataSource configuration as follows:

@Configuration
@Profile("dev")
public class StandaloneDataConfig {

    @Bean
    public DataSource dataSource() {
        return new EmbeddedDatabaseBuilder()
            .setType(EmbeddedDatabaseType.HSQL)
            .addScript("classpath:com/bank/config/sql/schema.sql")
            .addScript("classpath:com/bank/config/sql/test-data.sql")
            .build();
    }
}
@Configuration
@Profile("production")
public class JndiDataConfig {

    @Bean(destroyMethod="")
    public DataSource dataSource() throws Exception {
        Context ctx = new InitialContext();
        return (DataSource) ctx.lookup("java:comp/env/jdbc/datasource");
    }
}
Spring框架参考文档 Spring Framework Reference Documentation

As mentioned before, with @Bean methods, you will typically choose to use programmatic JNDI lookups: either using Spring’sJndiTemplate/JndiLocatorDelegate helpers or the straight JNDI InitialContext usage shown above, but not the JndiObjectFactoryBeanvariant which would force you to declare the return type as the FactoryBean type.

@Profile can be used as a meta-annotation for the purpose of creating a custom composed annotation. The following example defines a custom @Productionannotation that can be used as a drop-in replacement for @Profile("production"):

@Target(ElementType.TYPE)
@Retention(RetentionPolicy.RUNTIME)
@Profile("production")
public @interface Production {
}

@Profile can also be declared at the method level to include only one particular bean of a configuration class:

@Configuration
public class AppConfig {

    @Bean
    @Profile("dev")
    public DataSource devDataSource() {
        return new EmbeddedDatabaseBuilder()
            .setType(EmbeddedDatabaseType.HSQL)
            .addScript("classpath:com/bank/config/sql/schema.sql")
            .addScript("classpath:com/bank/config/sql/test-data.sql")
            .build();
    }

    @Bean
    @Profile("production")
    public DataSource productionDataSource() throws Exception {
        Context ctx = new InitialContext();
        return (DataSource) ctx.lookup("java:comp/env/jdbc/datasource");
    }
}
Spring框架参考文档 Spring Framework Reference Documentation

If a @Configuration class is marked with @Profile, all of the @Bean methods and @Import annotations associated with that class will be bypassed unless one or more of the specified profiles are active. If a @Component or @Configuration class is marked with @Profile({"p1", "p2"}), that class will not be registered/processed unless profiles 'p1' and/or 'p2' have been activated. If a given profile is prefixed with the NOT operator (!), the annotated element will be registered if the profile is not active. For example, given @Profile({"p1", "!p2"}), registration will occur if profile 'p1' is active or if profile 'p2' is not active.

7.13.2 XML bean definition profiles

The XML counterpart is the profile attribute of the <beans> element. Our sample configuration above can be rewritten in two XML files as follows:

<beans profile="dev"
    xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:jdbc="http://www.springframework.org/schema/jdbc"
    xsi:schemaLocation="...">

    <jdbc:embedded-database id="dataSource">
        <jdbc:script location="classpath:com/bank/config/sql/schema.sql"/>
        <jdbc:script location="classpath:com/bank/config/sql/test-data.sql"/>
    </jdbc:embedded-database>
</beans>
<beans profile="production"
    xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:jee="http://www.springframework.org/schema/jee"
    xsi:schemaLocation="...">

    <jee:jndi-lookup id="dataSource" jndi-name="java:comp/env/jdbc/datasource"/>
</beans>

It is also possible to avoid that split and nest <beans/> elements within the same file:

<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:jdbc="http://www.springframework.org/schema/jdbc"
    xmlns:jee="http://www.springframework.org/schema/jee"
    xsi:schemaLocation="...">

    <!-- other bean definitions -->

    <beans profile="dev">
        <jdbc:embedded-database id="dataSource">
            <jdbc:script location="classpath:com/bank/config/sql/schema.sql"/>
            <jdbc:script location="classpath:com/bank/config/sql/test-data.sql"/>
        </jdbc:embedded-database>
    </beans>

    <beans profile="production">
        <jee:jndi-lookup id="dataSource" jndi-name="java:comp/env/jdbc/datasource"/>
    </beans>
</beans>

The spring-bean.xsd has been constrained to allow such elements only as the last ones in the file. This should help provide flexibility without incurring clutter in the XML files.

Activating a profile

Now that we have updated our configuration, we still need to instruct Spring which profile is active. If we started our sample application right now, we would see aNoSuchBeanDefinitionException thrown, because the container could not find the Spring bean named dataSource.

Activating a profile can be done in several ways, but the most straightforward is to do it programmatically against the Environment API which is available via anApplicationContext:

AnnotationConfigApplicationContext ctx = new AnnotationConfigApplicationContext();
ctx.getEnvironment().setActiveProfiles("dev");
ctx.register(SomeConfig.class, StandaloneDataConfig.class, JndiDataConfig.class);
ctx.refresh();

In addition, profiles may also be activated declaratively through the spring.profiles.active property which may be specified through system environment variables, JVM system properties, servlet context parameters in web.xml, or even as an entry in JNDI (see Section 7.13.3, “PropertySource abstraction”). In integration tests, active profiles can be declared via the @ActiveProfiles annotation in the spring-test module (see the section called “Context configuration with environment profiles”).

Note that profiles are not an "either-or" proposition; it is possible to activate multiple profiles at once. Programmatically, simply provide multiple profile names to thesetActiveProfiles() method, which accepts String…​ varargs:

ctx.getEnvironment().setActiveProfiles("profile1", "profile2");

Declaratively, spring.profiles.active may accept a comma-separated list of profile names:

-Dspring.profiles.active="profile1,profile2"

Default profile

The default profile represents the profile that is enabled by default. Consider the following:

@Configuration
@Profile("default")
public class DefaultDataConfig {

    @Bean
    public DataSource dataSource() {
        return new EmbeddedDatabaseBuilder()
            .setType(EmbeddedDatabaseType.HSQL)
            .addScript("classpath:com/bank/config/sql/schema.sql")
            .build();
    }
}

If no profile is active, the dataSource above will be created; this can be seen as a way to provide a default definition for one or more beans. If any profile is enabled, thedefault profile will not apply.

The name of the default profile can be changed using setDefaultProfiles() on the Environment or declaratively using the spring.profiles.default property.

7.13.3 PropertySource abstraction

Spring’s Environment abstraction provides search operations over a configurable hierarchy of property sources. To explain fully, consider the following:

ApplicationContext ctx = new GenericApplicationContext();
Environment env = ctx.getEnvironment();
boolean containsFoo = env.containsProperty("foo");
System.out.println("Does my environment contain the 'foo' property? " + containsFoo);

In the snippet above, we see a high-level way of asking Spring whether the foo property is defined for the current environment. To answer this question, theEnvironment object performs a search over a set of PropertySource objects. A PropertySource is a simple abstraction over any source of key-value pairs, and Spring’s StandardEnvironment is configured with two PropertySource objects — one representing the set of JVM system properties (a la System.getProperties()) and one representing the set of system environment variables (a la System.getenv()).

Spring框架参考文档 Spring Framework Reference Documentation

These default property sources are present for StandardEnvironment, for use in standalone applications. StandardServletEnvironment is populated with additional default property sources including servlet config and servlet context parameters. StandardPortletEnvironment similarly has access to portlet config and portlet context parameters as property sources. Both can optionally enable a JndiPropertySource. See the javadocs for details.

Concretely, when using the StandardEnvironment, the call to env.containsProperty("foo") will return true if a foo system property or foo environment variable is present at runtime.

Spring框架参考文档 Spring Framework Reference Documentation

The search performed is hierarchical. By default, system properties have precedence over environment variables, so if the foo property happens to be set in both places during a call to env.getProperty("foo"), the system property value will 'win' and be returned preferentially over the environment variable. Note that property values will not get merged but rather completely overridden by a preceding entry.

For a common StandardServletEnvironment, the full hierarchy looks as follows, with the highest-precedence entries at the top: * ServletConfig parameters (if applicable, e.g. in case of a DispatcherServlet context) * ServletContext parameters (web.xml context-param entries) * JNDI environment variables ("java:comp/env/" entries) * JVM system properties ("-D" command-line arguments) * JVM system environment (operating system environment variables)

Most importantly, the entire mechanism is configurable. Perhaps you have a custom source of properties that you’d like to integrate into this search. No problem — simply implement and instantiate your own PropertySource and add it to the set of PropertySources for the current Environment:

ConfigurableApplicationContext ctx = new GenericApplicationContext();
MutablePropertySources sources = ctx.getEnvironment().getPropertySources();
sources.addFirst(new MyPropertySource());

In the code above, MyPropertySource has been added with highest precedence in the search. If it contains a foo property, it will be detected and returned ahead of any foo property in any other PropertySource. The MutablePropertySources API exposes a number of methods that allow for precise manipulation of the set of property sources.

7.13.4 @PropertySource

The @PropertySource annotation provides a convenient and declarative mechanism for adding a PropertySource to Spring’s Environment.

Given a file "app.properties" containing the key/value pair testbean.name=myTestBean, the following @Configuration class uses @PropertySource in such a way that a call to testBean.getName() will return "myTestBean".

@Configuration
@PropertySource("classpath:/com/myco/app.properties")
public class AppConfig {
    @Autowired
    Environment env;

    @Bean
    public TestBean testBean() {
        TestBean testBean = new TestBean();
        testBean.setName(env.getProperty("testbean.name"));
        return testBean;
    }
}

Any ${…​} placeholders present in a @PropertySource resource location will be resolved against the set of property sources already registered against the environment. For example:

@Configuration
@PropertySource("classpath:/com/${my.placeholder:default/path}/app.properties")
public class AppConfig {
    @Autowired
    Environment env;

    @Bean
    public TestBean testBean() {
        TestBean testBean = new TestBean();
        testBean.setName(env.getProperty("testbean.name"));
        return testBean;
    }
}

Assuming that "my.placeholder" is present in one of the property sources already registered, e.g. system properties or environment variables, the placeholder will be resolved to the corresponding value. If not, then "default/path" will be used as a default. If no default is specified and a property cannot be resolved, anIllegalArgumentException will be thrown.

7.13.5 Placeholder resolution in statements

Historically, the value of placeholders in elements could be resolved only against JVM system properties or environment variables. No longer is this the case. Because the Environment abstraction is integrated throughout the container, it’s easy to route resolution of placeholders through it. This means that you may configure the resolution process in any way you like: change the precedence of searching through system properties and environment variables, or remove them entirely; add your own property sources to the mix as appropriate.

Concretely, the following statement works regardless of where the customer property is defined, as long as it is available in the Environment:

<beans>
    <import resource="com/bank/service/${customer}-config.xml"/>
</beans>

7.14 Registering a LoadTimeWeaver

The LoadTimeWeaver is used by Spring to dynamically transform classes as they are loaded into the Java virtual machine (JVM).

To enable load-time weaving add the @EnableLoadTimeWeaving to one of your @Configuration classes:

@Configuration
@EnableLoadTimeWeaving
public class AppConfig {

}

Alternatively for XML configuration use the context:load-time-weaver element:

<beans>
    <context:load-time-weaver/>
</beans>

Once configured for the ApplicationContext. Any bean within that ApplicationContext may implement LoadTimeWeaverAware, thereby receiving a reference to the load-time weaver instance. This is particularly useful in combination with Spring’s JPA support where load-time weaving may be necessary for JPA class transformation. Consult the LocalContainerEntityManagerFactoryBean javadocs for more detail. For more on AspectJ load-time weaving, see Section 11.8.4, “Load-time weaving with AspectJ in the Spring Framework”.

7.15 Additional Capabilities of the ApplicationContext

As was discussed in the chapter introduction, the org.springframework.beans.factory package provides basic functionality for managing and manipulating beans, including in a programmatic way. The org.springframework.context package adds the ApplicationContext interface, which extends the BeanFactory interface, in addition to extending other interfaces to provide additional functionality in a more application framework-oriented style. Many people use the ApplicationContext in a completely declarative fashion, not even creating it programmatically, but instead relying on support classes such as ContextLoader to automatically instantiate anApplicationContext as part of the normal startup process of a Java EE web application.

To enhance BeanFactory functionality in a more framework-oriented style the context package also provides the following functionality:

  • Access to messages in i18n-style, through the MessageSource interface.
  • Access to resources, such as URLs and files, through the ResourceLoader interface.
  • Event publication to namely beans implementing the ApplicationListener interface, through the use of the ApplicationEventPublisher interface.
  • Loading of multiple (hierarchical) contexts, allowing each to be focused on one particular layer, such as the web layer of an application, through theHierarchicalBeanFactory interface.

7.15.1 Internationalization using MessageSource

The ApplicationContext interface extends an interface called MessageSource, and therefore provides internationalization (i18n) functionality. Spring also provides the interface HierarchicalMessageSource, which can resolve messages hierarchically. Together these interfaces provide the foundation upon which Spring effects message resolution. The methods defined on these interfaces include:

  • String getMessage(String code, Object[] args, String default, Locale loc): The basic method used to retrieve a message from theMessageSource. When no message is found for the specified locale, the default message is used. Any arguments passed in become replacement values, using theMessageFormat functionality provided by the standard library.
  • String getMessage(String code, Object[] args, Locale loc): Essentially the same as the previous method, but with one difference: no default message can be specified; if the message cannot be found, a NoSuchMessageException is thrown.
  • String getMessage(MessageSourceResolvable resolvable, Locale locale): All properties used in the preceding methods are also wrapped in a class named MessageSourceResolvable, which you can use with this method.

When an ApplicationContext is loaded, it automatically searches for a MessageSource bean defined in the context. The bean must have the namemessageSource. If such a bean is found, all calls to the preceding methods are delegated to the message source. If no message source is found, theApplicationContext attempts to find a parent containing a bean with the same name. If it does, it uses that bean as the MessageSource. If theApplicationContext cannot find any source for messages, an empty DelegatingMessageSource is instantiated in order to be able to accept calls to the methods defined above.

Spring provides two MessageSource implementations, ResourceBundleMessageSource and StaticMessageSource. Both implementHierarchicalMessageSource in order to do nested messaging. The StaticMessageSource is rarely used but provides programmatic ways to add messages to the source. The ResourceBundleMessageSource is shown in the following example:

<beans>
    <bean id="messageSource"
            class="org.springframework.context.support.ResourceBundleMessageSource">
        <property name="basenames">
            <list>
                <value>format</value>
                <value>exceptions</value>
                <value>windows</value>
            </list>
        </property>
    </bean>
</beans>

In the example it is assumed you have three resource bundles defined in your classpath called format, exceptions and windows. Any request to resolve a message will be handled in the JDK standard way of resolving messages through ResourceBundles. For the purposes of the example, assume the contents of two of the above resource bundle files are…​

# in format.properties
message=Alligators rock!
# in exceptions.properties
argument.required=The {0} argument is required.

A program to execute the MessageSource functionality is shown in the next example. Remember that all ApplicationContext implementations are alsoMessageSource implementations and so can be cast to the MessageSource interface.

public static void main(String[] args) {
    MessageSource resources = new ClassPathXmlApplicationContext("beans.xml");
    String message = resources.getMessage("message", null, "Default", null);
    System.out.println(message);
}

The resulting output from the above program will be…​

Alligators rock!

So to summarize, the MessageSource is defined in a file called beans.xml, which exists at the root of your classpath. The messageSource bean definition refers to a number of resource bundles through its basenames property. The three files that are passed in the list to the basenames property exist as files at the root of your classpath and are called format.properties, exceptions.properties, and windows.properties respectively.

The next example shows arguments passed to the message lookup; these arguments will be converted into Strings and inserted into placeholders in the lookup message.

<beans>

    <!-- this MessageSource is being used in a web application -->
    <bean id="messageSource" class="org.springframework.context.support.ResourceBundleMessageSource">
        <property name="basename" value="exceptions"/>
    </bean>

    <!-- lets inject the above MessageSource into this POJO -->
    <bean id="example" class="com.foo.Example">
        <property name="messages" ref="messageSource"/>
    </bean>

</beans>
public class Example {

    private MessageSource messages;

    public void setMessages(MessageSource messages) {
        this.messages = messages;
    }

    public void execute() {
        String message = this.messages.getMessage("argument.required",
            new Object [] {"userDao"}, "Required", null);
        System.out.println(message);
    }

}

The resulting output from the invocation of the execute() method will be…​

The userDao argument is required.

With regard to internationalization (i18n), Spring’s various MessageSource implementations follow the same locale resolution and fallback rules as the standard JDKResourceBundle. In short, and continuing with the example messageSource defined previously, if you want to resolve messages against the British (en-GB) locale, you would create files called format_en_GB.properties, exceptions_en_GB.properties, and windows_en_GB.properties respectively.

Typically, locale resolution is managed by the surrounding environment of the application. In this example, the locale against which (British) messages will be resolved is specified manually.

# in exceptions_en_GB.properties
argument.required=Ebagum lad, the {0} argument is required, I say, required.
public static void main(final String[] args) {
    MessageSource resources = new ClassPathXmlApplicationContext("beans.xml");
    String message = resources.getMessage("argument.required",
        new Object [] {"userDao"}, "Required", Locale.UK);
    System.out.println(message);
}

The resulting output from the running of the above program will be…​

Ebagum lad, the 'userDao' argument is required, I say, required.

You can also use the MessageSourceAware interface to acquire a reference to any MessageSource that has been defined. Any bean that is defined in anApplicationContext that implements the MessageSourceAware interface is injected with the application context’s MessageSource when the bean is created and configured.

Spring框架参考文档 Spring Framework Reference Documentation

As an alternative to ResourceBundleMessageSource, Spring provides a ReloadableResourceBundleMessageSource class. This variant supports the same bundle file format but is more flexible than the standard JDK based ResourceBundleMessageSource implementation. In particular, it allows for reading files from any Spring resource location (not just from the classpath) and supports hot reloading of bundle property files (while efficiently caching them in between). Check out the ReloadableResourceBundleMessageSource javadocs for details.

7.15.2 Standard and Custom Events

Event handling in the ApplicationContext is provided through the ApplicationEvent class and ApplicationListener interface. If a bean that implements theApplicationListener interface is deployed into the context, every time an ApplicationEvent gets published to the ApplicationContext, that bean is notified. Essentially, this is the standard Observer design pattern.

Spring框架参考文档 Spring Framework Reference Documentation

As of Spring 4.2, the event infrastructure has been significantly improved and offer an annotation-based model as well as the ability to publish any arbitrary event, that is an object that does not necessarily extend from ApplicationEvent. When such an object is published we wrap it in aPayloadApplicationEvent for you.

Spring provides the following standard events:

Table 7.7. Built-in Events

Event Explanation

ContextRefreshedEvent

Published when the ApplicationContext is initialized or refreshed, for example, using the refresh() method on theConfigurableApplicationContext interface. "Initialized" here means that all beans are loaded, post-processor beans are detected and activated, singletons are pre-instantiated, and the ApplicationContext object is ready for use. As long as the context has not been closed, a refresh can be triggered multiple times, provided that the chosen ApplicationContext actually supports such "hot" refreshes. For example, XmlWebApplicationContext supports hot refreshes, butGenericApplicationContext does not.

ContextStartedEvent

Published when the ApplicationContext is started, using the start() method on the ConfigurableApplicationContextinterface. "Started" here means that all Lifecycle beans receive an explicit start signal. Typically this signal is used to restart beans after an explicit stop, but it may also be used to start components that have not been configured for autostart , for example, components that have not already started on initialization.

ContextStoppedEvent

Published when the ApplicationContext is stopped, using the stop() method on the ConfigurableApplicationContextinterface. "Stopped" here means that all Lifecycle beans receive an explicit stop signal. A stopped context may be restarted through a start() call.

ContextClosedEvent

Published when the ApplicationContext is closed, using the close() method on the ConfigurableApplicationContextinterface. "Closed" here means that all singleton beans are destroyed. A closed context reaches its end of life; it cannot be refreshed or restarted.

RequestHandledEvent

A web-specific event telling all beans that an HTTP request has been serviced. This event is published after the request is complete. This event is only applicable to web applications using Spring’s DispatcherServlet.

You can also create and publish your own custom events. This example demonstrates a simple class that extends Spring’s ApplicationEvent base class:

public class BlackListEvent extends ApplicationEvent {

    private final String address;
    private final String test;

    public BlackListEvent(Object source, String address, String test) {
        super(source);
        this.address = address;
        this.test = test;
    }

    // accessor and other methods...

}

To publish a custom ApplicationEvent, call the publishEvent() method on an ApplicationEventPublisher. Typically this is done by creating a class that implements ApplicationEventPublisherAware and registering it as a Spring bean. The following example demonstrates such a class:

public class EmailService implements ApplicationEventPublisherAware {

    private List<String> blackList;
    private ApplicationEventPublisher publisher;

    public void setBlackList(List<String> blackList) {
        this.blackList = blackList;
    }

    public void setApplicationEventPublisher(ApplicationEventPublisher publisher) {
        this.publisher = publisher;
    }

    public void sendEmail(String address, String text) {
        if (blackList.contains(address)) {
            BlackListEvent event = new BlackListEvent(this, address, text);
            publisher.publishEvent(event);
            return;
        }
        // send email...
    }

}

At configuration time, the Spring container will detect that EmailService implements ApplicationEventPublisherAware and will automatically callsetApplicationEventPublisher(). In reality, the parameter passed in will be the Spring container itself; you’re simply interacting with the application context via itsApplicationEventPublisher interface.

To receive the custom ApplicationEvent, create a class that implements ApplicationListener and register it as a Spring bean. The following example demonstrates such a class:

public class BlackListNotifier implements ApplicationListener<BlackListEvent> {

    private String notificationAddress;

    public void setNotificationAddress(String notificationAddress) {
        this.notificationAddress = notificationAddress;
    }

    public void onApplicationEvent(BlackListEvent event) {
        // notify appropriate parties via notificationAddress...
    }

}

Notice that ApplicationListener is generically parameterized with the type of your custom event, BlackListEvent. This means that the onApplicationEvent()method can remain type-safe, avoiding any need for downcasting. You may register as many event listeners as you wish, but note that by default event listeners receive events synchronously. This means the publishEvent() method blocks until all listeners have finished processing the event. One advantage of this synchronous and single-threaded approach is that when a listener receives an event, it operates inside the transaction context of the publisher if a transaction context is available. If another strategy for event publication becomes necessary, refer to the JavaDoc for Spring’s ApplicationEventMulticaster interface.

The following example shows the bean definitions used to register and configure each of the classes above:

<bean id="emailService" class="example.EmailService">
    <property name="blackList">
        <list>
            <value>known.spammer@example.org</value>
            <value>known.hacker@example.org</value>
            <value>john.doe@example.org</value>
        </list>
    </property>
</bean>

<bean id="blackListNotifier" class="example.BlackListNotifier">
    <property name="notificationAddress" value="blacklist@example.org"/>
</bean>

Putting it all together, when the sendEmail() method of the emailService bean is called, if there are any emails that should be blacklisted, a custom event of typeBlackListEvent is published. The blackListNotifier bean is registered as an ApplicationListener and thus receives the BlackListEvent, at which point it can notify appropriate parties.

Spring框架参考文档 Spring Framework Reference Documentation

Spring’s eventing mechanism is designed for simple communication between Spring beans within the same application context. However, for more sophisticated enterprise integration needs, the separately-maintained Spring Integration project provides complete support for building lightweight, pattern-oriented, event-driven architectures that build upon the well-known Spring programming model.

Annotation-based Event Listeners

As of Spring 4.2, an event listener can be registered on any public method of a managed bean via the EventListener annotation. The BlackListNotifier can be rewritten as follows:

public class BlackListNotifier {

    private String notificationAddress;

    public void setNotificationAddress(String notificationAddress) {
        this.notificationAddress = notificationAddress;
    }

    @EventListener
    public void processBlackListEvent(BlackListEvent event) {
        // notify appropriate parties via notificationAddress...
    }

}

As you can see above, the method signature actually infer which even type it listens to. This also works for nested generics as long as the actual event resolves the generics parameter you would filter on.

If your method should listen to several events or if you want to define it with no parameter at all, the event type(s) can also be specified on the annotation itself:

@EventListener({ContextStartedEvent.class, ContextRefreshedEvent.class})
public void handleContextStart() {

}

It is also possible to add additional runtime filtering via the condition attribute of the annotation that defines a SpEL expression that should match to actually invoke the method for a particular event.

For instance, our notifier can be rewritten to be only invoked if the test attribute of the event is equal to foo:

@EventListener(condition = "#event.test == 'foo'")
public void processBlackListEvent(BlackListEvent event) {
    // notify appropriate parties via notificationAddress...
}

Each SpEL expression evaluates again a dedicated context. The next table lists the items made available to the context so one can use them for conditional event processing:

Table 7.8. Event SpEL available metadata

Name Location Description Example

event

root object

The actual ApplicationEvent

#root.event

args

root object

The arguments (as array) used for invoking the target

#root.args[0]

argument name

evaluation context

Name of any of the method arguments. If for some reason the names are not available (e.g. no debug information), the argument names are also available under the #a<#arg> where #arg stands for the argument index (starting from 0).

#iban or #a0 (one can also use #p0 or #p<#arg>notation as an alias).

Note that #root.event allows you to access to the underlying event, even if your method signature actually refers to an arbitrary object that was published.

If you need to publish an event as the result of processing another, just change the method signature to return the event that should be published, something like:

@EventListener
public ListUpdateEvent handleBlackListEvent(BlackListEvent event) {
    // notify appropriate parties via notificationAddress and
    // then publish a ListUpdateEvent...
}
Spring框架参考文档 Spring Framework Reference Documentation

This feature is not supported for asynchronous listeners.

This new method will publish a new ListUpdateEvent for every BlackListEvent handled by the method above. If you need to publish several events, just return aCollection of events instead.

Asynchronous Listeners

If you want a particular listener to process events asynchronously, simply reuse the regular @Async support:

@EventListener
@Async
public void processBlackListEvent(BlackListEvent event) {
    // BlackListEvent is processed in a separate thread
}

Be aware of the following limitations when using asynchronous events:

  1. If the event listener throws an Exception it will not be propagated to the caller, check AsyncUncaughtExceptionHandler for more details.
  2. Such event listener cannot send replies. If you need to send another event as the result of the processing, inject ApplicationEventPublisher to send the event manually.

Ordering Listeners

If you need the listener to be invoked before another one, just add the @Order annotation to the method declaration:

@EventListener
@Order(42)
public void processBlackListEvent(BlackListEvent event) {
    // notify appropriate parties via notificationAddress...
}

Generic Events

You may also use generics to further define the structure of your event. Consider an EntityCreatedEvent<T> where T is the type of the actual entity that got created. You can create the following listener definition to only receive EntityCreatedEvent for a Person:

@EventListener
public void onPersonCreated(EntityCreatedEvent<Person> event) {
    ...
}

Due to type erasure, this will only work if the event that is fired resolves the generic parameter(s) on which the event listener filters on (that is something likeclass PersonCreatedEvent extends EntityCreatedEvent<Person> { …​ }).

In certain circumstances, this may become quite tedious if all events follow the same structure (as it should be the case for the event above). In such a case, you can implement ResolvableTypeProvider to guide the framework beyond what the runtime environment provides:

public class EntityCreatedEvent<T>
        extends ApplicationEvent implements ResolvableTypeProvider {

    public EntityCreatedEvent(T entity) {
        super(entity);
    }

    @Override
    public ResolvableType getResolvableType() {
        return ResolvableType.forClassWithGenerics(getClass(),
                ResolvableType.forInstance(getSource()));
    }
}
Spring框架参考文档 Spring Framework Reference Documentation

This works not only for ApplicationEvent but any arbitrary object that you’d send as an event.

7.15.3 Convenient access to low-level resources

For optimal usage and understanding of application contexts, users should generally familiarize themselves with Spring’s Resource abstraction, as described in the chapter Chapter 8, Resources.

An application context is a ResourceLoader, which can be used to load Resources. A Resource is essentially a more feature rich version of the JDK classjava.net.URL, in fact, the implementations of the Resource wrap an instance of java.net.URL where appropriate. A Resource can obtain low-level resources from almost any location in a transparent fashion, including from the classpath, a filesystem location, anywhere describable with a standard URL, and some other variations. If the resource location string is a simple path without any special prefixes, where those resources come from is specific and appropriate to the actual application context type.

You can configure a bean deployed into the application context to implement the special callback interface, ResourceLoaderAware, to be automatically called back at initialization time with the application context itself passed in as the ResourceLoader. You can also expose properties of type Resource, to be used to access static resources; they will be injected into it like any other properties. You can specify those Resource properties as simple String paths, and rely on a special JavaBeanPropertyEditor that is automatically registered by the context, to convert those text strings to actual Resource objects when the bean is deployed.

The location path or paths supplied to an ApplicationContext constructor are actually resource strings, and in simple form are treated appropriately to the specific context implementation. ClassPathXmlApplicationContext treats a simple location path as a classpath location. You can also use location paths (resource strings) with special prefixes to force loading of definitions from the classpath or a URL, regardless of the actual context type.

7.15.4 Convenient ApplicationContext instantiation for web applications

You can create ApplicationContext instances declaratively by using, for example, a ContextLoader. Of course you can also create ApplicationContextinstances programmatically by using one of the ApplicationContext implementations.

You can register an ApplicationContext using the ContextLoaderListener as follows:

<context-param>
    <param-name>contextConfigLocation</param-name>
    <param-value>/WEB-INF/daoContext.xml /WEB-INF/applicationContext.xml</param-value>
</context-param>

<listener>
    <listener-class>org.springframework.web.context.ContextLoaderListener</listener-class>
</listener>

The listener inspects the contextConfigLocation parameter. If the parameter does not exist, the listener uses /WEB-INF/applicationContext.xml as a default. When the parameter does exist, the listener separates the String by using predefined delimiters (comma, semicolon and whitespace) and uses the values as locations where application contexts will be searched. Ant-style path patterns are supported as well. Examples are /WEB-INF/*Context.xml for all files with names ending with "Context.xml", residing in the "WEB-INF" directory, and /WEB-INF/**/*Context.xml, for all such files in any subdirectory of "WEB-INF".

7.15.5 Deploying a Spring ApplicationContext as a Java EE RAR file

It is possible to deploy a Spring ApplicationContext as a RAR file, encapsulating the context and all of its required bean classes and library JARs in a Java EE RAR deployment unit. This is the equivalent of bootstrapping a standalone ApplicationContext, just hosted in Java EE environment, being able to access the Java EE servers facilities. RAR deployment is more natural alternative to scenario of deploying a headless WAR file, in effect, a WAR file without any HTTP entry points that is used only for bootstrapping a Spring ApplicationContext in a Java EE environment.

RAR deployment is ideal for application contexts that do not need HTTP entry points but rather consist only of message endpoints and scheduled jobs. Beans in such a context can use application server resources such as the JTA transaction manager and JNDI-bound JDBC DataSources and JMS ConnectionFactory instances, and may also register with the platform’s JMX server - all through Spring’s standard transaction management and JNDI and JMX support facilities. Application components can also interact with the application server’s JCA WorkManager through Spring’s TaskExecutor abstraction.

Check out the JavaDoc of the SpringContextResourceAdapter class for the configuration details involved in RAR deployment.

For a simple deployment of a Spring ApplicationContext as a Java EE RAR file: package all application classes into a RAR file, which is a standard JAR file with a different file extension. Add all required library JARs into the root of the RAR archive. Add a "META-INF/ra.xml" deployment descriptor (as shown inSpringContextResourceAdapters JavaDoc) and the corresponding Spring XML bean definition file(s) (typically "META-INF/applicationContext.xml"), and drop the resulting RAR file into your application server’s deployment directory.

Spring框架参考文档 Spring Framework Reference Documentation

Such RAR deployment units are usually self-contained; they do not expose components to the outside world, not even to other modules of the same application. Interaction with a RAR-based ApplicationContext usually occurs through JMS destinations that it shares with other modules. A RAR-based ApplicationContext may also, for example, schedule some jobs, reacting to new files in the file system (or the like). If it needs to allow synchronous access from the outside, it could for example export RMI endpoints, which of course may be used by other application modules on the same machine.

7.16 The BeanFactory

The BeanFactory provides the underlying basis for Spring’s IoC functionality but it is only used directly in integration with other third-party frameworks and is now largely historical in nature for most users of Spring. The BeanFactory and related interfaces, such as BeanFactoryAware, InitializingBean, DisposableBean, are still present in Spring for the purposes of backward compatibility with the large number of third-party frameworks that integrate with Spring. Often third-party components that can not use more modern equivalents such as @PostConstruct or @PreDestroy in order to remain compatible with JDK 1.4 or to avoid a dependency on JSR-250.

This section provides additional background into the differences between the BeanFactory and ApplicationContext and how one might access the IoC container directly through a classic singleton lookup.

7.16.1 BeanFactory or ApplicationContext?

Use an ApplicationContext unless you have a good reason for not doing so.

Because the ApplicationContext includes all functionality of the BeanFactory, it is generally recommended over the BeanFactory, except for a few situations such as in embedded applications running on resource-constrained devices where memory consumption might be critical and a few extra kilobytes might make a difference. However, for most typical enterprise applications and systems, the ApplicationContext is what you will want to use. Spring makes heavy use of theBeanPostProcessor extension point (to effect proxying and so on). If you use only a plain BeanFactory, a fair amount of support such as transactions and AOP will not take effect, at least not without some extra steps on your part. This situation could be confusing because nothing is actually wrong with the configuration.

The following table lists features provided by the BeanFactory and ApplicationContext interfaces and implementations.

Table 7.9. Feature Matrix

Feature BeanFactory ApplicationContext

Bean instantiation/wiring

Yes

Yes

Automatic BeanPostProcessor registration

No

Yes

Automatic BeanFactoryPostProcessor registration

No

Yes

Convenient MessageSource access (for i18n)

No

Yes

ApplicationEvent publication

No

Yes

To explicitly register a bean post-processor with a BeanFactory implementation, you need to write code like this:

DefaultListableBeanFactory factory = new DefaultListableBeanFactory();
// populate the factory with bean definitions

// now register any needed BeanPostProcessor instances
MyBeanPostProcessor postProcessor = new MyBeanPostProcessor();
factory.addBeanPostProcessor(postProcessor);

// now start using the factory

To explicitly register a BeanFactoryPostProcessor when using a BeanFactory implementation, you must write code like this:

DefaultListableBeanFactory factory = new DefaultListableBeanFactory();
XmlBeanDefinitionReader reader = new XmlBeanDefinitionReader(factory);
reader.loadBeanDefinitions(new FileSystemResource("beans.xml"));

// bring in some property values from a Properties file
PropertyPlaceholderConfigurer cfg = new PropertyPlaceholderConfigurer();
cfg.setLocation(new FileSystemResource("jdbc.properties"));

// now actually do the replacement
cfg.postProcessBeanFactory(factory);

In both cases, the explicit registration step is inconvenient, which is one reason why the various ApplicationContext implementations are preferred above plainBeanFactory implementations in the vast majority of Spring-backed applications, especially when using BeanFactoryPostProcessors and BeanPostProcessors. These mechanisms implement important functionality such as property placeholder replacement and AOP.

7.16.2 Glue code and the evil singleton

It is best to write most application code in a dependency-injection (DI) style, where that code is served out of a Spring IoC container, has its own dependencies supplied by the container when it is created, and is completely unaware of the container. However, for the small glue layers of code that are sometimes needed to tie other code together, you sometimes need a singleton (or quasi-singleton) style access to a Spring IoC container. For example, third-party code may try to construct new objects directly ( Class.forName() style), without the ability to get these objects out of a Spring IoC container.If the object constructed by the third-party code is a small stub or proxy, which then uses a singleton style access to a Spring IoC container to get a real object to delegate to, then inversion of control has still been achieved for the majority of the code (the object coming out of the container). Thus most code is still unaware of the container or how it is accessed, and remains decoupled from other code, with all ensuing benefits. EJBs may also use this stub/proxy approach to delegate to a plain Java implementation object, retrieved from a Spring IoC container. While the Spring IoC container itself ideally does not have to be a singleton, it may be unrealistic in terms of memory usage or initialization times (when using beans in the Spring IoC container such as a Hibernate SessionFactory) for each bean to use its own, non-singleton Spring IoC container.

Looking up the application context in a service locator style is sometimes the only option for accessing shared Spring-managed components, such as in an EJB 2.1 environment, or when you want to share a single ApplicationContext as a parent to WebApplicationContexts across WAR files. In this case you should look into using the utility class ContextSingletonBeanFactoryLocator locator that is described in this Spring team blog entry.


8. Resources

8.1 Introduction

Java’s standard java.net.URL class and standard handlers for various URL prefixes unfortunately are not quite adequate enough for all access to low-level resources. For example, there is no standardized URL implementation that may be used to access a resource that needs to be obtained from the classpath, or relative to aServletContext. While it is possible to register new handlers for specialized URL prefixes (similar to existing handlers for prefixes such as http:), this is generally quite complicated, and the URL interface still lacks some desirable functionality, such as a method to check for the existence of the resource being pointed to.

8.2 The Resource interface

Spring’s Resource interface is meant to be a more capable interface for abstracting access to low-level resources.

public interface Resource extends InputStreamSource {

    boolean exists();

    boolean isOpen();

    URL getURL() throws IOException;

    File getFile() throws IOException;

    Resource createRelative(String relativePath) throws IOException;

    String getFilename();

    String getDescription();

}
public interface InputStreamSource {

    InputStream getInputStream() throws IOException;

}

Some of the most important methods from the Resource interface are:

  • getInputStream(): locates and opens the resource, returning an InputStream for reading from the resource. It is expected that each invocation returns a freshInputStream. It is the responsibility of the caller to close the stream.
  • exists(): returns a boolean indicating whether this resource actually exists in physical form.
  • isOpen(): returns a boolean indicating whether this resource represents a handle with an open stream. If true, the InputStream cannot be read multiple times, and must be read once only and then closed to avoid resource leaks. Will be false for all usual resource implementations, with the exception ofInputStreamResource.
  • getDescription(): returns a description for this resource, to be used for error output when working with the resource. This is often the fully qualified file name or the actual URL of the resource.

Other methods allow you to obtain an actual URL or File object representing the resource (if the underlying implementation is compatible, and supports that functionality).

The Resource abstraction is used extensively in Spring itself, as an argument type in many method signatures when a resource is needed. Other methods in some Spring APIs (such as the constructors to various ApplicationContext implementations), take a String which in unadorned or simple form is used to create aResource appropriate to that context implementation, or via special prefixes on the String path, allow the caller to specify that a specific Resource implementation must be created and used.

While the Resource interface is used a lot with Spring and by Spring, it’s actually very useful to use as a general utility class by itself in your own code, for access to resources, even when your code doesn’t know or care about any other parts of Spring. While this couples your code to Spring, it really only couples it to this small set of utility classes, which are serving as a more capable replacement for URL, and can be considered equivalent to any other library you would use for this purpose.

It is important to note that the Resource abstraction does not replace functionality: it wraps it where possible. For example, a UrlResource wraps a URL, and uses the wrapped URL to do its work.

8.3 Built-in Resource implementations

There are a number of Resource implementations that come supplied straight out of the box in Spring:

8.3.1 UrlResource

The UrlResource wraps a java.net.URL, and may be used to access any object that is normally accessible via a URL, such as files, an HTTP target, an FTP target, etc. All URLs have a standardized String representation, such that appropriate standardized prefixes are used to indicate one URL type from another. This includesfile: for accessing filesystem paths, http: for accessing resources via the HTTP protocol, ftp: for accessing resources via FTP, etc.

A UrlResource is created by Java code explicitly using the UrlResource constructor, but will often be created implicitly when you call an API method which takes aString argument which is meant to represent a path. For the latter case, a JavaBeans PropertyEditor will ultimately decide which type of Resource to create. If the path string contains a few well-known (to it, that is) prefixes such as classpath:, it will create an appropriate specialized Resource for that prefix. However, if it doesn’t recognize the prefix, it will assume the this is just a standard URL string, and will create a UrlResource.

8.3.2 ClassPathResource

This class represents a resource which should be obtained from the classpath. This uses either the thread context class loader, a given class loader, or a given class for loading resources.

This Resource implementation supports resolution as java.io.File if the class path resource resides in the file system, but not for classpath resources which reside in a jar and have not been expanded (by the servlet engine, or whatever the environment is) to the filesystem. To address this the various Resource implementations always support resolution as a java.net.URL.

A ClassPathResource is created by Java code explicitly using the ClassPathResource constructor, but will often be created implicitly when you call an API method which takes a String argument which is meant to represent a path. For the latter case, a JavaBeans PropertyEditor will recognize the special prefix classpath:on the string path, and create a ClassPathResource in that case.

8.3.3 FileSystemResource

This is a Resource implementation for java.io.File handles. It obviously supports resolution as a File, and as a URL.

8.3.4 ServletContextResource

This is a Resource implementation for ServletContext resources, interpreting relative paths within the relevant web application’s root directory.

This always supports stream access and URL access, but only allows java.io.File access when the web application archive is expanded and the resource is physically on the filesystem. Whether or not it’s expanded and on the filesystem like this, or accessed directly from the JAR or somewhere else like a DB (it’s conceivable) is actually dependent on the Servlet container.

8.3.5 InputStreamResource

A Resource implementation for a given InputStream. This should only be used if no specific Resource implementation is applicable. In particular, preferByteArrayResource or any of the file-based Resource implementations where possible.

In contrast to other Resource implementations, this is a descriptor for an already opened resource - therefore returning true from isOpen(). Do not use it if you need to keep the resource descriptor somewhere, or if you need to read a stream multiple times.

8.3.6 ByteArrayResource

This is a Resource implementation for a given byte array. It creates a ByteArrayInputStream for the given byte array.

It’s useful for loading content from any given byte array, without having to resort to a single-use InputStreamResource.

8.4 The ResourceLoader

The ResourceLoader interface is meant to be implemented by objects that can return (i.e. load) Resource instances.

public interface ResourceLoader {

    Resource getResource(String location);

}

All application contexts implement the ResourceLoader interface, and therefore all application contexts may be used to obtain Resource instances.

When you call getResource() on a specific application context, and the location path specified doesn’t have a specific prefix, you will get back a Resource type that is appropriate to that particular application context. For example, assume the following snippet of code was executed against a ClassPathXmlApplicationContextinstance:

Resource template = ctx.getResource("some/resource/path/myTemplate.txt");

What would be returned would be a ClassPathResource; if the same method was executed against a FileSystemXmlApplicationContext instance, you’d get back a FileSystemResource. For a WebApplicationContext, you’d get back a ServletContextResource, and so on.

As such, you can load resources in a fashion appropriate to the particular application context.

On the other hand, you may also force ClassPathResource to be used, regardless of the application context type, by specifying the special classpath: prefix:

Resource template = ctx.getResource("classpath:some/resource/path/myTemplate.txt");

Similarly, one can force a UrlResource to be used by specifying any of the standard java.net.URL prefixes:

Resource template = ctx.getResource("file:///some/resource/path/myTemplate.txt");
Resource template = ctx.getResource("http://myhost.com/resource/path/myTemplate.txt");

The following table summarizes the strategy for converting Strings to Resources:

Table 8.1. Resource strings

Prefix Example Explanation

classpath:

classpath:com/myapp/config.xml

Loaded from the classpath.

file:

file:///data/config.xml

Loaded as a URL, from the filesystem. [1]

http:

http://myserver/logo.png

Loaded as a URL.

(none)

/data/config.xml

Depends on the underlying ApplicationContext.

8.5 The ResourceLoaderAware interface

The ResourceLoaderAware interface is a special marker interface, identifying objects that expect to be provided with a ResourceLoader reference.

public interface ResourceLoaderAware {

    void setResourceLoader(ResourceLoader resourceLoader);
}

When a class implements ResourceLoaderAware and is deployed into an application context (as a Spring-managed bean), it is recognized as ResourceLoaderAwareby the application context. The application context will then invoke the setResourceLoader(ResourceLoader), supplying itself as the argument (remember, all application contexts in Spring implement the ResourceLoader interface).

Of course, since an ApplicationContext is a ResourceLoader, the bean could also implement the ApplicationContextAware interface and use the supplied application context directly to load resources, but in general, it’s better to use the specialized ResourceLoader interface if that’s all that’s needed. The code would just be coupled to the resource loading interface, which can be considered a utility interface, and not the whole Spring ApplicationContext interface.

As of Spring 2.5, you can rely upon autowiring of the ResourceLoader as an alternative to implementing the ResourceLoaderAware interface. The "traditional"constructor and byType autowiring modes (as described in Section 7.4.5, “Autowiring collaborators”) are now capable of providing a dependency of typeResourceLoader for either a constructor argument or setter method parameter respectively. For more flexibility (including the ability to autowire fields and multiple parameter methods), consider using the new annotation-based autowiring features. In that case, the ResourceLoader will be autowired into a field, constructor argument, or method parameter that is expecting the ResourceLoader type as long as the field, constructor, or method in question carries the @Autowired annotation. For more information, see Section 7.9.2, “@Autowired”.

8.6 Resources as dependencies

If the bean itself is going to determine and supply the resource path through some sort of dynamic process, it probably makes sense for the bean to use theResourceLoader interface to load resources. Consider as an example the loading of a template of some sort, where the specific resource that is needed depends on the role of the user. If the resources are static, it makes sense to eliminate the use of the ResourceLoader interface completely, and just have the bean expose theResource properties it needs, and expect that they will be injected into it.

What makes it trivial to then inject these properties, is that all application contexts register and use a special JavaBeans PropertyEditor which can convert Stringpaths to Resource objects. So if myBean has a template property of type Resource, it can be configured with a simple string for that resource, as follows:

<bean id="myBean" class="...">
    <property name="template" value="some/resource/path/myTemplate.txt"/>
</bean>

Note that the resource path has no prefix, so because the application context itself is going to be used as the ResourceLoader, the resource itself will be loaded via aClassPathResource, FileSystemResource, or ServletContextResource (as appropriate) depending on the exact type of the context.

If there is a need to force a specific Resource type to be used, then a prefix may be used. The following two examples show how to force a ClassPathResource and aUrlResource (the latter being used to access a filesystem file).

<property name="template" value="classpath:some/resource/path/myTemplate.txt">
<property name="template" value="file:///some/resource/path/myTemplate.txt"/>

8.7 Application contexts and Resource paths

8.7.1 Constructing application contexts

An application context constructor (for a specific application context type) generally takes a string or array of strings as the location path(s) of the resource(s) such as XML files that make up the definition of the context.

When such a location path doesn’t have a prefix, the specific Resource type built from that path and used to load the bean definitions, depends on and is appropriate to the specific application context. For example, if you create a ClassPathXmlApplicationContext as follows:

ApplicationContext ctx = new ClassPathXmlApplicationContext("conf/appContext.xml");

The bean definitions will be loaded from the classpath, as a ClassPathResource will be used. But if you create a FileSystemXmlApplicationContext as follows:

ApplicationContext ctx =
    new FileSystemXmlApplicationContext("conf/appContext.xml");

The bean definition will be loaded from a filesystem location, in this case relative to the current working directory.

Note that the use of the special classpath prefix or a standard URL prefix on the location path will override the default type of Resource created to load the definition. So this FileSystemXmlApplicationContext…​

ApplicationContext ctx =
    new FileSystemXmlApplicationContext("classpath:conf/appContext.xml");
  1. will actually load its bean definitions from the classpath. However, it is still a FileSystemXmlApplicationContext. If it is subsequently used as aResourceLoader, any unprefixed paths will still be treated as filesystem paths.

Constructing ClassPathXmlApplicationContext instances - shortcuts

The ClassPathXmlApplicationContext exposes a number of constructors to enable convenient instantiation. The basic idea is that one supplies merely a string array containing just the filenames of the XML files themselves (without the leading path information), and one also supplies a Class; theClassPathXmlApplicationContext will derive the path information from the supplied class.

An example will hopefully make this clear. Consider a directory layout that looks like this:

com/
  foo/
	services.xml
	daos.xml
    MessengerService.class

A ClassPathXmlApplicationContext instance composed of the beans defined in the 'services.xml' and 'daos.xml' could be instantiated like so…​

ApplicationContext ctx = new ClassPathXmlApplicationContext(
    new String[] {"services.xml", "daos.xml"}, MessengerService.class);

Please do consult the ClassPathXmlApplicationContext javadocs for details on the various constructors.

8.7.2 Wildcards in application context constructor resource paths

The resource paths in application context constructor values may be a simple path (as shown above) which has a one-to-one mapping to a target Resource, or alternately may contain the special "classpath*:" prefix and/or internal Ant-style regular expressions (matched using Spring’s PathMatcher utility). Both of the latter are effectively wildcards

One use for this mechanism is when doing component-style application assembly. All components can 'publish' context definition fragments to a well-known location path, and when the final application context is created using the same path prefixed via classpath*:, all component fragments will be picked up automatically.

Note that this wildcarding is specific to use of resource paths in application context constructors (or when using the PathMatcher utility class hierarchy directly), and is resolved at construction time. It has nothing to do with the Resource type itself. It’s not possible to use the classpath*: prefix to construct an actual Resource, as a resource points to just one resource at a time.

Ant-style Patterns

When the path location contains an Ant-style pattern, for example:

/WEB-INF/*-context.xml
  com/mycompany/**/applicationContext.xml
  file:C:/some/path/*-context.xml
  classpath:com/mycompany/**/applicationContext.xml
  1. the resolver follows a more complex but defined procedure to try to resolve the wildcard. It produces a Resource for the path up to the last non-wildcard segment and obtains a URL from it. If this URL is not a "jar:" URL or container-specific variant (e.g. "zip:`" in WebLogic, " `wsjar`" in WebSphere, etc.), then a `java.io.File is obtained from it and used to resolve the wildcard by traversing the filesystem. In the case of a jar URL, the resolver either gets a java.net.JarURLConnection from it or manually parses the jar URL and then traverses the contents of the jar file to resolve the wildcards.
Implications on portability

If the specified path is already a file URL (either explicitly, or implicitly because the base ResourceLoader is a filesystem one, then wildcarding is guaranteed to work in a completely portable fashion.

If the specified path is a classpath location, then the resolver must obtain the last non-wildcard path segment URL via a Classloader.getResource() call. Since this is just a node of the path (not the file at the end) it is actually undefined (in the ClassLoader javadocs) exactly what sort of a URL is returned in this case. In practice, it is always a java.io.File representing the directory, where the classpath resource resolves to a filesystem location, or a jar URL of some sort, where the classpath resource resolves to a jar location. Still, there is a portability concern on this operation.

If a jar URL is obtained for the last non-wildcard segment, the resolver must be able to get a java.net.JarURLConnection from it, or manually parse the jar URL, to be able to walk the contents of the jar, and resolve the wildcard. This will work in most environments, but will fail in others, and it is strongly recommended that the wildcard resolution of resources coming from jars be thoroughly tested in your specific environment before you rely on it.

The Classpath*: portability classpath*: prefix

When constructing an XML-based application context, a location string may use the special classpath*: prefix:

ApplicationContext ctx =
    new ClassPathXmlApplicationContext("classpath*:conf/appContext.xml");

This special prefix specifies that all classpath resources that match the given name must be obtained (internally, this essentially happens via aClassLoader.getResources(…​) call), and then merged to form the final application context definition.

Spring框架参考文档 Spring Framework Reference Documentation

The wildcard classpath relies on the getResources() method of the underlying classloader. As most application servers nowadays supply their own classloader implementation, the behavior might differ especially when dealing with jar files. A simple test to check if classpath* works is to use the classloader to load a file from within a jar on the classpath: getClass().getClassLoader().getResources("<someFileInsideTheJar>"). Try this test with files that have the same name but are placed inside two different locations. In case an inappropriate result is returned, check the application server documentation for settings that might affect the classloader behavior.

The " classpath*:`" prefix can also be combined with a `PathMatcher pattern in the rest of the location path, for example " `classpath*:META-INF/*-beans.xml`". In this case, the resolution strategy is fairly simple: a ClassLoader.getResources() call is used on the last non-wildcard path segment to get all the matching resources in the class loader hierarchy, and then off each resource the same PathMatcher resolution strategy described above is used for the wildcard subpath.

Other notes relating to wildcards

Please note that classpath*: when combined with Ant-style patterns will only work reliably with at least one root directory before the pattern starts, unless the actual target files reside in the file system. This means that a pattern like "classpath*:*.xml`" will not retrieve files from the root of jar files but rather only from the root of expanded directories. This originates from a limitation in the JDK’s `ClassLoader.getResources()method which only returns file system locations for a passed-in empty string (indicating potential roots to search).

Ant-style patterns with " `classpath:`" resources are not guaranteed to find matching resources if the root package to search is available in multiple class path locations. This is because a resource such as

com/mycompany/package1/service-context.xml

may be in only one location, but when a path such as

classpath:com/mycompany/**/service-context.xml

is used to try to resolve it, the resolver will work off the (first) URL returned by getResource("com/mycompany");. If this base package node exists in multiple classloader locations, the actual end resource may not be underneath. Therefore, preferably, use " `classpath*:`" with the same Ant-style pattern in such a case, which will search all class path locations that contain the root package.

8.7.3 FileSystemResource caveats

A FileSystemResource that is not attached to a FileSystemApplicationContext (that is, a FileSystemApplicationContext is not the actual ResourceLoader) will treat absolute vs. relative paths as you would expect. Relative paths are relative to the current working directory, while absolute paths are relative to the root of the filesystem.

For backwards compatibility (historical) reasons however, this changes when the FileSystemApplicationContext is the ResourceLoader. TheFileSystemApplicationContext simply forces all attached FileSystemResource instances to treat all location paths as relative, whether they start with a leading slash or not. In practice, this means the following are equivalent:

ApplicationContext ctx =
    new FileSystemXmlApplicationContext("conf/context.xml");
ApplicationContext ctx =
    new FileSystemXmlApplicationContext("/conf/context.xml");

As are the following: (Even though it would make sense for them to be different, as one case is relative and the other absolute.)

FileSystemXmlApplicationContext ctx = ...;
ctx.getResource("some/resource/path/myTemplate.txt");
FileSystemXmlApplicationContext ctx = ...;
ctx.getResource("/some/resource/path/myTemplate.txt");

In practice, if true absolute filesystem paths are needed, it is better to forgo the use of absolute paths with FileSystemResource /FileSystemXmlApplicationContext, and just force the use of a UrlResource, by using the file: URL prefix.

// actual context type doesn't matter, the Resource will always be UrlResource
ctx.getResource("file:///some/resource/path/myTemplate.txt");
// force this FileSystemXmlApplicationContext to load its definition via a UrlResource
ApplicationContext ctx =
    new FileSystemXmlApplicationContext("file:///conf/context.xml");

9. Validation, Data Binding, and Type Conversion

9.1 Introduction

JSR-303/JSR-349 Bean Validation

Spring Framework 4.0 supports Bean Validation 1.0 (JSR-303) and Bean Validation 1.1 (JSR-349) in terms of setup support, also adapting it to Spring’sValidator interface.

An application can choose to enable Bean Validation once globally, as described in Section 9.8, “Spring Validation”, and use it exclusively for all validation needs.

An application can also register additional Spring Validator instances per DataBinder instance, as described in Section 9.8.3, “Configuring a DataBinder”. This may be useful for plugging in validation logic without the use of annotations.

There are pros and cons for considering validation as business logic, and Spring offers a design for validation (and data binding) that does not exclude either one of them. Specifically validation should not be tied to the web tier, should be easy to localize and it should be possible to plug in any validator available. Considering the above, Spring has come up with a Validator interface that is both basic and eminently usable in every layer of an application.

Data binding is useful for allowing user input to be dynamically bound to the domain model of an application (or whatever objects you use to process user input). Spring provides the so-called DataBinder to do exactly that. The Validator and the DataBinder make up the validation package, which is primarily used in but not limited to the MVC framework.

The BeanWrapper is a fundamental concept in the Spring Framework and is used in a lot of places. However, you probably will not have the need to use theBeanWrapper directly. Because this is reference documentation however, we felt that some explanation might be in order. We will explain the BeanWrapper in this chapter since, if you were going to use it at all, you would most likely do so when trying to bind data to objects.

Spring’s DataBinder and the lower-level BeanWrapper both use PropertyEditors to parse and format property values. The PropertyEditor concept is part of the JavaBeans specification, and is also explained in this chapter. Spring 3 introduces a "core.convert" package that provides a general type conversion facility, as well as a higher-level "format" package for formatting UI field values. These new packages may be used as simpler alternatives to PropertyEditors, and will also be discussed in this chapter.

9.2 Validation using Spring’s Validator interface

Spring features a Validator interface that you can use to validate objects. The Validator interface works using an Errors object so that while validating, validators can report validation failures to the Errors object.

Let’s consider a small data object:

public class Person {

    private String name;
    private int age;

    // the usual getters and setters...
}

We’re going to provide validation behavior for the Person class by implementing the following two methods of the org.springframework.validation.Validatorinterface:

  • supports(Class) - Can this Validator validate instances of the supplied Class?
  • validate(Object, org.springframework.validation.Errors) - validates the given object and in case of validation errors, registers those with the givenErrors object

Implementing a Validator is fairly straightforward, especially when you know of the ValidationUtils helper class that the Spring Framework also provides.

public class PersonValidator implements Validator {

    /**
     * This Validator validates *just* Person instances
     */
    public boolean supports(Class clazz) {
        return Person.class.equals(clazz);
    }

    public void validate(Object obj, Errors e) {
        ValidationUtils.rejectIfEmpty(e, "name", "name.empty");
        Person p = (Person) obj;
        if (p.getAge() < 0) {
            e.rejectValue("age", "negativevalue");
        } else if (p.getAge() > 110) {
            e.rejectValue("age", "too.darn.old");
        }
    }
}

As you can see, the static rejectIfEmpty(..) method on the ValidationUtils class is used to reject the 'name' property if it is null or the empty string. Have a look at the ValidationUtils javadocs to see what functionality it provides besides the example shown previously.

While it is certainly possible to implement a single Validator class to validate each of the nested objects in a rich object, it may be better to encapsulate the validation logic for each nested class of object in its own Validator implementation. A simple example of a 'rich' object would be a Customer that is composed of two Stringproperties (a first and second name) and a complex Address object. Address objects may be used independently of Customer objects, and so a distinctAddressValidator has been implemented. If you want your CustomerValidator to reuse the logic contained within the AddressValidator class without resorting to copy-and-paste, you can dependency-inject or instantiate an AddressValidator within your CustomerValidator, and use it like so:

public class CustomerValidator implements Validator {

    private final Validator addressValidator;

    public CustomerValidator(Validator addressValidator) {
        if (addressValidator == null) {
            throw new IllegalArgumentException("The supplied [Validator] is " +
                "required and must not be null.");
        }
        if (!addressValidator.supports(Address.class)) {
            throw new IllegalArgumentException("The supplied [Validator] must " +
                support the validation of [Address] instances.");
        }
        this.addressValidator = addressValidator;
    }

    /**
     * This Validator validates Customer instances, and any subclasses of Customer too
     */
    public boolean supports(Class clazz) {
        return Customer.class.isAssignableFrom(clazz);
    }

    public void validate(Object target, Errors errors) {
        ValidationUtils.rejectIfEmptyOrWhitespace(errors, "firstName", "field.required");
        ValidationUtils.rejectIfEmptyOrWhitespace(errors, "surname", "field.required");
        Customer customer = (Customer) target;
        try {
            errors.pushNestedPath("address");
            ValidationUtils.invokeValidator(this.addressValidator, customer.getAddress(), errors);
        } finally {
            errors.popNestedPath();
        }
    }
}

Validation errors are reported to the Errors object passed to the validator. In case of Spring Web MVC you can use <spring:bind/> tag to inspect the error messages, but of course you can also inspect the errors object yourself. More information about the methods it offers can be found in the javadocs.

9.3 Resolving codes to error messages

We’ve talked about databinding and validation. Outputting messages corresponding to validation errors is the last thing we need to discuss. In the example we’ve shown above, we rejected the name and the age field. If we’re going to output the error messages by using a MessageSource, we will do so using the error code we’ve given when rejecting the field ('name' and 'age' in this case). When you call (either directly, or indirectly, using for example the ValidationUtils class) rejectValue or one of the other reject methods from the Errors interface, the underlying implementation will not only register the code you’ve passed in, but also a number of additional error codes. What error codes it registers is determined by the MessageCodesResolver that is used. By default, the DefaultMessageCodesResolver is used, which for example not only registers a message with the code you gave, but also messages that include the field name you passed to the reject method. So in case you reject a field using rejectValue("age", "too.darn.old"), apart from the too.darn.old code, Spring will also register too.darn.old.age and too.darn.old.age.int(so the first will include the field name and the second will include the type of the field); this is done as a convenience to aid developers in targeting error messages and suchlike.

More information on the MessageCodesResolver and the default strategy can be found online in the javadocs of MessageCodesResolver andDefaultMessageCodesResolver, respectively.

9.4 Bean manipulation and the BeanWrapper

The org.springframework.beans package adheres to the JavaBeans standard provided by Oracle. A JavaBean is simply a class with a default no-argument constructor, which follows a naming convention where (by way of an example) a property named bingoMadness would have a setter method setBingoMadness(..)and a getter method getBingoMadness(). For more information about JavaBeans and the specification, please refer to Oracle’s website ( javabeans).

One quite important class in the beans package is the BeanWrapper interface and its corresponding implementation ( BeanWrapperImpl). As quoted from the javadocs, the BeanWrapper offers functionality to set and get property values (individually or in bulk), get property descriptors, and to query properties to determine if they are readable or writable. Also, the BeanWrapper offers support for nested properties, enabling the setting of properties on sub-properties to an unlimited depth. Then, the BeanWrapper supports the ability to add standard JavaBeans PropertyChangeListeners and VetoableChangeListeners, without the need for supporting code in the target class. Last but not least, the BeanWrapper provides support for the setting of indexed properties. The BeanWrapper usually isn’t used by application code directly, but by the DataBinder and the BeanFactory.

The way the BeanWrapper works is partly indicated by its name: it wraps a bean to perform actions on that bean, like setting and retrieving properties.

9.4.1 Setting and getting basic and nested properties

Setting and getting properties is done using the setPropertyValue(s) and getPropertyValue(s) methods that both come with a couple of overloaded variants. They’re all described in more detail in the javadocs Spring comes with. What’s important to know is that there are a couple of conventions for indicating properties of an object. A couple of examples:

Table 9.1. Examples of properties

Expression Explanation

name

Indicates the property name corresponding to the methods getName() or isName() and setName(..)

account.name

Indicates the nested property name of the property account corresponding e.g. to the methods getAccount().setName() orgetAccount().getName()

account[2]

Indicates the third element of the indexed property account. Indexed properties can be of type array, list or other naturally ordered collection

account[COMPANYNAME]

Indicates the value of the map entry indexed by the key COMPANYNAME of the Map property account

Below you’ll find some examples of working with the BeanWrapper to get and set properties.

(This next section is not vitally important to you if you’re not planning to work with the BeanWrapper directly. If you’re just using the DataBinder and the BeanFactoryand their out-of-the-box implementation, you should skip ahead to the section about PropertyEditors.)

Consider the following two classes:

public class Company {

    private String name;
    private Employee managingDirector;

    public String getName() {
        return this.name;
    }

    public void setName(String name) {
        this.name = name;
    }

    public Employee getManagingDirector() {
        return this.managingDirector;
    }

    public void setManagingDirector(Employee managingDirector) {
        this.managingDirector = managingDirector;
    }
}
public class Employee {

    private String name;

    private float salary;

    public String getName() {
        return this.name;
    }

    public void setName(String name) {
        this.name = name;
    }

    public float getSalary() {
        return salary;
    }

    public void setSalary(float salary) {
        this.salary = salary;
    }
}

The following code snippets show some examples of how to retrieve and manipulate some of the properties of instantiated Companies and Employees:

BeanWrapper company = new BeanWrapperImpl(new Company());
// setting the company name..
company.setPropertyValue("name", "Some Company Inc.");
// ... can also be done like this:
PropertyValue value = new PropertyValue("name", "Some Company Inc.");
company.setPropertyValue(value);

// ok, let's create the director and tie it to the company:
BeanWrapper jim = new BeanWrapperImpl(new Employee());
jim.setPropertyValue("name", "Jim Stravinsky");
company.setPropertyValue("managingDirector", jim.getWrappedInstance());

// retrieving the salary of the managingDirector through the company
Float salary = (Float) company.getPropertyValue("managingDirector.salary");

9.4.2 Built-in PropertyEditor implementations

Spring uses the concept of PropertyEditors to effect the conversion between an Object and a String. If you think about it, it sometimes might be handy to be able to represent properties in a different way than the object itself. For example, a Date can be represented in a human readable way (as the String '2007-14-09'), while we’re still able to convert the human readable form back to the original date (or even better: convert any date entered in a human readable form, back to Dateobjects). This behavior can be achieved by registering custom editors, of type java.beans.PropertyEditor. Registering custom editors on a BeanWrapper or alternately in a specific IoC container as mentioned in the previous chapter, gives it the knowledge of how to convert properties to the desired type. Read more aboutPropertyEditors in the javadocs of the java.beans package provided by Oracle.

A couple of examples where property editing is used in Spring:

  • setting properties on beans is done using PropertyEditors. When mentioning java.lang.String as the value of a property of some bean you’re declaring in XML file, Spring will (if the setter of the corresponding property has a Class-parameter) use the ClassEditor to try to resolve the parameter to a Class object.
  • parsing HTTP request parameters in Spring’s MVC framework is done using all kinds of PropertyEditors that you can manually bind in all subclasses of theCommandController.

Spring has a number of built-in PropertyEditors to make life easy. Each of those is listed below and they are all located in theorg.springframework.beans.propertyeditors package. Most, but not all (as indicated below), are registered by default by BeanWrapperImpl. Where the property editor is configurable in some fashion, you can of course still register your own variant to override the default one:

Table 9.2. Built-in PropertyEditors

Class Explanation

ByteArrayPropertyEditor

Editor for byte arrays. Strings will simply be converted to their corresponding byte representations. Registered by default byBeanWrapperImpl.

ClassEditor

Parses Strings representing classes to actual classes and the other way around. When a class is not found, anIllegalArgumentException is thrown. Registered by default by BeanWrapperImpl.

CustomBooleanEditor

Customizable property editor for Boolean properties. Registered by default by BeanWrapperImpl, but, can be overridden by registering custom instance of it as custom editor.

CustomCollectionEditor

Property editor for Collections, converting any source Collection to a given target Collection type.

CustomDateEditor

Customizable property editor for java.util.Date, supporting a custom DateFormat. NOT registered by default. Must be user registered as needed with appropriate format.

CustomNumberEditor

Customizable property editor for any Number subclass like Integer, Long, Float, Double. Registered by default byBeanWrapperImpl, but can be overridden by registering custom instance of it as a custom editor.

FileEditor

Capable of resolving Strings to java.io.File objects. Registered by default by BeanWrapperImpl.

InputStreamEditor

One-way property editor, capable of taking a text string and producing (via an intermediate ResourceEditor and Resource) an InputStream, so InputStream properties may be directly set as Strings. Note that the default usage will not close theInputStream for you! Registered by default by BeanWrapperImpl.

LocaleEditor

Capable of resolving Strings to Locale objects and vice versa (the String format is [country][variant], which is the same thing the toString() method of Locale provides). Registered by default by BeanWrapperImpl.

PatternEditor

Capable of resolving Strings to java.util.regex.Pattern objects and vice versa.

PropertiesEditor

Capable of converting Strings (formatted using the format as defined in the javadocs of the java.util.Properties class) toProperties objects. Registered by default by BeanWrapperImpl.

StringTrimmerEditor

Property editor that trims Strings. Optionally allows transforming an empty string into a null value. NOT registered by default; must be user registered as needed.

URLEditor

Capable of resolving a String representation of a URL to an actual URL object. Registered by default by BeanWrapperImpl.

Spring uses the java.beans.PropertyEditorManager to set the search path for property editors that might be needed. The search path also includessun.bean.editors, which includes PropertyEditor implementations for types such as Font, Color, and most of the primitive types. Note also that the standard JavaBeans infrastructure will automatically discover PropertyEditor classes (without you having to register them explicitly) if they are in the same package as the class they handle, and have the same name as that class, with 'Editor' appended; for example, one could have the following class and package structure, which would be sufficient for the FooEditor class to be recognized and used as the PropertyEditor for Foo-typed properties.

com
  chank
    pop
      Foo
      FooEditor // the PropertyEditor for the Foo class

Note that you can also use the standard BeanInfo JavaBeans mechanism here as well (described in not-amazing-detail here). Find below an example of using theBeanInfo mechanism for explicitly registering one or more PropertyEditor instances with the properties of an associated class.

com
  chank
    pop
      Foo
      FooBeanInfo // the BeanInfo for the Foo class

Here is the Java source code for the referenced FooBeanInfo class. This would associate a CustomNumberEditor with the age property of the Foo class.

public class FooBeanInfo extends SimpleBeanInfo {

    public PropertyDescriptor[] getPropertyDescriptors() {
        try {
            final PropertyEditor numberPE = new CustomNumberEditor(Integer.class, true);
            PropertyDescriptor ageDescriptor = new PropertyDescriptor("age", Foo.class) {
                public PropertyEditor createPropertyEditor(Object bean) {
                    return numberPE;
                };
            };
            return new PropertyDescriptor[] { ageDescriptor };
        }
        catch (IntrospectionException ex) {
            throw new Error(ex.toString());
        }
    }
}

Registering additional custom PropertyEditors

When setting bean properties as a string value, a Spring IoC container ultimately uses standard JavaBeans PropertyEditors to convert these Strings to the complex type of the property. Spring pre-registers a number of custom PropertyEditors (for example, to convert a classname expressed as a string into a real Class object). Additionally, Java’s standard JavaBeans PropertyEditor lookup mechanism allows a PropertyEditor for a class simply to be named appropriately and placed in the same package as the class it provides support for, to be found automatically.

If there is a need to register other custom PropertyEditors, there are several mechanisms available. The most manual approach, which is not normally convenient or recommended, is to simply use the registerCustomEditor() method of the ConfigurableBeanFactory interface, assuming you have a BeanFactory reference. Another, slightly more convenient, mechanism is to use a special bean factory post-processor called CustomEditorConfigurer. Although bean factory post-processors can be used with BeanFactory implementations, the CustomEditorConfigurer has a nested property setup, so it is strongly recommended that it is used with theApplicationContext, where it may be deployed in similar fashion to any other bean, and automatically detected and applied.

Note that all bean factories and application contexts automatically use a number of built-in property editors, through their use of something called a BeanWrapper to handle property conversions. The standard property editors that the BeanWrapper registers are listed in the previous section. Additionally, ApplicationContexts also override or add an additional number of editors to handle resource lookups in a manner appropriate to the specific application context type.

Standard JavaBeans PropertyEditor instances are used to convert property values expressed as strings to the actual complex type of the property.CustomEditorConfigurer, a bean factory post-processor, may be used to conveniently add support for additional PropertyEditor instances to anApplicationContext.

Consider a user class ExoticType, and another class DependsOnExoticType which needs ExoticType set as a property:

package example;

public class ExoticType {

    private String name;

    public ExoticType(String name) {
        this.name = name;
    }
}

public class DependsOnExoticType {

    private ExoticType type;

    public void setType(ExoticType type) {
        this.type = type;
    }
}

When things are properly set up, we want to be able to assign the type property as a string, which a PropertyEditor will behind the scenes convert into an actualExoticType instance:

<bean id="sample" class="example.DependsOnExoticType">
    <property name="type" value="aNameForExoticType"/>
</bean>

The PropertyEditor implementation could look similar to this:

// converts string representation to ExoticType object
package example;

public class ExoticTypeEditor extends PropertyEditorSupport {

    public void setAsText(String text) {
        setValue(new ExoticType(text.toUpperCase()));
    }
}

Finally, we use CustomEditorConfigurer to register the new PropertyEditor with the ApplicationContext, which will then be able to use it as needed:

<bean class="org.springframework.beans.factory.config.CustomEditorConfigurer">
    <property name="customEditors">
        <map>
            <entry key="example.ExoticType" value="example.ExoticTypeEditor"/>
        </map>
    </property>
</bean>
Using PropertyEditorRegistrars

Another mechanism for registering property editors with the Spring container is to create and use a PropertyEditorRegistrar. This interface is particularly useful when you need to use the same set of property editors in several different situations: write a corresponding registrar and reuse that in each case.PropertyEditorRegistrars work in conjunction with an interface called PropertyEditorRegistry, an interface that is implemented by the Spring BeanWrapper(and DataBinder). PropertyEditorRegistrars are particularly convenient when used in conjunction with the CustomEditorConfigurer (introduced here), which exposes a property called setPropertyEditorRegistrars(..): PropertyEditorRegistrars added to a CustomEditorConfigurer in this fashion can easily be shared with DataBinder and Spring MVC Controllers. Furthermore, it avoids the need for synchronization on custom editors: a PropertyEditorRegistrar is expected to create fresh PropertyEditor instances for each bean creation attempt.

Using a PropertyEditorRegistrar is perhaps best illustrated with an example. First off, you need to create your own PropertyEditorRegistrar implementation:

package com.foo.editors.spring;

public final class CustomPropertyEditorRegistrar implements PropertyEditorRegistrar {

    public void registerCustomEditors(PropertyEditorRegistry registry) {

        // it is expected that new PropertyEditor instances are created
        registry.registerCustomEditor(ExoticType.class, new ExoticTypeEditor());

        // you could register as many custom property editors as are required here...
    }
}

See also the org.springframework.beans.support.ResourceEditorRegistrar for an example PropertyEditorRegistrar implementation. Notice how in its implementation of the registerCustomEditors(..) method it creates new instances of each property editor.

Next we configure a CustomEditorConfigurer and inject an instance of our CustomPropertyEditorRegistrar into it:

<bean class="org.springframework.beans.factory.config.CustomEditorConfigurer">
    <property name="propertyEditorRegistrars">
        <list>
            <ref bean="customPropertyEditorRegistrar"/>
        </list>
    </property>
</bean>

<bean id="customPropertyEditorRegistrar"
    class="com.foo.editors.spring.CustomPropertyEditorRegistrar"/>

Finally, and in a bit of a departure from the focus of this chapter, for those of you using Spring’s MVC web framework, using PropertyEditorRegistrars in conjunction with data-binding Controllers (such as SimpleFormController) can be very convenient. Find below an example of using a PropertyEditorRegistrar in the implementation of an initBinder(..) method:

public final class RegisterUserController extends SimpleFormController {

    private final PropertyEditorRegistrar customPropertyEditorRegistrar;

    public RegisterUserController(PropertyEditorRegistrar propertyEditorRegistrar) {
        this.customPropertyEditorRegistrar = propertyEditorRegistrar;
    }

    protected void initBinder(HttpServletRequest request,
            ServletRequestDataBinder binder) throws Exception {
        this.customPropertyEditorRegistrar.registerCustomEditors(binder);
    }

    // other methods to do with registering a User
}

This style of PropertyEditor registration can lead to concise code (the implementation of initBinder(..) is just one line long!), and allows commonPropertyEditor registration code to be encapsulated in a class and then shared amongst as many Controllers as needed.

9.5 Spring Type Conversion

Spring 3 introduces a core.convert package that provides a general type conversion system. The system defines an SPI to implement type conversion logic, as well as an API to execute type conversions at runtime. Within a Spring container, this system can be used as an alternative to PropertyEditors to convert externalized bean property value strings to required property types. The public API may also be used anywhere in your application where type conversion is needed.

9.5.1 Converter SPI

The SPI to implement type conversion logic is simple and strongly typed:

package org.springframework.core.convert.converter;

public interface Converter<S, T> {

    T convert(S source);

}

To create your own converter, simply implement the interface above. Parameterize S as the type you are converting from, and T as the type you are converting to. Such a converter can also be applied transparently if a collection or array of S needs to be converted to an array or collection of T, provided that a delegating array/collection converter has been registered as well (which DefaultConversionService does by default).

For each call to convert(S), the source argument is guaranteed to be NOT null. Your Converter may throw any unchecked exception if conversion fails; specifically, anIllegalArgumentException should be thrown to report an invalid source value. Take care to ensure that your Converter implementation is thread-safe.

Several converter implementations are provided in the core.convert.support package as a convenience. These include converters from Strings to Numbers and other common types. Consider StringToInteger as an example for a typical Converter implementation:

package org.springframework.core.convert.support;

final class StringToInteger implements Converter<String, Integer> {

    public Integer convert(String source) {
        return Integer.valueOf(source);
    }

}

9.5.2 ConverterFactory

When you need to centralize the conversion logic for an entire class hierarchy, for example, when converting from String to java.lang.Enum objects, implementConverterFactory:

package org.springframework.core.convert.converter;

public interface ConverterFactory<S, R> {

    <T extends R> Converter<S, T> getConverter(Class<T> targetType);

}

Parameterize S to be the type you are converting from and R to be the base type defining the range of classes you can convert to. Then implement getConverter(Class<T>), where T is a subclass of R.

Consider the StringToEnum ConverterFactory as an example:

package org.springframework.core.convert.support;

final class StringToEnumConverterFactory implements ConverterFactory<String, Enum> {

    public <T extends Enum> Converter<String, T> getConverter(Class<T> targetType) {
        return new StringToEnumConverter(targetType);
    }

    private final class StringToEnumConverter<T extends Enum> implements Converter<String, T> {

        private Class<T> enumType;

        public StringToEnumConverter(Class<T> enumType) {
            this.enumType = enumType;
        }

        public T convert(String source) {
            return (T) Enum.valueOf(this.enumType, source.trim());
        }
    }
}

9.5.3 GenericConverter

When you require a sophisticated Converter implementation, consider the GenericConverter interface. With a more flexible but less strongly typed signature, a GenericConverter supports converting between multiple source and target types. In addition, a GenericConverter makes available source and target field context you can use when implementing your conversion logic. Such context allows a type conversion to be driven by a field annotation, or generic information declared on a field signature.

package org.springframework.core.convert.converter;

public interface GenericConverter {

    public Set<ConvertiblePair> getConvertibleTypes();

    Object convert(Object source, TypeDescriptor sourceType, TypeDescriptor targetType);

}

To implement a GenericConverter, have getConvertibleTypes() return the supported source→target type pairs. Then implement convert(Object, TypeDescriptor, TypeDescriptor) to implement your conversion logic. The source TypeDescriptor provides access to the source field holding the value being converted. The target TypeDescriptor provides access to the target field where the converted value will be set.

A good example of a GenericConverter is a converter that converts between a Java Array and a Collection. Such an ArrayToCollectionConverter introspects the field that declares the target Collection type to resolve the Collection’s element type. This allows each element in the source array to be converted to the Collection element type before the Collection is set on the target field.

Spring框架参考文档 Spring Framework Reference Documentation

Because GenericConverter is a more complex SPI interface, only use it when you need it. Favor Converter or ConverterFactory for basic type conversion needs.

ConditionalGenericConverter

Sometimes you only want a Converter to execute if a specific condition holds true. For example, you might only want to execute a Converter if a specific annotation is present on the target field. Or you might only want to execute a Converter if a specific method, such as a static valueOf method, is defined on the target class.ConditionalGenericConverter is the union of the GenericConverter and ConditionalConverter interfaces that allows you to define such custom matching criteria:

public interface ConditionalGenericConverter
        extends GenericConverter, ConditionalConverter {

    boolean matches(TypeDescriptor sourceType, TypeDescriptor targetType);

}

A good example of a ConditionalGenericConverter is an EntityConverter that converts between an persistent entity identifier and an entity reference. Such a EntityConverter might only match if the target entity type declares a static finder method e.g. findAccount(Long). You would perform such a finder method check in the implementation of matches(TypeDescriptor, TypeDescriptor).

9.5.4 ConversionService API

The ConversionService defines a unified API for executing type conversion logic at runtime. Converters are often executed behind this facade interface:

package org.springframework.core.convert;

public interface ConversionService {

    boolean canConvert(Class<?> sourceType, Class<?> targetType);

    <T> T convert(Object source, Class<T> targetType);

    boolean canConvert(TypeDescriptor sourceType, TypeDescriptor targetType);

    Object convert(Object source, TypeDescriptor sourceType, TypeDescriptor targetType);

}

Most ConversionService implementations also implement ConverterRegistry, which provides an SPI for registering converters. Internally, a ConversionService implementation delegates to its registered converters to carry out type conversion logic.

A robust ConversionService implementation is provided in the core.convert.support package. GenericConversionService is the general-purpose implementation suitable for use in most environments. ConversionServiceFactory provides a convenient factory for creating common ConversionService configurations.

9.5.5 Configuring a ConversionService

A ConversionService is a stateless object designed to be instantiated at application startup, then shared between multiple threads. In a Spring application, you typically configure a ConversionService instance per Spring container (or ApplicationContext). That ConversionService will be picked up by Spring and then used whenever a type conversion needs to be performed by the framework. You may also inject this ConversionService into any of your beans and invoke it directly.

Spring框架参考文档 Spring Framework Reference Documentation

If no ConversionService is registered with Spring, the original PropertyEditor-based system is used.

To register a default ConversionService with Spring, add the following bean definition with id conversionService:

<bean id="conversionService"
    class="org.springframework.context.support.ConversionServiceFactoryBean"/>

A default ConversionService can convert between strings, numbers, enums, collections, maps, and other common types. To supplement or override the default converters with your own custom converter(s), set the converters property. Property values may implement either of the Converter, ConverterFactory, or GenericConverter interfaces.

<bean id="conversionService"
        class="org.springframework.context.support.ConversionServiceFactoryBean">
    <property name="converters">
        <set>
            <bean class="example.MyCustomConverter"/>
        </set>
    </property>
</bean>

It is also common to use a ConversionService within a Spring MVC application. See Section 22.16.3, “Conversion and Formatting” in the Spring MVC chapter.

In certain situations you may wish to apply formatting during conversion. See Section 9.6.3, “FormatterRegistry SPI” for details on usingFormattingConversionServiceFactoryBean.

9.5.6 Using a ConversionService programmatically

To work with a ConversionService instance programmatically, simply inject a reference to it like you would for any other bean:

@Service
public class MyService {

    @Autowired
    public MyService(ConversionService conversionService) {
        this.conversionService = conversionService;
    }

    public void doIt() {
        this.conversionService.convert(...)
    }
}

For most use cases, the convert method specifying the targetType can be used but it will not work with more complex types such as a collection of a parameterized element. If you want to convert a List of Integer to a List of String programmatically, for instance, you need to provide a formal definition of the source and target types.

Fortunately, TypeDescriptor provides various options to make that straightforward:

DefaultConversionService cs = new DefaultConversionService();

List<Integer> input = ....
cs.convert(input,
    TypeDescriptor.forObject(input), // List<Integer> type descriptor
    TypeDescriptor.collection(List.class, TypeDescriptor.valueOf(String.class)));

Note that DefaultConversionService registers converters automatically which are appropriate for most environments. This includes collection converters, scalar converters, and also basic Object to String converters. The same converters can be registered with any ConverterRegistry using the staticaddDefaultConverters method on the DefaultConversionService class.

Converters for value types will be reused for arrays and collections, so there is no need to create a specific converter to convert from a Collection of S to aCollection of T, assuming that standard collection handling is appropriate.

9.6 Spring Field Formatting

As discussed in the previous section, core.convert is a general-purpose type conversion system. It provides a unified ConversionService API as well as a strongly-typed Converter SPI for implementing conversion logic from one type to another. A Spring Container uses this system to bind bean property values. In addition, both the Spring Expression Language (SpEL) and DataBinder use this system to bind field values. For example, when SpEL needs to coerce a Short to a Long to complete anexpression.setValue(Object bean, Object value) attempt, the core.convert system performs the coercion.

Now consider the type conversion requirements of a typical client environment such as a web or desktop application. In such environments, you typically convert from String to support the client postback process, as well as back to String to support the view rendering process. In addition, you often need to localize String values. The more general core.convert Converter SPI does not address such formatting requirements directly. To directly address them, Spring 3 introduces a convenient Formatter SPI that provides a simple and robust alternative to PropertyEditors for client environments.

In general, use the Converter SPI when you need to implement general-purpose type conversion logic; for example, for converting between a java.util.Date and and java.lang.Long. Use the Formatter SPI when you’re working in a client environment, such as a web application, and need to parse and print localized field values. The ConversionService provides a unified type conversion API for both SPIs.

9.6.1 Formatter SPI

The Formatter SPI to implement field formatting logic is simple and strongly typed:

package org.springframework.format;

public interface Formatter<T> extends Printer<T>, Parser<T> {
}

Where Formatter extends from the Printer and Parser building-block interfaces:

public interface Printer<T> {
    String print(T fieldValue, Locale locale);
}
import java.text.ParseException;

public interface Parser<T> {
    T parse(String clientValue, Locale locale) throws ParseException;
}

To create your own Formatter, simply implement the Formatter interface above. Parameterize T to be the type of object you wish to format, for example,java.util.Date. Implement the print() operation to print an instance of T for display in the client locale. Implement the parse() operation to parse an instance of T from the formatted representation returned from the client locale. Your Formatter should throw a ParseException or IllegalArgumentException if a parse attempt fails. Take care to ensure your Formatter implementation is thread-safe.

Several Formatter implementations are provided in format subpackages as a convenience. The number package provides a NumberFormatter,CurrencyFormatter, and PercentFormatter to format java.lang.Number objects using a java.text.NumberFormat. The datetime package provides aDateFormatter to format java.util.Date objects with a java.text.DateFormat. The datetime.joda package provides comprehensive datetime formatting support based on the Joda Time library.

Consider DateFormatter as an example Formatter implementation:

package org.springframework.format.datetime;

public final class DateFormatter implements Formatter<Date> {

    private String pattern;

    public DateFormatter(String pattern) {
        this.pattern = pattern;
    }

    public String print(Date date, Locale locale) {
        if (date == null) {
            return "";
        }
        return getDateFormat(locale).format(date);
    }

    public Date parse(String formatted, Locale locale) throws ParseException {
        if (formatted.length() == 0) {
            return null;
        }
        return getDateFormat(locale).parse(formatted);
    }

    protected DateFormat getDateFormat(Locale locale) {
        DateFormat dateFormat = new SimpleDateFormat(this.pattern, locale);
        dateFormat.setLenient(false);
        return dateFormat;
    }

}

The Spring team welcomes community-driven Formatter contributions; see jira.spring.io to contribute.

9.6.2 Annotation-driven Formatting

As you will see, field formatting can be configured by field type or annotation. To bind an Annotation to a formatter, implement AnnotationFormatterFactory:

package org.springframework.format;

public interface AnnotationFormatterFactory<A extends Annotation> {

    Set<Class<?>> getFieldTypes();

    Printer<?> getPrinter(A annotation, Class<?> fieldType);

    Parser<?> getParser(A annotation, Class<?> fieldType);

}

Parameterize A to be the field annotationType you wish to associate formatting logic with, for example org.springframework.format.annotation.DateTimeFormat. Have getFieldTypes() return the types of fields the annotation may be used on. Have getPrinter() return a Printer to print the value of an annotated field. HavegetParser() return a Parser to parse a clientValue for an annotated field.

The example AnnotationFormatterFactory implementation below binds the @NumberFormat Annotation to a formatter. This annotation allows either a number style or pattern to be specified:

public final class NumberFormatAnnotationFormatterFactory
        implements AnnotationFormatterFactory<NumberFormat> {

    public Set<Class<?>> getFieldTypes() {
        return new HashSet<Class<?>>(asList(new Class<?>[] {
            Short.class, Integer.class, Long.class, Float.class,
            Double.class, BigDecimal.class, BigInteger.class }));
    }

    public Printer<Number> getPrinter(NumberFormat annotation, Class<?> fieldType) {
        return configureFormatterFrom(annotation, fieldType);
    }

    public Parser<Number> getParser(NumberFormat annotation, Class<?> fieldType) {
        return configureFormatterFrom(annotation, fieldType);
    }

    private Formatter<Number> configureFormatterFrom(NumberFormat annotation,
            Class<?> fieldType) {
        if (!annotation.pattern().isEmpty()) {
            return new NumberFormatter(annotation.pattern());
        } else {
            Style style = annotation.style();
            if (style == Style.PERCENT) {
                return new PercentFormatter();
            } else if (style == Style.CURRENCY) {
                return new CurrencyFormatter();
            } else {
                return new NumberFormatter();
            }
        }
    }
}

To trigger formatting, simply annotate fields with @NumberFormat:

public class MyModel {

    @NumberFormat(style=Style.CURRENCY)
    private BigDecimal decimal;

}

Format Annotation API

A portable format annotation API exists in the org.springframework.format.annotation package. Use @NumberFormat to format java.lang.Number fields. Use @DateTimeFormat to format java.util.Date, java.util.Calendar, java.util.Long, or Joda Time fields.

The example below uses @DateTimeFormat to format a java.util.Date as a ISO Date (yyyy-MM-dd):

public class MyModel {

    @DateTimeFormat(iso=ISO.DATE)
    private Date date;

}

9.6.3 FormatterRegistry SPI

The FormatterRegistry is an SPI for registering formatters and converters. FormattingConversionService is an implementation of FormatterRegistry suitable for most environments. This implementation may be configured programmatically or declaratively as a Spring bean using FormattingConversionServiceFactoryBean. Because this implementation also implements ConversionService, it can be directly configured for use with Spring’s DataBinder and the Spring Expression Language (SpEL).

Review the FormatterRegistry SPI below:

package org.springframework.format;

public interface FormatterRegistry extends ConverterRegistry {

    void addFormatterForFieldType(Class<?> fieldType, Printer<?> printer, Parser<?> parser);

    void addFormatterForFieldType(Class<?> fieldType, Formatter<?> formatter);

    void addFormatterForFieldType(Formatter<?> formatter);

    void addFormatterForAnnotation(AnnotationFormatterFactory<?, ?> factory);

}

As shown above, Formatters can be registered by fieldType or annotation.

The FormatterRegistry SPI allows you to configure Formatting rules centrally, instead of duplicating such configuration across your Controllers. For example, you might want to enforce that all Date fields are formatted a certain way, or fields with a specific annotation are formatted in a certain way. With a shared FormatterRegistry, you define these rules once and they are applied whenever formatting is needed.

9.6.4 FormatterRegistrar SPI

The FormatterRegistrar is an SPI for registering formatters and converters through the FormatterRegistry:

package org.springframework.format;

public interface FormatterRegistrar {

    void registerFormatters(FormatterRegistry registry);

}

A FormatterRegistrar is useful when registering multiple related converters and formatters for a given formatting category, such as Date formatting. It can also be useful where declarative registration is insufficient. For example when a formatter needs to be indexed under a specific field type different from its own <T> or when registering a Printer/Parser pair. The next section provides more information on converter and formatter registration.

9.6.5 Configuring Formatting in Spring MVC

See Section 22.16.3, “Conversion and Formatting” in the Spring MVC chapter.

9.7 Configuring a global date & time format

By default, date and time fields that are not annotated with @DateTimeFormat are converted from strings using the DateFormat.SHORT style. If you prefer, you can change this by defining your own global format.

You will need to ensure that Spring does not register default formatters, and instead you should register all formatters manually. Use theorg.springframework.format.datetime.joda.JodaTimeFormatterRegistrar or org.springframework.format.datetime.DateFormatterRegistrar class depending on whether you use the Joda Time library.

For example, the following Java configuration will register a global ' `yyyyMMdd’ format. This example does not depend on the Joda Time library:

@Configuration
public class AppConfig {

    @Bean
    public FormattingConversionService conversionService() {

        // Use the DefaultFormattingConversionService but do not register defaults
        DefaultFormattingConversionService conversionService = new DefaultFormattingConversionService(false);

        // Ensure @NumberFormat is still supported
        conversionService.addFormatterForFieldAnnotation(new NumberFormatAnnotationFormatterFactory());

        // Register date conversion with a specific global format
        DateFormatterRegistrar registrar = new DateFormatterRegistrar();
        registrar.setFormatter(new DateFormatter("yyyyMMdd"));
        registrar.registerFormatters(conversionService);

        return conversionService;
    }
}

If you prefer XML based configuration you can use a FormattingConversionServiceFactoryBean. Here is the same example, this time using Joda Time:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="
        http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd>

    <bean >>
        <property name="registerDefaultFormatters" value="false" />
        <property name="formatters">
            <set>
                <bean class="org.springframework.format.number.NumberFormatAnnotationFormatterFactory" />
            </set>
        </property>
        <property name="formatterRegistrars">
            <set>
                <bean class="org.springframework.format.datetime.joda.JodaTimeFormatterRegistrar">
                    <property name="dateFormatter">
                        <bean class="org.springframework.format.datetime.joda.DateTimeFormatterFactoryBean">
                            <property name="pattern" value="yyyyMMdd"/>
                        </bean>
                    </property>
                </bean>
            </set>
        </property>
    </bean>
</beans>
Spring框架参考文档 Spring Framework Reference Documentation

Joda Time provides separate distinct types to represent date, time and date-time values. The dateFormatter, timeFormatter anddateTimeFormatter properties of the JodaTimeFormatterRegistrar should be used to configure the different formats for each type. TheDateTimeFormatterFactoryBean provides a convenient way to create formatters.

If you are using Spring MVC remember to explicitly configure the conversion service that is used. For Java based @Configuration this means extending theWebMvcConfigurationSupport class and overriding the mvcConversionService() method. For XML you should use the 'conversion-service' attribute of themvc:annotation-driven element. See Section 22.16.3, “Conversion and Formatting” for details.

9.8 Spring Validation

Spring 3 introduces several enhancements to its validation support. First, the JSR-303 Bean Validation API is now fully supported. Second, when used programmatically, Spring’s DataBinder can now validate objects as well as bind to them. Third, Spring MVC now has support for declaratively validating @Controller inputs.

9.8.1 Overview of the JSR-303 Bean Validation API

JSR-303 standardizes validation constraint declaration and metadata for the Java platform. Using this API, you annotate domain model properties with declarative validation constraints and the runtime enforces them. There are a number of built-in constraints you can take advantage of. You may also define your own custom constraints.

To illustrate, consider a simple PersonForm model with two properties:

public class PersonForm {
    private String name;
    private int age;
}

JSR-303 allows you to define declarative validation constraints against such properties:

public class PersonForm {

    @NotNull
    @Size(max=64)
    private String name;

    @Min(0)
    private int age;

}

When an instance of this class is validated by a JSR-303 Validator, these constraints will be enforced.

For general information on JSR-303/JSR-349, see the Bean Validation website. For information on the specific capabilities of the default reference implementation, see the Hibernate Validator documentation. To learn how to setup a Bean Validation provider as a Spring bean, keep reading.

9.8.2 Configuring a Bean Validation Provider

Spring provides full support for the Bean Validation API. This includes convenient support for bootstrapping a JSR-303/JSR-349 Bean Validation provider as a Spring bean. This allows for a javax.validation.ValidatorFactory or javax.validation.Validator to be injected wherever validation is needed in your application.

Use the LocalValidatorFactoryBean to configure a default Validator as a Spring bean:

<bean id="validator"
    class="org.springframework.validation.beanvalidation.LocalValidatorFactoryBean"/>

The basic configuration above will trigger Bean Validation to initialize using its default bootstrap mechanism. A JSR-303/JSR-349 provider, such as Hibernate Validator, is expected to be present in the classpath and will be detected automatically.

Injecting a Validator

LocalValidatorFactoryBean implements both javax.validation.ValidatorFactory and javax.validation.Validator, as well as Spring’sorg.springframework.validation.Validator. You may inject a reference to either of these interfaces into beans that need to invoke validation logic.

Inject a reference to javax.validation.Validator if you prefer to work with the Bean Validation API directly:

import javax.validation.Validator;

@Service
public class MyService {

    @Autowired
    private Validator validator;

Inject a reference to org.springframework.validation.Validator if your bean requires the Spring Validation API:

import org.springframework.validation.Validator;

@Service
public class MyService {

    @Autowired
    private Validator validator;

}

Configuring Custom Constraints

Each Bean Validation constraint consists of two parts. First, a @Constraint annotation that declares the constraint and its configurable properties. Second, an implementation of the javax.validation.ConstraintValidator interface that implements the constraint’s behavior. To associate a declaration with an implementation, each @Constraint annotation references a corresponding ValidationConstraint implementation class. At runtime, a ConstraintValidatorFactoryinstantiates the referenced implementation when the constraint annotation is encountered in your domain model.

By default, the LocalValidatorFactoryBean configures a SpringConstraintValidatorFactory that uses Spring to create ConstraintValidator instances. This allows your custom ConstraintValidators to benefit from dependency injection like any other Spring bean.

Shown below is an example of a custom @Constraint declaration, followed by an associated ConstraintValidator implementation that uses Spring for dependency injection:

@Target({ElementType.METHOD, ElementType.FIELD})
@Retention(RetentionPolicy.RUNTIME)
@Constraint(validatedBy=MyConstraintValidator.class)
public @interface MyConstraint {
}
import javax.validation.ConstraintValidator;

public class MyConstraintValidator implements ConstraintValidator {

    @Autowired;
    private Foo aDependency;

    ...
}

As you can see, a ConstraintValidator implementation may have its dependencies @Autowired like any other Spring bean.

Spring-driven Method Validation

The method validation feature supported by Bean Validation 1.1, and as a custom extension also by Hibernate Validator 4.3, can be integrated into a Spring context through a MethodValidationPostProcessor bean definition:

<bean class="org.springframework.validation.beanvalidation.MethodValidationPostProcessor"/>

In order to be eligible for Spring-driven method validation, all target classes need to be annotated with Spring’s @Validated annotation, optionally declaring the validation groups to use. Check out the MethodValidationPostProcessor javadocs for setup details with Hibernate Validator and Bean Validation 1.1 providers.

Additional Configuration Options

The default LocalValidatorFactoryBean configuration should prove sufficient for most cases. There are a number of configuration options for various Bean Validation constructs, from message interpolation to traversal resolution. See the LocalValidatorFactoryBean javadocs for more information on these options.

9.8.3 Configuring a DataBinder

Since Spring 3, a DataBinder instance can be configured with a Validator. Once configured, the Validator may be invoked by calling binder.validate(). Any validation Errors are automatically added to the binder’s BindingResult.

When working with the DataBinder programmatically, this can be used to invoke validation logic after binding to a target object:

Foo target = new Foo();
DataBinder binder = new DataBinder(target);
binder.setValidator(new FooValidator());

// bind to the target object
binder.bind(propertyValues);

// validate the target object
binder.validate();

// get BindingResult that includes any validation errors
BindingResult results = binder.getBindingResult();

A DataBinder can also be configured with multiple Validator instances via dataBinder.addValidators and dataBinder.replaceValidators. This is useful when combining globally configured Bean Validation with a Spring Validator configured locally on a DataBinder instance. See ???.

9.8.4 Spring MVC 3 Validation

See Section 22.16.4, “Validation” in the Spring MVC chapter.

10. Spring Expression Language (SpEL)

10.1 Introduction

The Spring Expression Language (SpEL for short) is a powerful expression language that supports querying and manipulating an object graph at runtime. The language syntax is similar to Unified EL but offers additional features, most notably method invocation and basic string templating functionality.

While there are several other Java expression languages available, OGNL, MVEL, and JBoss EL, to name a few, the Spring Expression Language was created to provide the Spring community with a single well supported expression language that can be used across all the products in the Spring portfolio. Its language features are driven by the requirements of the projects in the Spring portfolio, including tooling requirements for code completion support within the eclipse based Spring Tool Suite. That said, SpEL is based on a technology agnostic API allowing other expression language implementations to be integrated should the need arise.

While SpEL serves as the foundation for expression evaluation within the Spring portfolio, it is not directly tied to Spring and can be used independently. In order to be self contained, many of the examples in this chapter use SpEL as if it were an independent expression language. This requires creating a few bootstrapping infrastructure classes such as the parser. Most Spring users will not need to deal with this infrastructure and will instead only author expression strings for evaluation. An example of this typical use is the integration of SpEL into creating XML or annotated based bean definitions as shown in the section Expression support for defining bean definitions.

This chapter covers the features of the expression language, its API, and its language syntax. In several places an Inventor and Inventor’s Society class are used as the target objects for expression evaluation. These class declarations and the data used to populate them are listed at the end of the chapter.

10.2 Feature Overview

The expression language supports the following functionality

  • Literal expressions
  • Boolean and relational operators
  • Regular expressions
  • Class expressions
  • Accessing properties, arrays, lists, maps
  • Method invocation
  • Relational operators
  • Assignment
  • Calling constructors
  • Bean references
  • Array construction
  • Inline lists
  • Inline maps
  • Ternary operator
  • Variables
  • User defined functions
  • Collection projection
  • Collection selection
  • Templated expressions

10.3 Expression Evaluation using Spring’s Expression Interface

This section introduces the simple use of SpEL interfaces and its expression language. The complete language reference can be found in the section Language Reference.

The following code introduces the SpEL API to evaluate the literal string expression 'Hello World'.

ExpressionParser parser = new SpelExpressionParser();
Expression exp = parser.parseExpression("'Hello World'");
String message = (String) exp.getValue();

The value of the message variable is simply 'Hello World'.

The SpEL classes and interfaces you are most likely to use are located in the packages org.springframework.expression and its sub packages andspel.support.

The interface ExpressionParser is responsible for parsing an expression string. In this example the expression string is a string literal denoted by the surrounding single quotes. The interface Expression is responsible for evaluating the previously defined expression string. There are two exceptions that can be thrown,ParseException and EvaluationException when calling parser.parseExpression and exp.getValue respectively.

SpEL supports a wide range of features, such as calling methods, accessing properties, and calling constructors.

As an example of method invocation, we call the concat method on the string literal.

ExpressionParser parser = new SpelExpressionParser();
Expression exp = parser.parseExpression("'Hello World'.concat('!')");
String message = (String) exp.getValue();

The value of message is now 'Hello World!'.

As an example of calling a JavaBean property, the String property Bytes can be called as shown below.

ExpressionParser parser = new SpelExpressionParser();

// invokes 'getBytes()'
Expression exp = parser.parseExpression("'Hello World'.bytes");
byte[] bytes = (byte[]) exp.getValue();

SpEL also supports nested properties using standard dot notation, i.e. prop1.prop2.prop3 and the setting of property values

Public fields may also be accessed.

ExpressionParser parser = new SpelExpressionParser();

// invokes 'getBytes().length'
Expression exp = parser.parseExpression("'Hello World'.bytes.length");
int length = (Integer) exp.getValue();

The String’s constructor can be called instead of using a string literal.

ExpressionParser parser = new SpelExpressionParser();
Expression exp = parser.parseExpression("new String('hello world').toUpperCase()");
String message = exp.getValue(String.class);

Note the use of the generic method public <T> T getValue(Class<T> desiredResultType). Using this method removes the need to cast the value of the expression to the desired result type. An EvaluationException will be thrown if the value cannot be cast to the type T or converted using the registered type converter.

The more common usage of SpEL is to provide an expression string that is evaluated against a specific object instance (called the root object). There are two options here and which to choose depends on whether the object against which the expression is being evaluated will be changing with each call to evaluate the expression. In the following example we retrieve the name property from an instance of the Inventor class.

// Create and set a calendar
GregorianCalendar c = new GregorianCalendar();
c.set(1856, 7, 9);

// The constructor arguments are name, birthday, and nationality.
Inventor tesla = new Inventor("Nikola Tesla", c.getTime(), "Serbian");

ExpressionParser parser = new SpelExpressionParser();
Expression exp = parser.parseExpression("name");

EvaluationContext context = new StandardEvaluationContext(tesla);
String name = (String) exp.getValue(context);

In the last line, the value of the string variable name will be set to "Nikola Tesla". The class StandardEvaluationContext is where you can specify which object the "name" property will be evaluated against. This is the mechanism to use if the root object is unlikely to change, it can simply be set once in the evaluation context. If the root object is likely to change repeatedly, it can be supplied on each call to getValue, as this next example shows:

/ Create and set a calendar
GregorianCalendar c = new GregorianCalendar();
c.set(1856, 7, 9);

// The constructor arguments are name, birthday, and nationality.
Inventor tesla = new Inventor("Nikola Tesla", c.getTime(), "Serbian");

ExpressionParser parser = new SpelExpressionParser();
Expression exp = parser.parseExpression("name");
String name = (String) exp.getValue(tesla);

In this case the inventor tesla has been supplied directly to getValue and the expression evaluation infrastructure creates and manages a default evaluation context internally - it did not require one to be supplied.

The StandardEvaluationContext is relatively expensive to construct and during repeated usage it builds up cached state that enables subsequent expression evaluations to be performed more quickly. For this reason it is better to cache and reuse them where possible, rather than construct a new one for each expression evaluation.

In some cases it can be desirable to use a configured evaluation context and yet still supply a different root object on each call to getValue. getValue allows both to be specified on the same call. In these situations the root object passed on the call is considered to override any (which maybe null) specified on the evaluation context.

Spring框架参考文档 Spring Framework Reference Documentation

In standalone usage of SpEL there is a need to create the parser, parse expressions and perhaps provide evaluation contexts and a root context object. However, more common usage is to provide only the SpEL expression string as part of a configuration file, for example for Spring bean or Spring Web Flow definitions. In this case, the parser, evaluation context, root object and any predefined variables are all set up implicitly, requiring the user to specify nothing other than the expressions.

As a final introductory example, the use of a boolean operator is shown using the Inventor object in the previous example.

Expression exp = parser.parseExpression("name == 'Nikola Tesla'");
boolean result = exp.getValue(context, Boolean.class); // evaluates to true

10.3.1 The EvaluationContext interface

The interface EvaluationContext is used when evaluating an expression to resolve properties, methods, fields, and to help perform type conversion. The out-of-the-box implementation, StandardEvaluationContext, uses reflection to manipulate the object, caching java.lang.reflect.Method, java.lang.reflect.Field, and java.lang.reflect.Constructor instances for increased performance.

The StandardEvaluationContext is where you may specify the root object to evaluate against via the method setRootObject() or passing the root object into the constructor. You can also specify variables and functions that will be used in the expression using the methods setVariable() and registerFunction(). The use of variables and functions are described in the language reference sections Variables and Functions. The StandardEvaluationContext is also where you can register custom ConstructorResolvers, MethodResolvers, and PropertyAccessors to extend how SpEL evaluates expressions. Please refer to the JavaDoc of these classes for more details.

Type Conversion

By default SpEL uses the conversion service available in Spring core ( org.springframework.core.convert.ConversionService). This conversion service comes with many converters built in for common conversions but is also fully extensible so custom conversions between types can be added. Additionally it has the key capability that it is generics aware. This means that when working with generic types in expressions, SpEL will attempt conversions to maintain type correctness for any objects it encounters.

What does this mean in practice? Suppose assignment, using setValue(), is being used to set a List property. The type of the property is actually List<Boolean>. SpEL will recognize that the elements of the list need to be converted to Boolean before being placed in it. A simple example:

class Simple {
    public List<Boolean> booleanList = new ArrayList<Boolean>();
}

Simple simple = new Simple();

simple.booleanList.add(true);

StandardEvaluationContext simpleContext = new StandardEvaluationContext(simple);

// false is passed in here as a string. SpEL and the conversion service will
// correctly recognize that it needs to be a Boolean and convert it
parser.parseExpression("booleanList[0]").setValue(simpleContext, "false");

// b will be false
Boolean b = simple.booleanList.get(0);

10.3.2 Parser configuration

It is possible to configure the SpEL expression parser using a parser configuration object (org.springframework.expression.spel.SpelParserConfiguration). The configuration object controls the behavior of some of the expression components. For example, if indexing into an array or collection and the element at the specified index is null it is possible to automatically create the element. This is useful when using expressions made up of a chain of property references. If indexing into an array or list and specifying an index that is beyond the end of the current size of the array or list it is possible to automatically grow the array or list to accommodate that index.

class Demo {
    public List<String> list;
}

// Turn on:
// - auto null reference initialization
// - auto collection growing
SpelParserConfiguration config = new SpelParserConfiguration(true,true);

ExpressionParser parser = new SpelExpressionParser(config);

Expression expression = parser.parseExpression("list[3]");

Demo demo = new Demo();

Object o = expression.getValue(demo);

// demo.list will now be a real collection of 4 entries
// Each entry is a new empty String

It is also possible to configure the behaviour of the SpEL expression compiler.

10.3.3 SpEL compilation

Spring Framework 4.1 includes a basic expression compiler. Expressions are usually interpreted which provides a lot of dynamic flexibility during evaluation but does not provide the optimum performance. For occasional expression usage this is fine, but when used by other components like Spring Integration, performance can be very important and there is no real need for the dynamism.

The new SpEL compiler is intended to address this need. The compiler will generate a real Java class on the fly during evaluation that embodies the expression behavior and use that to achieve much faster expression evaluation. Due to the lack of typing around expressions the compiler uses information gathered during the interpreted evaluations of an expression when performing compilation. For example, it does not know the type of a property reference purely from the expression but during the first interpreted evaluation it will find out what it is. Of course, basing the compilation on this information could cause trouble later if the types of the various expression elements change over time. For this reason compilation is best suited to expressions whose type information is not going to change on repeated evaluations.

For a basic expression like this:

someArray[0].someProperty.someOtherProperty < 0.1

which involves array access, some property derefencing and numeric operations, the performance gain can be very noticeable. In an example micro benchmark run of 50000 iterations, it was taking 75ms to evaluate using only the interpreter and just 3ms using the compiled version of the expression.

Compiler configuration

The compiler is not turned on by default, but there are two ways to turn it on. It can be turned on using the parser configuration process discussed earlier or via a system property when SpEL usage is embedded inside another component. This section discusses both of these options.

Is is important to understand that there are a few modes the compiler can operate in, captured in an enum (org.springframework.expression.spel.SpelCompilerMode). The modes are as follows:

  • OFF - The compiler is switched off; this is the default.
  • IMMEDIATE - In immediate mode the expressions are compiled as soon as possible. This is typically after the first interpreted evaluation. If the compiled expression fails (typically due to a type changing, as described above) then the caller of the expression evaluation will receive an exception.
  • MIXED - In mixed mode the expressions silently switch between interpreted and compiled mode over time. After some number of interpreted runs they will switch to compiled form and if something goes wrong with the compiled form (like a type changing, as described above) then the expression will automatically switch back to interpreted form again. Sometime later it may generate another compiled form and switch to it. Basically the exception that the user gets in IMMEDIATE mode is instead handled internally.

IMMEDIATE mode exists because MIXED mode could cause issues for expressions that have side effects. If a compiled expression blows up after partially succeeding it may have already done something that has affected the state of the system. If this has happened the caller may not want it to silently re-run in interpreted mode since part of the expression may be running twice.

After selecting a mode, use the SpelParserConfiguration to configure the parser:

SpelParserConfiguration config = new SpelParserConfiguration(SpelCompilerMode.IMMEDIATE,
    this.getClass().getClassLoader());

SpelExpressionParser parser = new SpelExpressionParser(config);

Expression expr = parser.parseExpression("payload");

MyMessage message = new MyMessage();

Object payload = expr.getValue(message);

When specifying the compiler mode it is also possible to specify a classloader (passing null is allowed). Compiled expressions will be defined in a child classloader created under any that is supplied. It is important to ensure if a classloader is specified it can see all the types involved in the expression evaluation process. If none is specified then a default classloader will be used (typically the context classloader for the thread that is running during expression evaluation).

The second way to configure the compiler is for use when SpEL is embedded inside some other component and it may not be possible to configure via a configuration object. In these cases it is possible to use a system property. The property spring.expression.compiler.mode can be set to one of the SpelCompilerMode enum values (off, immediate, or mixed).

Compiler limitations

With Spring Framework 4.1 the basic compilation framework is in place. However, the framework does not yet support compiling every kind of expression. The initial focus has been on the common expressions that are likely to be used in performance critical contexts. These kinds of expression cannot be compiled at the moment:

  • expressions involving assignment
  • expressions relying on the conversion service
  • expressions using custom resolvers or accessors
  • expressions using selection or projection

More and more types of expression will be compilable in the future.

10.4 Expression support for defining bean definitions

SpEL expressions can be used with XML or annotation-based configuration metadata for defining BeanDefinitions. In both cases the syntax to define the expression is of the form #{ <expression string> }.

10.4.1 XML based configuration

A property or constructor-arg value can be set using expressions as shown below.

<bean id="numberGuess" class="org.spring.samples.NumberGuess">
    <property name="randomNumber" value="#{ T(java.lang.Math).random() * 100.0 }"/>

    <!-- other properties -->
</bean>

The variable systemProperties is predefined, so you can use it in your expressions as shown below. Note that you do not have to prefix the predefined variable with the # symbol in this context.

<bean id="taxCalculator" class="org.spring.samples.TaxCalculator">
    <property name="defaultLocale" value="#{ systemProperties['user.region'] }"/>

    <!-- other properties -->
</bean>

You can also refer to other bean properties by name, for example.

<bean id="numberGuess" class="org.spring.samples.NumberGuess">
    <property name="randomNumber" value="#{ T(java.lang.Math).random() * 100.0 }"/>

    <!-- other properties -->
</bean>

<bean id="shapeGuess" class="org.spring.samples.ShapeGuess">
    <property name="initialShapeSeed" value="#{ numberGuess.randomNumber }"/>

    <!-- other properties -->
</bean>

10.4.2 Annotation-based configuration

The @Value annotation can be placed on fields, methods and method/constructor parameters to specify a default value.

Here is an example to set the default value of a field variable.

public static class FieldValueTestBean

    @Value("#{ systemProperties['user.region'] }")
    private String defaultLocale;

    public void setDefaultLocale(String defaultLocale) {
        this.defaultLocale = defaultLocale;
    }

    public String getDefaultLocale() {
        return this.defaultLocale;
    }

}

The equivalent but on a property setter method is shown below.

public static class PropertyValueTestBean

    private String defaultLocale;

    @Value("#{ systemProperties['user.region'] }")
    public void setDefaultLocale(String defaultLocale) {
        this.defaultLocale = defaultLocale;
    }

    public String getDefaultLocale() {
        return this.defaultLocale;
    }

}

Autowired methods and constructors can also use the @Value annotation.

public class SimpleMovieLister {

    private MovieFinder movieFinder;
    private String defaultLocale;

    @Autowired
    public void configure(MovieFinder movieFinder,
            @Value("#{ systemProperties['user.region'] }") String defaultLocale) {
        this.movieFinder = movieFinder;
        this.defaultLocale = defaultLocale;
    }

    // ...
}
public class MovieRecommender {

    private String defaultLocale;

    private CustomerPreferenceDao customerPreferenceDao;

    @Autowired
    public MovieRecommender(CustomerPreferenceDao customerPreferenceDao,
            @Value("#{systemProperties['user.country']}") String defaultLocale) {
        this.customerPreferenceDao = customerPreferenceDao;
        this.defaultLocale = defaultLocale;
    }

    // ...
}

10.5 Language Reference

10.5.1 Literal expressions

The types of literal expressions supported are strings, dates, numeric values (int, real, and hex), boolean and null. Strings are delimited by single quotes. To put a single quote itself in a string use two single quote characters. The following listing shows simple usage of literals. Typically they would not be used in isolation like this, but as part of a more complex expression, for example using a literal on one side of a logical comparison operator.

ExpressionParser parser = new SpelExpressionParser();

// evals to "Hello World"
String helloWorld = (String) parser.parseExpression("'Hello World'").getValue();

double avogadrosNumber = (Double) parser.parseExpression("6.0221415E+23").getValue();

// evals to 2147483647
int maxValue = (Integer) parser.parseExpression("0x7FFFFFFF").getValue();

boolean trueValue = (Boolean) parser.parseExpression("true").getValue();

Object nullValue = parser.parseExpression("null").getValue();

Numbers support the use of the negative sign, exponential notation, and decimal points. By default real numbers are parsed using Double.parseDouble().

10.5.2 Properties, Arrays, Lists, Maps, Indexers

Navigating with property references is easy: just use a period to indicate a nested property value. The instances of the Inventor class, pupin, and tesla, were populated with data listed in the section Classes used in the examples. To navigate "down" and get Tesla’s year of birth and Pupin’s city of birth the following expressions are used.

// evals to 1856
int year = (Integer) parser.parseExpression("Birthdate.Year + 1900").getValue(context);

String city = (String) parser.parseExpression("placeOfBirth.City").getValue(context);

Case insensitivity is allowed for the first letter of property names. The contents of arrays and lists are obtained using square bracket notation.

ExpressionParser parser = new SpelExpressionParser();

// Inventions Array
StandardEvaluationContext teslaContext = new StandardEvaluationContext(tesla);

// evaluates to "Induction motor"
String invention = parser.parseExpression("inventions[3]").getValue(
        teslaContext, String.class);

// Members List
StandardEvaluationContext societyContext = new StandardEvaluationContext(ieee);

// evaluates to "Nikola Tesla"
String name = parser.parseExpression("Members[0].Name").getValue(
        societyContext, String.class);

// List and Array navigation
// evaluates to "Wireless communication"
String invention = parser.parseExpression("Members[0].Inventions[6]").getValue(
        societyContext, String.class);

The contents of maps are obtained by specifying the literal key value within the brackets. In this case, because keys for the Officers map are strings, we can specify string literals.

// Officer's Dictionary

Inventor pupin = parser.parseExpression("Officers['president']").getValue(
        societyContext, Inventor.class);

// evaluates to "Idvor"
String city = parser.parseExpression("Officers['president'].PlaceOfBirth.City").getValue(
        societyContext, String.class);

// setting values
parser.parseExpression("Officers['advisors'][0].PlaceOfBirth.Country").setValue(
        societyContext, "Croatia");

10.5.3 Inline lists

Lists can be expressed directly in an expression using {} notation.

// evaluates to a Java list containing the four numbers
List numbers = (List) parser.parseExpression("{1,2,3,4}").getValue(context);

List listOfLists = (List) parser.parseExpression("{{'a','b'},{'x','y'}}").getValue(context);

{} by itself means an empty list. For performance reasons, if the list is itself entirely composed of fixed literals then a constant list is created to represent the expression, rather than building a new list on each evaluation.

10.5.4 Inline Maps

Maps can also be expressed directly in an expression using {key:value} notation.

// evaluates to a Java map containing the two entries
Map inventorInfo = (Map) parser.parseExpression("{name:'Nikola',dob:'10-July-1856'}").getValue(context);

Map mapOfMaps = (Map) parser.parseExpression("{name:{first:'Nikola',last:'Tesla'},dob:{day:10,month:'July',year:1856}}").getValue(context);

{:} by itself means an empty map. For performance reasons, if the map is itself composed of fixed literals or other nested constant structures (lists or maps) then a constant map is created to represent the expression, rather than building a new map on each evaluation. Quoting of the map keys is optional, the examples above are not using quoted keys.

10.5.5 Array construction

Arrays can be built using the familiar Java syntax, optionally supplying an initializer to have the array populated at construction time.

int[] numbers1 = (int[]) parser.parseExpression("new int[4]").getValue(context);

// Array with initializer
int[] numbers2 = (int[]) parser.parseExpression("new int[]{1,2,3}").getValue(context);

// Multi dimensional array
int[][] numbers3 = (int[][]) parser.parseExpression("new int[4][5]").getValue(context);

It is not currently allowed to supply an initializer when constructing a multi-dimensional array.

10.5.6 Methods

Methods are invoked using typical Java programming syntax. You may also invoke methods on literals. Varargs are also supported.

// string literal, evaluates to "bc"
String c = parser.parseExpression("'abc'.substring(2, 3)").getValue(String.class);

// evaluates to true
boolean isMember = parser.parseExpression("isMember('Mihajlo Pupin')").getValue(
        societyContext, Boolean.class);

10.5.7 Operators

Relational operators

The relational operators; equal, not equal, less than, less than or equal, greater than, and greater than or equal are supported using standard operator notation.

// evaluates to true
boolean trueValue = parser.parseExpression("2 == 2").getValue(Boolean.class);

// evaluates to false
boolean falseValue = parser.parseExpression("2 < -5.0").getValue(Boolean.class);

// evaluates to true
boolean trueValue = parser.parseExpression("'black' < 'block'").getValue(Boolean.class);

In addition to standard relational operators SpEL supports the instanceof and regular expression based matches operator.

// evaluates to false
boolean falseValue = parser.parseExpression(
        "'xyz' instanceof T(Integer.class)").getValue(Boolean.class);

// evaluates to true
boolean trueValue = parser.parseExpression(
        "'5.00' matches '\^-?\\d+(\\.\\d{2})?$'").getValue(Boolean.class);

//evaluates to false
boolean falseValue = parser.parseExpression(
        "'5.0067' matches '\^-?\\d+(\\.\\d{2})?$'").getValue(Boolean.class);
Spring框架参考文档 Spring Framework Reference Documentation

Be careful with primitive types as they are immediately boxed up to the wrapper type, so 1 instanceof T(int) evaluates to false while1 instanceof T(Integer.class) evaluates to true, as expected.

Each symbolic operator can also be specified as a purely alphabetic equivalent. This avoids problems where the symbols used have special meaning for the document type in which the expression is embedded (eg. an XML document). The textual equivalents are shown here: lt (<), gt (>), le (⇐), ge (>=), eq (==), ne (!=),div (/), mod (%), not (!). These are case insensitive.

Logical operators

The logical operators that are supported are and, or, and not. Their use is demonstrated below.

// -- AND --

// evaluates to false
boolean falseValue = parser.parseExpression("true and false").getValue(Boolean.class);

// evaluates to true
String expression = "isMember('Nikola Tesla') and isMember('Mihajlo Pupin')";
boolean trueValue = parser.parseExpression(expression).getValue(societyContext, Boolean.class);

// -- OR --

// evaluates to true
boolean trueValue = parser.parseExpression("true or false").getValue(Boolean.class);

// evaluates to true
String expression = "isMember('Nikola Tesla') or isMember('Albert Einstein')";
boolean trueValue = parser.parseExpression(expression).getValue(societyContext, Boolean.class);

// -- NOT --

// evaluates to false
boolean falseValue = parser.parseExpression("!true").getValue(Boolean.class);

// -- AND and NOT --
String expression = "isMember('Nikola Tesla') and !isMember('Mihajlo Pupin')";
boolean falseValue = parser.parseExpression(expression).getValue(societyContext, Boolean.class);

Mathematical operators

The addition operator can be used on both numbers and strings. Subtraction, multiplication and division can be used only on numbers. Other mathematical operators supported are modulus (%) and exponential power (^). Standard operator precedence is enforced. These operators are demonstrated below.

// Addition
int two = parser.parseExpression("1 + 1").getValue(Integer.class); // 2

String testString = parser.parseExpression(
        "'test' + ' ' + 'string'").getValue(String.class); // 'test string'

// Subtraction
int four = parser.parseExpression("1 - -3").getValue(Integer.class); // 4

double d = parser.parseExpression("1000.00 - 1e4").getValue(Double.class); // -9000

// Multiplication
int six = parser.parseExpression("-2 * -3").getValue(Integer.class); // 6

double twentyFour = parser.parseExpression("2.0 * 3e0 * 4").getValue(Double.class); // 24.0

// Division
int minusTwo = parser.parseExpression("6 / -3").getValue(Integer.class); // -2

double one = parser.parseExpression("8.0 / 4e0 / 2").getValue(Double.class); // 1.0

// Modulus
int three = parser.parseExpression("7 % 4").getValue(Integer.class); // 3

int one = parser.parseExpression("8 / 5 % 2").getValue(Integer.class); // 1

// Operator precedence
int minusTwentyOne = parser.parseExpression("1+2-3*8").getValue(Integer.class); // -21

10.5.8 Assignment

Setting of a property is done by using the assignment operator. This would typically be done within a call to setValue but can also be done inside a call to getValue.

Inventor inventor = new Inventor();
StandardEvaluationContext inventorContext = new StandardEvaluationContext(inventor);

parser.parseExpression("Name").setValue(inventorContext, "Alexander Seovic2");

// alternatively

String aleks = parser.parseExpression(
        "Name = 'Alexandar Seovic'").getValue(inventorContext, String.class);

10.5.9 Types

The special T operator can be used to specify an instance of java.lang.Class (the type). Static methods are invoked using this operator as well. TheStandardEvaluationContext uses a TypeLocator to find types and the StandardTypeLocator (which can be replaced) is built with an understanding of the java.lang package. This means T() references to types within java.lang do not need to be fully qualified, but all other type references must be.

Class dateClass = parser.parseExpression("T(java.util.Date)").getValue(Class.class);

Class stringClass = parser.parseExpression("T(String)").getValue(Class.class);

boolean trueValue = parser.parseExpression(
        "T(java.math.RoundingMode).CEILING < T(java.math.RoundingMode).FLOOR")
        .getValue(Boolean.class);

10.5.10 Constructors

Constructors can be invoked using the new operator. The fully qualified class name should be used for all but the primitive type and String (where int, float, etc, can be used).

Inventor einstein = p.parseExpression(
        "new org.spring.samples.spel.inventor.Inventor('Albert Einstein', 'German')")
        .getValue(Inventor.class);

//create new inventor instance within add method of List
p.parseExpression(
        "Members.add(new org.spring.samples.spel.inventor.Inventor(
            'Albert Einstein', 'German'))").getValue(societyContext);

10.5.11 Variables

Variables can be referenced in the expression using the syntax #variableName. Variables are set using the method setVariable on the StandardEvaluationContext.

Inventor tesla = new Inventor("Nikola Tesla", "Serbian");
StandardEvaluationContext context = new StandardEvaluationContext(tesla);
context.setVariable("newName", "Mike Tesla");

parser.parseExpression("Name = #newName").getValue(context);

System.out.println(tesla.getName()) // "Mike Tesla"

The #this and #root variables

The variable #this is always defined and refers to the current evaluation object (against which unqualified references are resolved). The variable #root is always defined and refers to the root context object. Although #this may vary as components of an expression are evaluated, #root always refers to the root.

// create an array of integers
List<Integer> primes = new ArrayList<Integer>();
primes.addAll(Arrays.asList(2,3,5,7,11,13,17));

// create parser and set variable 'primes' as the array of integers
ExpressionParser parser = new SpelExpressionParser();
StandardEvaluationContext context = new StandardEvaluationContext();
context.setVariable("primes",primes);

// all prime numbers > 10 from the list (using selection ?{...})
// evaluates to [11, 13, 17]
List<Integer> primesGreaterThanTen = (List<Integer>) parser.parseExpression(
        "#primes.?[#this>10]").getValue(context);

10.5.12 Functions

You can extend SpEL by registering user defined functions that can be called within the expression string. The function is registered with theStandardEvaluationContext using the method.

public void registerFunction(String name, Method m)

A reference to a Java Method provides the implementation of the function. For example, a utility method to reverse a string is shown below.

public abstract class StringUtils {

    public static String reverseString(String input) {
        StringBuilder backwards = new StringBuilder();
        for (int i = 0; i < input.length(); i++)
            backwards.append(input.charAt(input.length() - 1 - i));
        }
        return backwards.toString();
    }
}

This method is then registered with the evaluation context and can be used within an expression string.

ExpressionParser parser = new SpelExpressionParser();
StandardEvaluationContext context = new StandardEvaluationContext();

context.registerFunction("reverseString",
    StringUtils.class.getDeclaredMethod("reverseString", new Class[] { String.class }));

String helloWorldReversed = parser.parseExpression(
    "#reverseString('hello')").getValue(context, String.class);

10.5.13 Bean references

If the evaluation context has been configured with a bean resolver it is possible to lookup beans from an expression using the (@) symbol.

ExpressionParser parser = new SpelExpressionParser();
StandardEvaluationContext context = new StandardEvaluationContext();
context.setBeanResolver(new MyBeanResolver());

// This will end up calling resolve(context,"foo") on MyBeanResolver during evaluation
Object bean = parser.parseExpression("@foo").getValue(context);

To access a factory bean itself, the bean name should instead be prefixed with a (&) symbol.

ExpressionParser parser = new SpelExpressionParser();
StandardEvaluationContext context = new StandardEvaluationContext();
context.setBeanResolver(new MyBeanResolver());

// This will end up calling resolve(context,"&foo") on MyBeanResolver during evaluation
Object bean = parser.parseExpression("&foo").getValue(context);

10.5.14 Ternary Operator (If-Then-Else)

You can use the ternary operator for performing if-then-else conditional logic inside the expression. A minimal example is:

String falseString = parser.parseExpression(
        "false ? 'trueExp' : 'falseExp'").getValue(String.class);

In this case, the boolean false results in returning the string value 'falseExp'. A more realistic example is shown below.

parser.parseExpression("Name").setValue(societyContext, "IEEE");
societyContext.setVariable("queryName", "Nikola Tesla");

expression = "isMember(#queryName)? #queryName + ' is a member of the ' " +
        "+ Name + ' Society' : #queryName + ' is not a member of the ' + Name + ' Society'";

String queryResultString = parser.parseExpression(expression)
        .getValue(societyContext, String.class);
// queryResultString = "Nikola Tesla is a member of the IEEE Society"

Also see the next section on the Elvis operator for an even shorter syntax for the ternary operator.

10.5.15 The Elvis Operator

The Elvis operator is a shortening of the ternary operator syntax and is used in the Groovy language. With the ternary operator syntax you usually have to repeat a variable twice, for example:

String name = "Elvis Presley";
String displayName = name != null ? name : "Unknown";

Instead you can use the Elvis operator, named for the resemblance to Elvis' hair style.

ExpressionParser parser = new SpelExpressionParser();

String name = parser.parseExpression("name?:'Unknown'").getValue(String.class);

System.out.println(name); // 'Unknown'

Here is a more complex example.

ExpressionParser parser = new SpelExpressionParser();

Inventor tesla = new Inventor("Nikola Tesla", "Serbian");
StandardEvaluationContext context = new StandardEvaluationContext(tesla);

String name = parser.parseExpression("Name?:'Elvis Presley'").getValue(context, String.class);

System.out.println(name); // Nikola Tesla

tesla.setName(null);

name = parser.parseExpression("Name?:'Elvis Presley'").getValue(context, String.class);

System.out.println(name); // Elvis Presley

10.5.16 Safe Navigation operator

The Safe Navigation operator is used to avoid a NullPointerException and comes from the Groovy language. Typically when you have a reference to an object you might need to verify that it is not null before accessing methods or properties of the object. To avoid this, the safe navigation operator will simply return null instead of throwing an exception.

ExpressionParser parser = new SpelExpressionParser();

Inventor tesla = new Inventor("Nikola Tesla", "Serbian");
tesla.setPlaceOfBirth(new PlaceOfBirth("Smiljan"));

StandardEvaluationContext context = new StandardEvaluationContext(tesla);

String city = parser.parseExpression("PlaceOfBirth?.City").getValue(context, String.class);
System.out.println(city); // Smiljan

tesla.setPlaceOfBirth(null);

city = parser.parseExpression("PlaceOfBirth?.City").getValue(context, String.class);

System.out.println(city); // null - does not throw NullPointerException!!!
Spring框架参考文档 Spring Framework Reference Documentation

The Elvis operator can be used to apply default values in expressions, e.g. in an @Value expression:

@Value("#{systemProperties['pop3.port'] ?: 25}")

This will inject a system property pop3.port if it is defined or 25 if not.

10.5.17 Collection Selection

Selection is a powerful expression language feature that allows you to transform some source collection into another by selecting from its entries.

Selection uses the syntax ?[selectionExpression]. This will filter the collection and return a new collection containing a subset of the original elements. For example, selection would allow us to easily get a list of Serbian inventors:

List<Inventor> list = (List<Inventor>) parser.parseExpression(
        "Members.?[Nationality == 'Serbian']").getValue(societyContext);

Selection is possible upon both lists and maps. In the former case the selection criteria is evaluated against each individual list element whilst against a map the selection criteria is evaluated against each map entry (objects of the Java type Map.Entry). Map entries have their key and value accessible as properties for use in the selection.

This expression will return a new map consisting of those elements of the original map where the entry value is less than 27.

Map newMap = parser.parseExpression("map.?[value<27]").getValue();

In addition to returning all the selected elements, it is possible to retrieve just the first or the last value. To obtain the first entry matching the selection the syntax is ^[…​]whilst to obtain the last matching selection the syntax is $[…​].

10.5.18 Collection Projection

Projection allows a collection to drive the evaluation of a sub-expression and the result is a new collection. The syntax for projection is ![projectionExpression]. Most easily understood by example, suppose we have a list of inventors but want the list of cities where they were born. Effectively we want to evaluate 'placeOfBirth.city' for every entry in the inventor list. Using projection:

// returns ['Smiljan', 'Idvor' ]
List placesOfBirth = (List)parser.parseExpression("Members.![placeOfBirth.city]");

A map can also be used to drive projection and in this case the projection expression is evaluated against each entry in the map (represented as a Java Map.Entry). The result of a projection across a map is a list consisting of the evaluation of the projection expression against each map entry.

10.5.19 Expression templating

Expression templates allow a mixing of literal text with one or more evaluation blocks. Each evaluation block is delimited with prefix and suffix characters that you can define, a common choice is to use #{ } as the delimiters. For example,

String randomPhrase = parser.parseExpression(
        "random number is #{T(java.lang.Math).random()}",
        new TemplateParserContext()).getValue(String.class);

// evaluates to "random number is 0.7038186818312008"

The string is evaluated by concatenating the literal text 'random number is ' with the result of evaluating the expression inside the #{ } delimiter, in this case the result of calling that random() method. The second argument to the method parseExpression() is of the type ParserContext. The ParserContext interface is used to influence how the expression is parsed in order to support the expression templating functionality. The definition of TemplateParserContext is shown below.

public class TemplateParserContext implements ParserContext {

    public String getExpressionPrefix() {
        return "#{";
    }

    public String getExpressionSuffix() {
        return "}";
    }

    public boolean isTemplate() {
        return true;
    }
}

10.6 Classes used in the examples

Inventor.java

package org.spring.samples.spel.inventor;

import java.util.Date;
import java.util.GregorianCalendar;

public class Inventor {

    private String name;
    private String nationality;
    private String[] inventions;
    private Date birthdate;
    private PlaceOfBirth placeOfBirth;

    public Inventor(String name, String nationality) {
        GregorianCalendar c= new GregorianCalendar();
        this.name = name;
        this.nationality = nationality;
        this.birthdate = c.getTime();
    }

    public Inventor(String name, Date birthdate, String nationality) {
        this.name = name;
        this.nationality = nationality;
        this.birthdate = birthdate;
    }

    public Inventor() {
    }

    public String getName() {
        return name;
    }

    public void setName(String name) {
        this.name = name;
    }

    public String getNationality() {
        return nationality;
    }

    public void setNationality(String nationality) {
        this.nationality = nationality;
    }

    public Date getBirthdate() {
        return birthdate;
    }

    public void setBirthdate(Date birthdate) {
        this.birthdate = birthdate;
    }

    public PlaceOfBirth getPlaceOfBirth() {
        return placeOfBirth;
    }

    public void setPlaceOfBirth(PlaceOfBirth placeOfBirth) {
        this.placeOfBirth = placeOfBirth;
    }

    public void setInventions(String[] inventions) {
        this.inventions = inventions;
    }

    public String[] getInventions() {
        return inventions;
    }
}

PlaceOfBirth.java

package org.spring.samples.spel.inventor;

public class PlaceOfBirth {

    private String city;
    private String country;

    public PlaceOfBirth(String city) {
        this.city=city;
    }

    public PlaceOfBirth(String city, String country) {
        this(city);
        this.country = country;
    }

    public String getCity() {
        return city;
    }

    public void setCity(String s) {
        this.city = s;
    }

    public String getCountry() {
        return country;
    }

    public void setCountry(String country) {
        this.country = country;
    }

}

Society.java

package org.spring.samples.spel.inventor;

import java.util.*;

public class Society {

    private String name;

    public static String Advisors = "advisors";
    public static String President = "president";

    private List<Inventor> members = new ArrayList<Inventor>();
    private Map officers = new HashMap();

    public List getMembers() {
        return members;
    }

    public Map getOfficers() {
        return officers;
    }

    public String getName() {
        return name;
    }

    public void setName(String name) {
        this.name = name;
    }

    public boolean isMember(String name) {
        for (Inventor inventor : members) {
            if (inventor.getName().equals(name)) {
                return true;
            }
        }
        return false;
    }

}

11. Aspect Oriented Programming with Spring

11.1 Introduction

Aspect-Oriented Programming (AOP) complements Object-Oriented Programming (OOP) by providing another way of thinking about program structure. The key unit of modularity in OOP is the class, whereas in AOP the unit of modularity is the aspect. Aspects enable the modularization of concerns such as transaction management that cut across multiple types and objects. (Such concerns are often termed crosscutting concerns in AOP literature.)

One of the key components of Spring is the AOP framework. While the Spring IoC container does not depend on AOP, meaning you do not need to use AOP if you don’t want to, AOP complements Spring IoC to provide a very capable middleware solution.

Spring 2.0 AOP

Spring 2.0 introduces a simpler and more powerful way of writing custom aspects using either a schema-based approach or the @AspectJ annotation style. Both of these styles offer fully typed advice and use of the AspectJ pointcut language, while still using Spring AOP for weaving.

The Spring 2.0 schema- and @AspectJ-based AOP support is discussed in this chapter. Spring 2.0 AOP remains fully backwards compatible with Spring 1.2 AOP, and the lower-level AOP support offered by the Spring 1.2 APIs is discussed in the following chapter.

AOP is used in the Spring Framework to…​

  • …​ provide declarative enterprise services, especially as a replacement for EJB declarative services. The most important such service is declarative transaction management.
  • …​ allow users to implement custom aspects, complementing their use of OOP with AOP.
Spring框架参考文档 Spring Framework Reference Documentation

If you are interested only in generic declarative services or other pre-packaged declarative middleware services such as pooling, you do not need to work directly with Spring AOP, and can skip most of this chapter.

11.1.1 AOP concepts

Let us begin by defining some central AOP concepts and terminology. These terms are not Spring-specific…​ unfortunately, AOP terminology is not particularly intuitive; however, it would be even more confusing if Spring used its own terminology.

  • Aspect: a modularization of a concern that cuts across multiple classes. Transaction management is a good example of a crosscutting concern in enterprise Java applications. In Spring AOP, aspects are implemented using regular classes (the schema-based approach) or regular classes annotated with the @Aspectannotation (the @AspectJ style).
  • Join point: a point during the execution of a program, such as the execution of a method or the handling of an exception. In Spring AOP, a join point alwaysrepresents a method execution.
  • Advice: action taken by an aspect at a particular join point. Different types of advice include "around," "before" and "after" advice. (Advice types are discussed below.) Many AOP frameworks, including Spring, model an advice as an interceptor, maintaining a chain of interceptors around the join point.
  • Pointcut: a predicate that matches join points. Advice is associated with a pointcut expression and runs at any join point matched by the pointcut (for example, the execution of a method with a certain name). The concept of join points as matched by pointcut expressions is central to AOP, and Spring uses the AspectJ pointcut expression language by default.
  • Introduction: declaring additional methods or fields on behalf of a type. Spring AOP allows you to introduce new interfaces (and a corresponding implementation) to any advised object. For example, you could use an introduction to make a bean implement an IsModified interface, to simplify caching. (An introduction is known as an inter-type declaration in the AspectJ community.)
  • Target object: object being advised by one or more aspects. Also referred to as the advised object. Since Spring AOP is implemented using runtime proxies, this object will always be a proxied object.
  • AOP proxy: an object created by the AOP framework in order to implement the aspect contracts (advise method executions and so on). In the Spring Framework, an AOP proxy will be a JDK dynamic proxy or a CGLIB proxy.
  • Weaving: linking aspects with other application types or objects to create an advised object. This can be done at compile time (using the AspectJ compiler, for example), load time, or at runtime. Spring AOP, like other pure Java AOP frameworks, performs weaving at runtime.

Types of advice:

  • Before advice: Advice that executes before a join point, but which does not have the ability to prevent execution flow proceeding to the join point (unless it throws an exception).
  • After returning advice: Advice to be executed after a join point completes normally: for example, if a method returns without throwing an exception.
  • After throwing advice: Advice to be executed if a method exits by throwing an exception.
  • After (finally) advice: Advice to be executed regardless of the means by which a join point exits (normal or exceptional return).
  • Around advice: Advice that surrounds a join point such as a method invocation. This is the most powerful kind of advice. Around advice can perform custom behavior before and after the method invocation. It is also responsible for choosing whether to proceed to the join point or to shortcut the advised method execution by returning its own return value or throwing an exception.

Around advice is the most general kind of advice. Since Spring AOP, like AspectJ, provides a full range of advice types, we recommend that you use the least powerful advice type that can implement the required behavior. For example, if you need only to update a cache with the return value of a method, you are better off implementing an after returning advice than an around advice, although an around advice can accomplish the same thing. Using the most specific advice type provides a simpler programming model with less potential for errors. For example, you do not need to invoke the proceed() method on the JoinPoint used for around advice, and hence cannot fail to invoke it.

In Spring 2.0, all advice parameters are statically typed, so that you work with advice parameters of the appropriate type (the type of the return value from a method execution for example) rather than Object arrays.

The concept of join points, matched by pointcuts, is the key to AOP which distinguishes it from older technologies offering only interception. Pointcuts enable advice to be targeted independently of the Object-Oriented hierarchy. For example, an around advice providing declarative transaction management can be applied to a set of methods spanning multiple objects (such as all business operations in the service layer).

11.1.2 Spring AOP capabilities and goals

Spring AOP is implemented in pure Java. There is no need for a special compilation process. Spring AOP does not need to control the class loader hierarchy, and is thus suitable for use in a Servlet container or application server.

Spring AOP currently supports only method execution join points (advising the execution of methods on Spring beans). Field interception is not implemented, although support for field interception could be added without breaking the core Spring AOP APIs. If you need to advise field access and update join points, consider a language such as AspectJ.

Spring AOP’s approach to AOP differs from that of most other AOP frameworks. The aim is not to provide the most complete AOP implementation (although Spring AOP is quite capable); it is rather to provide a close integration between AOP implementation and Spring IoC to help solve common problems in enterprise applications.

Thus, for example, the Spring Framework’s AOP functionality is normally used in conjunction with the Spring IoC container. Aspects are configured using normal bean definition syntax (although this allows powerful "autoproxying" capabilities): this is a crucial difference from other AOP implementations. There are some things you cannot do easily or efficiently with Spring AOP, such as advise very fine-grained objects (such as domain objects typically): AspectJ is the best choice in such cases. However, our experience is that Spring AOP provides an excellent solution to most problems in enterprise Java applications that are amenable to AOP.

Spring AOP will never strive to compete with AspectJ to provide a comprehensive AOP solution. We believe that both proxy-based frameworks like Spring AOP and full-blown frameworks such as AspectJ are valuable, and that they are complementary, rather than in competition. Spring seamlessly integrates Spring AOP and IoC with AspectJ, to enable all uses of AOP to be catered for within a consistent Spring-based application architecture. This integration does not affect the Spring AOP API or the AOP Alliance API: Spring AOP remains backward-compatible. See the following chapter for a discussion of the Spring AOP APIs.

Spring框架参考文档 Spring Framework Reference Documentation

One of the central tenets of the Spring Framework is that of non-invasiveness; this is the idea that you should not be forced to introduce framework-specific classes and interfaces into your business/domain model. However, in some places the Spring Framework does give you the option to introduce Spring Framework-specific dependencies into your codebase: the rationale in giving you such options is because in certain scenarios it might be just plain easier to read or code some specific piece of functionality in such a way. The Spring Framework (almost) always offers you the choice though: you have the freedom to make an informed decision as to which option best suits your particular use case or scenario.

One such choice that is relevant to this chapter is that of which AOP framework (and which AOP style) to choose. You have the choice of AspectJ and/or Spring AOP, and you also have the choice of either the @AspectJ annotation-style approach or the Spring XML configuration-style approach. The fact that this chapter chooses to introduce the @AspectJ-style approach first should not be taken as an indication that the Spring team favors the @AspectJ annotation-style approach over the Spring XML configuration-style.

See Section 11.4, “Choosing which AOP declaration style to use” for a more complete discussion of the whys and wherefores of each style.

11.1.3 AOP Proxies

Spring AOP defaults to using standard JDK dynamic proxies for AOP proxies. This enables any interface (or set of interfaces) to be proxied.

Spring AOP can also use CGLIB proxies. This is necessary to proxy classes rather than interfaces. CGLIB is used by default if a business object does not implement an interface. As it is good practice to program to interfaces rather than classes; business classes normally will implement one or more business interfaces. It is possible toforce the use of CGLIB, in those (hopefully rare) cases where you need to advise a method that is not declared on an interface, or where you need to pass a proxied object to a method as a concrete type.

It is important to grasp the fact that Spring AOP is proxy-based. See Section 11.6.1, “Understanding AOP proxies” for a thorough examination of exactly what this implementation detail actually means.

11.2 @AspectJ support

@AspectJ refers to a style of declaring aspects as regular Java classes annotated with annotations. The @AspectJ style was introduced by the AspectJ project as part of the AspectJ 5 release. Spring interprets the same annotations as AspectJ 5, using a library supplied by AspectJ for pointcut parsing and matching. The AOP runtime is still pure Spring AOP though, and there is no dependency on the AspectJ compiler or weaver.

Spring框架参考文档 Spring Framework Reference Documentation

Using the AspectJ compiler and weaver enables use of the full AspectJ language, and is discussed in Section 11.8, “Using AspectJ with Spring applications”.

11.2.1 Enabling @AspectJ Support

To use @AspectJ aspects in a Spring configuration you need to enable Spring support for configuring Spring AOP based on @AspectJ aspects, and autoproxying beans based on whether or not they are advised by those aspects. By autoproxying we mean that if Spring determines that a bean is advised by one or more aspects, it will automatically generate a proxy for that bean to intercept method invocations and ensure that advice is executed as needed.

The @AspectJ support can be enabled with XML or Java style configuration. In either case you will also need to ensure that AspectJ’s aspectjweaver.jar library is on the classpath of your application (version 1.6.8 or later). This library is available in the 'lib' directory of an AspectJ distribution or via the Maven Central repository.

Enabling @AspectJ Support with Java configuration

To enable @AspectJ support with Java @Configuration add the @EnableAspectJAutoProxy annotation:

@Configuration
@EnableAspectJAutoProxy
public class AppConfig {

}

Enabling @AspectJ Support with XML configuration

To enable @AspectJ support with XML based configuration use the aop:aspectj-autoproxy element:

<aop:aspectj-autoproxy/>

This assumes that you are using schema support as described in Chapter 41, XML Schema-based configuration. See Section 41.2.7, “the aop schema” for how to import the tags in the aop namespace.

11.2.2 Declaring an aspect

With the @AspectJ support enabled, any bean defined in your application context with a class that is an @AspectJ aspect (has the @Aspect annotation) will be automatically detected by Spring and used to configure Spring AOP. The following example shows the minimal definition required for a not-very-useful aspect:

A regular bean definition in the application context, pointing to a bean class that has the @Aspect annotation:

<bean id="myAspect" class="org.xyz.NotVeryUsefulAspect">
    <!-- configure properties of aspect here as normal -->
</bean>

And the NotVeryUsefulAspect class definition, annotated with org.aspectj.lang.annotation.Aspect annotation;

package org.xyz;
import org.aspectj.lang.annotation.Aspect;

@Aspect
public class NotVeryUsefulAspect {

}

Aspects (classes annotated with @Aspect) may have methods and fields just like any other class. They may also contain pointcut, advice, and introduction (inter-type) declarations.

Spring框架参考文档 Spring Framework Reference Documentation

You may register aspect classes as regular beans in your Spring XML configuration, or autodetect them through classpath scanning - just like any other Spring-managed bean. However, note that the @Aspect annotation is not sufficient for autodetection in the classpath: For that purpose, you need to add a separate @Component annotation (or alternatively a custom stereotype annotation that qualifies, as per the rules of Spring’s component scanner).
Spring框架参考文档 Spring Framework Reference Documentation

In Spring AOP, it is not possible to have aspects themselves be the target of advice from other aspects. The @Aspect annotation on a class marks it as an aspect, and hence excludes it from auto-proxying.

11.2.3 Declaring a pointcut

Recall that pointcuts determine join points of interest, and thus enable us to control when advice executes. Spring AOP only supports method execution join points for Spring beans, so you can think of a pointcut as matching the execution of methods on Spring beans. A pointcut declaration has two parts: a signature comprising a name and any parameters, and a pointcut expression that determines exactly which method executions we are interested in. In the @AspectJ annotation-style of AOP, a pointcut signature is provided by a regular method definition, and the pointcut expression is indicated using the @Pointcut annotation (the method serving as the pointcut signature must have a void return type).

An example will help make this distinction between a pointcut signature and a pointcut expression clear. The following example defines a pointcut named'anyOldTransfer' that will match the execution of any method named 'transfer':

@Pointcut("execution(* transfer(..))")// the pointcut expression
private void anyOldTransfer() {}// the pointcut signature

The pointcut expression that forms the value of the @Pointcut annotation is a regular AspectJ 5 pointcut expression. For a full discussion of AspectJ’s pointcut language, see the AspectJ Programming Guide (and for extensions, the AspectJ 5 Developers Notebook) or one of the books on AspectJ such as "Eclipse AspectJ" by Colyer et. al. or "AspectJ in Action" by Ramnivas Laddad.

Supported Pointcut Designators

Spring AOP supports the following AspectJ pointcut designators (PCD) for use in pointcut expressions:

Other pointcut types

The full AspectJ pointcut language supports additional pointcut designators that are not supported in Spring. These are:call, get, set, preinitialization, staticinitialization, initialization, handler, adviceexecution, withincode, cflow, cflowbelow, if, @this, and @withincode. Use of these pointcut designators in pointcut expressions interpreted by Spring AOP will result in an IllegalArgumentException being thrown.

The set of pointcut designators supported by Spring AOP may be extended in future releases to support more of the AspectJ pointcut designators.

  • execution - for matching method execution join points, this is the primary pointcut designator you will use when working with Spring AOP
  • within - limits matching to join points within certain types (simply the execution of a method declared within a matching type when using Spring AOP)
  • this - limits matching to join points (the execution of methods when using Spring AOP) where the bean reference (Spring AOP proxy) is an instance of the given type
  • target - limits matching to join points (the execution of methods when using Spring AOP) where the target object (application object being proxied) is an instance of the given type
  • args - limits matching to join points (the execution of methods when using Spring AOP) where the arguments are instances of the given types
  • @target - limits matching to join points (the execution of methods when using Spring AOP) where the class of the executing object has an annotation of the given type
  • @args - limits matching to join points (the execution of methods when using Spring AOP) where the runtime type of the actual arguments passed have annotations of the given type(s)
  • @within - limits matching to join points within types that have the given annotation (the execution of methods declared in types with the given annotation when using Spring AOP)
  • @annotation - limits matching to join points where the subject of the join point (method being executed in Spring AOP) has the given annotation

Because Spring AOP limits matching to only method execution join points, the discussion of the pointcut designators above gives a narrower definition than you will find in the AspectJ programming guide. In addition, AspectJ itself has type-based semantics and at an execution join point both this and target refer to the same object - the object executing the method. Spring AOP is a proxy-based system and differentiates between the proxy object itself (bound to this) and the target object behind the proxy (bound to target).

Spring框架参考文档 Spring Framework Reference Documentation

Due to the proxy-based nature of Spring’s AOP framework, protected methods are by definition not intercepted, neither for JDK proxies (where this isn’t applicable) nor for CGLIB proxies (where this is technically possible but not recommendable for AOP purposes). As a consequence, any given pointcut will be matched against public methods only!

If your interception needs include protected/private methods or even constructors, consider the use of Spring-driven native AspectJ weaving instead of Spring’s proxy-based AOP framework. This constitutes a different mode of AOP usage with different characteristics, so be sure to make yourself familiar with weaving first before making a decision.

Spring AOP also supports an additional PCD named bean. This PCD allows you to limit the matching of join points to a particular named Spring bean, or to a set of named Spring beans (when using wildcards). The bean PCD has the following form:

bean(idOrNameOfBean)

The idOrNameOfBean token can be the name of any Spring bean: limited wildcard support using the * character is provided, so if you establish some naming conventions for your Spring beans you can quite easily write a bean PCD expression to pick them out. As is the case with other pointcut designators, the bean PCD can be &&'ed, ||'ed, and ! (negated) too.

Spring框架参考文档 Spring Framework Reference Documentation

Please note that the bean PCD is only supported in Spring AOP - and not in native AspectJ weaving. It is a Spring-specific extension to the standard PCDs that AspectJ defines and therefore not available for aspects declared in the @Aspect model.

The bean PCD operates at the instance level (building on the Spring bean name concept) rather than at the type level only (which is what weaving-based AOP is limited to). Instance-based pointcut designators are a special capability of Spring’s proxy-based AOP framework and its close integration with the Spring bean factory, where it is natural and straightforward to identify specific beans by name.

Combining pointcut expressions

Pointcut expressions can be combined using '&&', '||' and '!'. It is also possible to refer to pointcut expressions by name. The following example shows three pointcut expressions: anyPublicOperation (which matches if a method execution join point represents the execution of any public method); inTrading (which matches if a method execution is in the trading module), and tradingOperation (which matches if a method execution represents any public method in the trading module).

@Pointcut("execution(public * *(..))")
private void anyPublicOperation() {}

@Pointcut("within(com.xyz.someapp.trading..*)")
private void inTrading() {}

@Pointcut("anyPublicOperation() && inTrading()")
private void tradingOperation() {}

It is a best practice to build more complex pointcut expressions out of smaller named components as shown above. When referring to pointcuts by name, normal Java visibility rules apply (you can see private pointcuts in the same type, protected pointcuts in the hierarchy, public pointcuts anywhere and so on). Visibility does not affect pointcut matching.

Sharing common pointcut definitions

When working with enterprise applications, you often want to refer to modules of the application and particular sets of operations from within several aspects. We recommend defining a "SystemArchitecture" aspect that captures common pointcut expressions for this purpose. A typical such aspect would look as follows:

package com.xyz.someapp;

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Pointcut;

@Aspect
public class SystemArchitecture {

    /**
     * A join point is in the web layer if the method is defined
     * in a type in the com.xyz.someapp.web package or any sub-package
     * under that.
     */
    @Pointcut("within(com.xyz.someapp.web..*)")
    public void inWebLayer() {}

    /**
     * A join point is in the service layer if the method is defined
     * in a type in the com.xyz.someapp.service package or any sub-package
     * under that.
     */
    @Pointcut("within(com.xyz.someapp.service..*)")
    public void inServiceLayer() {}

    /**
     * A join point is in the data access layer if the method is defined
     * in a type in the com.xyz.someapp.dao package or any sub-package
     * under that.
     */
    @Pointcut("within(com.xyz.someapp.dao..*)")
    public void inDataAccessLayer() {}

    /**
     * A business service is the execution of any method defined on a service
     * interface. This definition assumes that interfaces are placed in the
     * "service" package, and that implementation types are in sub-packages.
     *
     * If you group service interfaces by functional area (for example,
     * in packages com.xyz.someapp.abc.service and com.xyz.someapp.def.service) then
     * the pointcut expression "execution(* com.xyz.someapp..service.*.*(..))"
     * could be used instead.
     *
     * Alternatively, you can write the expression using the 'bean'
     * PCD, like so "bean(*Service)". (This assumes that you have
     * named your Spring service beans in a consistent fashion.)
     */
    @Pointcut("execution(* com.xyz.someapp..service.*.*(..))")
    public void businessService() {}

    /**
     * A data access operation is the execution of any method defined on a
     * dao interface. This definition assumes that interfaces are placed in the
     * "dao" package, and that implementation types are in sub-packages.
     */
    @Pointcut("execution(* com.xyz.someapp.dao.*.*(..))")
    public void dataAccessOperation() {}

}

The pointcuts defined in such an aspect can be referred to anywhere that you need a pointcut expression. For example, to make the service layer transactional, you could write:

<aop:config>
    <aop:advisor
        pointcut="com.xyz.someapp.SystemArchitecture.businessService()"
        advice-ref="tx-advice"/>
</aop:config>

<tx:advice id="tx-advice">
    <tx:attributes>
        <tx:method name="*" propagation="REQUIRED"/>
    </tx:attributes>
</tx:advice>

The <aop:config> and <aop:advisor> elements are discussed in Section 11.3, “Schema-based AOP support”. The transaction elements are discussed inChapter 17, Transaction Management.

Examples

Spring AOP users are likely to use the execution pointcut designator the most often. The format of an execution expression is:

execution(modifiers-pattern? ret-type-pattern declaring-type-pattern?name-pattern(param-pattern)
            throws-pattern?)

All parts except the returning type pattern (ret-type-pattern in the snippet above), name pattern, and parameters pattern are optional. The returning type pattern determines what the return type of the method must be in order for a join point to be matched. Most frequently you will use * as the returning type pattern, which matches any return type. A fully-qualified type name will match only when the method returns the given type. The name pattern matches the method name. You can use the * wildcard as all or part of a name pattern. If specifying a declaring type pattern then include a trailing . to join it to the name pattern component. The parameters pattern is slightly more complex: () matches a method that takes no parameters, whereas (..) matches any number of parameters (zero or more). The pattern (*)matches a method taking one parameter of any type, (*,String) matches a method taking two parameters, the first can be of any type, the second must be a String. Consult the Language Semantics section of the AspectJ Programming Guide for more information.

Some examples of common pointcut expressions are given below.

  • the execution of any public method:
execution(public * *(..))
  • the execution of any method with a name beginning with "set":
execution(* set*(..))
  • the execution of any method defined by the AccountService interface:
execution(* com.xyz.service.AccountService.*(..))
  • the execution of any method defined in the service package:
execution(* com.xyz.service.*.*(..))
  • the execution of any method defined in the service package or a sub-package:
execution(* com.xyz.service..*.*(..))
  • any join point (method execution only in Spring AOP) within the service package:
within(com.xyz.service.*)
  • any join point (method execution only in Spring AOP) within the service package or a sub-package:
within(com.xyz.service..*)
  • any join point (method execution only in Spring AOP) where the proxy implements the AccountService interface:
this(com.xyz.service.AccountService)
Spring框架参考文档 Spring Framework Reference Documentation

'this' is more commonly used in a binding form :- see the following section on advice for how to make the proxy object available in the advice body.
  • any join point (method execution only in Spring AOP) where the target object implements the AccountService interface:
target(com.xyz.service.AccountService)
Spring框架参考文档 Spring Framework Reference Documentation

'target' is more commonly used in a binding form :- see the following section on advice for how to make the target object available in the advice body.
  • any join point (method execution only in Spring AOP) which takes a single parameter, and where the argument passed at runtime is Serializable:
args(java.io.Serializable)
Spring框架参考文档 Spring Framework Reference Documentation

'args' is more commonly used in a binding form :- see the following section on advice for how to make the method arguments available in the advice body.

Note that the pointcut given in this example is different to execution(* *(java.io.Serializable)): the args version matches if the argument passed at runtime is Serializable, the execution version matches if the method signature declares a single parameter of type Serializable.

  • any join point (method execution only in Spring AOP) where the target object has an @Transactional annotation:
@target(org.springframework.transaction.annotation.Transactional)
Spring框架参考文档 Spring Framework Reference Documentation

'@target' can also be used in a binding form :- see the following section on advice for how to make the annotation object available in the advice body.
  • any join point (method execution only in Spring AOP) where the declared type of the target object has an @Transactional annotation:
@within(org.springframework.transaction.annotation.Transactional)
Spring框架参考文档 Spring Framework Reference Documentation

'@within' can also be used in a binding form :- see the following section on advice for how to make the annotation object available in the advice body.
  • any join point (method execution only in Spring AOP) where the executing method has an @Transactional annotation:
@annotation(org.springframework.transaction.annotation.Transactional)
Spring框架参考文档 Spring Framework Reference Documentation

'@annotation' can also be used in a binding form :- see the following section on advice for how to make the annotation object available in the advice body.
  • any join point (method execution only in Spring AOP) which takes a single parameter, and where the runtime type of the argument passed has the @Classifiedannotation:
@args(com.xyz.security.Classified)
Spring框架参考文档 Spring Framework Reference Documentation

'@args' can also be used in a binding form :- see the following section on advice for how to make the annotation object(s) available in the advice body.
  • any join point (method execution only in Spring AOP) on a Spring bean named tradeService:
bean(tradeService)
  • any join point (method execution only in Spring AOP) on Spring beans having names that match the wildcard expression *Service:
bean(*Service)

Writing good pointcuts

During compilation, AspectJ processes pointcuts in order to try and optimize matching performance. Examining code and determining if each join point matches (statically or dynamically) a given pointcut is a costly process. (A dynamic match means the match cannot be fully determined from static analysis and a test will be placed in the code to determine if there is an actual match when the code is running). On first encountering a pointcut declaration, AspectJ will rewrite it into an optimal form for the matching process. What does this mean? Basically pointcuts are rewritten in DNF (Disjunctive Normal Form) and the components of the pointcut are sorted such that those components that are cheaper to evaluate are checked first. This means you do not have to worry about understanding the performance of various pointcut designators and may supply them in any order in a pointcut declaration.

However, AspectJ can only work with what it is told, and for optimal performance of matching you should think about what they are trying to achieve and narrow the search space for matches as much as possible in the definition. The existing designators naturally fall into one of three groups: kinded, scoping and context:

  • Kinded designators are those which select a particular kind of join point. For example: execution, get, set, call, handler
  • Scoping designators are those which select a group of join points of interest (of probably many kinds). For example: within, withincode
  • Contextual designators are those that match (and optionally bind) based on context. For example: this, target, @annotation

A well written pointcut should try and include at least the first two types (kinded and scoping), whilst the contextual designators may be included if wishing to match based on join point context, or bind that context for use in the advice. Supplying either just a kinded designator or just a contextual designator will work but could affect weaving performance (time and memory used) due to all the extra processing and analysis. Scoping designators are very fast to match and their usage means AspectJ can very quickly dismiss groups of join points that should not be further processed - that is why a good pointcut should always include one if possible.

11.2.4 Declaring advice

Advice is associated with a pointcut expression, and runs before, after, or around method executions matched by the pointcut. The pointcut expression may be either a simple reference to a named pointcut, or a pointcut expression declared in place.

Before advice

Before advice is declared in an aspect using the @Before annotation:

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Before;

@Aspect
public class BeforeExample {

    @Before("com.xyz.myapp.SystemArchitecture.dataAccessOperation()")
    public void doAccessCheck() {
        // ...
    }

}

If using an in-place pointcut expression we could rewrite the above example as:

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Before;

@Aspect
public class BeforeExample {

    @Before("execution(* com.xyz.myapp.dao.*.*(..))")
    public void doAccessCheck() {
        // ...
    }

}

After returning advice

After returning advice runs when a matched method execution returns normally. It is declared using the @AfterReturning annotation:

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.AfterReturning;

@Aspect
public class AfterReturningExample {

    @AfterReturning("com.xyz.myapp.SystemArchitecture.dataAccessOperation()")
    public void doAccessCheck() {
        // ...
    }

}
Spring框架参考文档 Spring Framework Reference Documentation

Note: it is of course possible to have multiple advice declarations, and other members as well, all inside the same aspect. We’re just showing a single advice declaration in these examples to focus on the issue under discussion at the time.

Sometimes you need access in the advice body to the actual value that was returned. You can use the form of @AfterReturning that binds the return value for this:

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.AfterReturning;

@Aspect
public class AfterReturningExample {

    @AfterReturning(
        pointcut="com.xyz.myapp.SystemArchitecture.dataAccessOperation()",
        returning="retVal")
    public void doAccessCheck(Object retVal) {
        // ...
    }

}

The name used in the returning attribute must correspond to the name of a parameter in the advice method. When a method execution returns, the return value will be passed to the advice method as the corresponding argument value. A returning clause also restricts matching to only those method executions that return a value of the specified type ( Object in this case, which will match any return value).

Please note that it is not possible to return a totally different reference when using after-returning advice.

After throwing advice

After throwing advice runs when a matched method execution exits by throwing an exception. It is declared using the @AfterThrowing annotation:

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.AfterThrowing;

@Aspect
public class AfterThrowingExample {

    @AfterThrowing("com.xyz.myapp.SystemArchitecture.dataAccessOperation()")
    public void doRecoveryActions() {
        // ...
    }

}

Often you want the advice to run only when exceptions of a given type are thrown, and you also often need access to the thrown exception in the advice body. Use thethrowing attribute to both restrict matching (if desired, use Throwable as the exception type otherwise) and bind the thrown exception to an advice parameter.

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.AfterThrowing;

@Aspect
public class AfterThrowingExample {

    @AfterThrowing(
        pointcut="com.xyz.myapp.SystemArchitecture.dataAccessOperation()",
        throwing="ex")
    public void doRecoveryActions(DataAccessException ex) {
        // ...
    }

}

The name used in the throwing attribute must correspond to the name of a parameter in the advice method. When a method execution exits by throwing an exception, the exception will be passed to the advice method as the corresponding argument value. A throwing clause also restricts matching to only those method executions that throw an exception of the specified type ( DataAccessException in this case).

After (finally) advice

After (finally) advice runs however a matched method execution exits. It is declared using the @After annotation. After advice must be prepared to handle both normal and exception return conditions. It is typically used for releasing resources, etc.

import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.After;

@Aspect
public class AfterFinallyExample {

    @After("com.xyz.myapp.SystemArchitecture.dataAccessOperation()")
    public void doReleaseLock() {
        // ...
    }

}

Around advice

The final kind of advice is around advice. Around advice runs "around" a matched method execution. It has the opportunity to do work both before and after the method executes, and to determine when, how, and even if, the method actually gets to execute at all. Around advice is often used if you need to share state before and after a method execution in a thread-safe manner (starting and stopping a timer for example). Always use the least powerful form of advice that meets your requirements (i.e. don’t use around advice if simple before advice would do).

Around advice is declared using the @Around annotation. The first parameter of the advice method must be of type ProceedingJoinPoint. Within the body of the advice, calling proceed() on the ProceedingJoinPoint causes the underlying method to execute. The proceed method may also be called passing in anObject[] - the values in the array will be used as the arguments to the method execution when it proceeds.

Spring框架参考文档 Spring Framework Reference Documentation

The behavior of proceed when called with an Object[] is a little different than the behavior of proceed for around advice compiled by the AspectJ compiler. For around advice written using the traditional AspectJ language, the number of arguments passed to proceed must match the number of arguments passed to the around advice (not the number of arguments taken by the underlying join point), and the value passed to proceed in a given argument position supplants the original value at the join point for the entity the value was bound to (Don’t worry if this doesn’t make sense right now!). The approach taken by Spring is simpler and a better match to its proxy-based, execution only semantics. You only need to be aware of this difference if you are compiling @AspectJ aspects written for Spring and using proceed with arguments with the AspectJ compiler and weaver. There is a way to write such aspects that is 100% compatible across both Spring AOP and AspectJ, and this is discussed in the following section on advice parameters.

 
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.ProceedingJoinPoint;

@Aspect
public class AroundExample {

    @Around("com.xyz.myapp.SystemArchitecture.businessService()")
    public Object doBasicProfiling(ProceedingJoinPoint pjp) throws Throwable {
        // start stopwatch
        Object retVal = pjp.proceed();
        // stop stopwatch
        return retVal;
    }

}

The value returned by the around advice will be the return value seen by the caller of the method. A simple caching aspect for example could return a value from a cache if it has one, and invoke proceed() if it does not. Note that proceed may be invoked once, many times, or not at all within the body of the around advice, all of these are quite legal.

Advice parameters

Spring offers fully typed advice - meaning that you declare the parameters you need in the advice signature (as we saw for the returning and throwing examples above) rather than work with Object[] arrays all the time. We’ll see how to make argument and other contextual values available to the advice body in a moment. First let’s take a look at how to write generic advice that can find out about the method the advice is currently advising.

Access to the current JoinPoint

Any advice method may declare as its first parameter, a parameter of type org.aspectj.lang.JoinPoint (please note that around advice is required to declare a first parameter of type ProceedingJoinPoint, which is a subclass of JoinPoint. The JoinPoint interface provides a number of useful methods such as getArgs()(returns the method arguments), getThis() (returns the proxy object), getTarget() (returns the target object), getSignature() (returns a description of the method that is being advised) and toString() (prints a useful description of the method being advised). Please do consult the javadocs for full details.

Passing parameters to advice

We’ve already seen how to bind the returned value or exception value (using after returning and after throwing advice). To make argument values available to the advice body, you can use the binding form of args. If a parameter name is used in place of a type name in an args expression, then the value of the corresponding argument will be passed as the parameter value when the advice is invoked. An example should make this clearer. Suppose you want to advise the execution of dao operations that take an Account object as the first parameter, and you need access to the account in the advice body. You could write the following:

@Before("com.xyz.myapp.SystemArchitecture.dataAccessOperation() && args(account,..)")
public void validateAccount(Account account) {
    // ...
}

The args(account,..) part of the pointcut expression serves two purposes: firstly, it restricts matching to only those method executions where the method takes at least one parameter, and the argument passed to that parameter is an instance of Account; secondly, it makes the actual Account object available to the advice via the account parameter.

Another way of writing this is to declare a pointcut that "provides" the Account object value when it matches a join point, and then just refer to the named pointcut from the advice. This would look as follows:

@Pointcut("com.xyz.myapp.SystemArchitecture.dataAccessOperation() && args(account,..)")
private void accountDataAccessOperation(Account account) {}

@Before("accountDataAccessOperation(account)")
public void validateAccount(Account account) {
    // ...
}

The interested reader is once more referred to the AspectJ programming guide for more details.

The proxy object ( this), target object ( target), and annotations ( @within, @target, @annotation, @args) can all be bound in a similar fashion. The following example shows how you could match the execution of methods annotated with an @Auditable annotation, and extract the audit code.

First the definition of the @Auditable annotation:

@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.METHOD)
public @interface Auditable {
    AuditCode value();
}

And then the advice that matches the execution of @Auditable methods:

@Before("com.xyz.lib.Pointcuts.anyPublicMethod() && @annotation(auditable)")
public void audit(Auditable auditable) {
    AuditCode code = auditable.value();
    // ...
}
Advice parameters and generics

Spring AOP can handle generics used in class declarations and method parameters. Suppose you have a generic type like this:

public interface Sample<T> {
    void sampleGenericMethod(T param);
    void sampleGenericCollectionMethod(Collection<T> param);
}

You can restrict interception of method types to certain parameter types by simply typing the advice parameter to the parameter type you want to intercept the method for:

@Before("execution(* ..Sample+.sampleGenericMethod(*)) && args(param)")
public void beforeSampleMethod(MyType param) {
    // Advice implementation
}

That this works is pretty obvious as we already discussed above. However, it’s worth pointing out that this won’t work for generic collections. So you cannot define a pointcut like this:

@Before("execution(* ..Sample+.sampleGenericCollectionMethod(*)) && args(param)")
public void beforeSampleMethod(Collection<MyType> param) {
    // Advice implementation
}

To make this work we would have to inspect every element of the collection, which is not reasonable as we also cannot decide how to treat null values in general. To achieve something similar to this you have to type the parameter to Collection<?> and manually check the type of the elements.

Determining argument names

The parameter binding in advice invocations relies on matching names used in pointcut expressions to declared parameter names in (advice and pointcut) method signatures. Parameter names are not available through Java reflection, so Spring AOP uses the following strategies to determine parameter names:

  • If the parameter names have been specified by the user explicitly, then the specified parameter names are used: both the advice and the pointcut annotations have an optional "argNames" attribute which can be used to specify the argument names of the annotated method - these argument names are available at runtime. For example: 
@Before(value="com.xyz.lib.Pointcuts.anyPublicMethod() && target(bean) && @annotation(auditable)",
        argNames="bean,auditable")
public void audit(Object bean, Auditable auditable) {
    AuditCode code = auditable.value();
    // ... use code and bean
}

If the first parameter is of the JoinPoint, ProceedingJoinPoint, or JoinPoint.StaticPart type, you may leave out the name of the parameter from the value of the "argNames" attribute. For example, if you modify the preceding advice to receive the join point object, the "argNames" attribute need not include it:

@Before(value="com.xyz.lib.Pointcuts.anyPublicMethod() && target(bean) && @annotation(auditable)",
        argNames="bean,auditable")
public void audit(JoinPoint jp, Object bean, Auditable auditable) {
    AuditCode code = auditable.value();
    // ... use code, bean, and jp
}

The special treatment given to the first parameter of the JoinPoint, ProceedingJoinPoint, and JoinPoint.StaticPart types is particularly convenient for advice that do not collect any other join point context. In such situations, you may simply omit the "argNames" attribute. For example, the following advice need not declare the "argNames" attribute:

@Before("com.xyz.lib.Pointcuts.anyPublicMethod()")
public void audit(JoinPoint jp) {
    // ... use jp
}
  • Using the 'argNames' attribute is a little clumsy, so if the 'argNames' attribute has not been specified, then Spring AOP will look at the debug information for the class and try to determine the parameter names from the local variable table. This information will be present as long as the classes have been compiled with debug information ( '-g:vars' at a minimum). The consequences of compiling with this flag on are: (1) your code will be slightly easier to understand (reverse engineer), (2) the class file sizes will be very slightly bigger (typically inconsequential), (3) the optimization to remove unused local variables will not be applied by your compiler. In other words, you should encounter no difficulties building with this flag on.
Spring框架参考文档 Spring Framework Reference Documentation

If an @AspectJ aspect has been compiled by the AspectJ compiler (ajc) even without the debug information then there is no need to add the argNames attribute as the compiler will retain the needed information.
  • If the code has been compiled without the necessary debug information, then Spring AOP will attempt to deduce the pairing of binding variables to parameters (for example, if only one variable is bound in the pointcut expression, and the advice method only takes one parameter, the pairing is obvious!). If the binding of variables is ambiguous given the available information, then an AmbiguousBindingException will be thrown.
  • If all of the above strategies fail then an IllegalArgumentException will be thrown.
Proceeding with arguments

We remarked earlier that we would describe how to write a proceed call with arguments that works consistently across Spring AOP and AspectJ. The solution is simply to ensure that the advice signature binds each of the method parameters in order. For example:

@Around("execution(List<Account> find*(..)) && " +
        "com.xyz.myapp.SystemArchitecture.inDataAccessLayer() && " +
        "args(accountHolderNamePattern)")
public Object preProcessQueryPattern(ProceedingJoinPoint pjp,
        String accountHolderNamePattern) throws Throwable {
    String newPattern = preProcess(accountHolderNamePattern);
    return pjp.proceed(new Object[] {newPattern});
}

In many cases you will be doing this binding anyway (as in the example above).

Advice ordering

What happens when multiple pieces of advice all want to run at the same join point? Spring AOP follows the same precedence rules as AspectJ to determine the order of advice execution. The highest precedence advice runs first "on the way in" (so given two pieces of before advice, the one with highest precedence runs first). "On the way out" from a join point, the highest precedence advice runs last (so given two pieces of after advice, the one with the highest precedence will run second).

When two pieces of advice defined in different aspects both need to run at the same join point, unless you specify otherwise the order of execution is undefined. You can control the order of execution by specifying precedence. This is done in the normal Spring way by either implementing the org.springframework.core.Orderedinterface in the aspect class or annotating it with the Order annotation. Given two aspects, the aspect returning the lower value from Ordered.getValue() (or the annotation value) has the higher precedence.

When two pieces of advice defined in the same aspect both need to run at the same join point, the ordering is undefined (since there is no way to retrieve the declaration order via reflection for javac-compiled classes). Consider collapsing such advice methods into one advice method per join point in each aspect class, or refactor the pieces of advice into separate aspect classes - which can be ordered at the aspect level.

11.2.5 Introductions

Introductions (known as inter-type declarations in AspectJ) enable an aspect to declare that advised objects implement a given interface, and to provide an implementation of that interface on behalf of those objects.

An introduction is made using the @DeclareParents annotation. This annotation is used to declare that matching types have a new parent (hence the name). For example, given an interface UsageTracked, and an implementation of that interface DefaultUsageTracked, the following aspect declares that all implementors of service interfaces also implement the UsageTracked interface. (In order to expose statistics via JMX for example.)

@Aspect
public class UsageTracking {

    @DeclareParents(value="com.xzy.myapp.service.*+", defaultImpl=DefaultUsageTracked.class)
    public static UsageTracked mixin;

    @Before("com.xyz.myapp.SystemArchitecture.businessService() && this(usageTracked)")
    public void recordUsage(UsageTracked usageTracked) {
        usageTracked.incrementUseCount();
    }

}

The interface to be implemented is determined by the type of the annotated field. The value attribute of the @DeclareParents annotation is an AspectJ type pattern :- any bean of a matching type will implement the UsageTracked interface. Note that in the before advice of the above example, service beans can be directly used as implementations of the UsageTracked interface. If accessing a bean programmatically you would write the following:

UsageTracked usageTracked = (UsageTracked) context.getBean("myService");

11.2.6 Aspect instantiation models

Spring框架参考文档 Spring Framework Reference Documentation

(This is an advanced topic, so if you are just starting out with AOP you can safely skip it until later.)

By default there will be a single instance of each aspect within the application context. AspectJ calls this the singleton instantiation model. It is possible to define aspects with alternate lifecycles :- Spring supports AspectJ’s perthis and pertarget instantiation models ( percflow, percflowbelow, and pertypewithin are not currently supported).

A "perthis" aspect is declared by specifying a perthis clause in the @Aspect annotation. Let’s look at an example, and then we’ll explain how it works.

@Aspect("perthis(com.xyz.myapp.SystemArchitecture.businessService())")
public class MyAspect {

    private int someState;

    @Before(com.xyz.myapp.SystemArchitecture.businessService())
    public void recordServiceUsage() {
        // ...
    }

}

The effect of the 'perthis' clause is that one aspect instance will be created for each unique service object executing a business service (each unique object bound to 'this' at join points matched by the pointcut expression). The aspect instance is created the first time that a method is invoked on the service object. The aspect goes out of scope when the service object goes out of scope. Before the aspect instance is created, none of the advice within it executes. As soon as the aspect instance has been created, the advice declared within it will execute at matched join points, but only when the service object is the one this aspect is associated with. See the AspectJ programming guide for more information on per-clauses.

The 'pertarget' instantiation model works in exactly the same way as perthis, but creates one aspect instance for each unique target object at matched join points.

11.2.7 Example

Now that you have seen how all the constituent parts work, let’s put them together to do something useful!

The execution of business services can sometimes fail due to concurrency issues (for example, deadlock loser). If the operation is retried, it is quite likely to succeed next time round. For business services where it is appropriate to retry in such conditions (idempotent operations that don’t need to go back to the user for conflict resolution), we’d like to transparently retry the operation to avoid the client seeing a PessimisticLockingFailureException. This is a requirement that clearly cuts across multiple services in the service layer, and hence is ideal for implementing via an aspect.

Because we want to retry the operation, we will need to use around advice so that we can call proceed multiple times. Here’s how the basic aspect implementation looks:

@Aspect
public class ConcurrentOperationExecutor implements Ordered {

    private static final int DEFAULT_MAX_RETRIES = 2;

    private int maxRetries = DEFAULT_MAX_RETRIES;
    private int order = 1;

    public void setMaxRetries(int maxRetries) {
        this.maxRetries = maxRetries;
    }

    public int getOrder() {
        return this.order;
    }

    public void setOrder(int order) {
        this.order = order;
    }

    @Around("com.xyz.myapp.SystemArchitecture.businessService()")
    public Object doConcurrentOperation(ProceedingJoinPoint pjp) throws Throwable {
        int numAttempts = 0;
        PessimisticLockingFailureException lockFailureException;
        do {
            numAttempts++;
            try {
                return pjp.proceed();
            }
            catch(PessimisticLockingFailureException ex) {
                lockFailureException = ex;
            }
        } while(numAttempts <= this.maxRetries);
        throw lockFailureException;
    }

}

Note that the aspect implements the Ordered interface so we can set the precedence of the aspect higher than the transaction advice (we want a fresh transaction each time we retry). The maxRetries and order properties will both be configured by Spring. The main action happens in the doConcurrentOperation around advice. Notice that for the moment we’re applying the retry logic to all businessService()s. We try to proceed, and if we fail with anPessimisticLockingFailureException we simply try again unless we have exhausted all of our retry attempts.

The corresponding Spring configuration is:

<aop:aspectj-autoproxy/>

<bean id="concurrentOperationExecutor" class="com.xyz.myapp.service.impl.ConcurrentOperationExecutor">
    <property name="maxRetries" value="3"/>
    <property name="order" value="100"/>
</bean>

To refine the aspect so that it only retries idempotent operations, we might define an Idempotent annotation:

@Retention(RetentionPolicy.RUNTIME)
public @interface Idempotent {
    // marker annotation
}

and use the annotation to annotate the implementation of service operations. The change to the aspect to only retry idempotent operations simply involves refining the pointcut expression so that only @Idempotent operations match:

@Around("com.xyz.myapp.SystemArchitecture.businessService() && " +
        "@annotation(com.xyz.myapp.service.Idempotent)")
public Object doConcurrentOperation(ProceedingJoinPoint pjp) throws Throwable {
    ...
}

11.3 Schema-based AOP support

If you prefer an XML-based format, then Spring also offers support for defining aspects using the new "aop" namespace tags. The exact same pointcut expressions and advice kinds are supported as when using the @AspectJ style, hence in this section we will focus on the new syntax and refer the reader to the discussion in the previous section (Section 11.2, “@AspectJ support”) for an understanding of writing pointcut expressions and the binding of advice parameters.

To use the aop namespace tags described in this section, you need to import the spring-aop schema as described in Chapter 41, XML Schema-based configuration. See Section 41.2.7, “the aop schema” for how to import the tags in the aop namespace.

Within your Spring configurations, all aspect and advisor elements must be placed within an <aop:config> element (you can have more than one <aop:config>element in an application context configuration). An <aop:config> element can contain pointcut, advisor, and aspect elements (note these must be declared in that order).

Spring框架参考文档 Spring Framework Reference Documentation

The <aop:config> style of configuration makes heavy use of Spring’s auto-proxying mechanism. This can cause issues (such as advice not being woven) if you are already using explicit auto-proxying via the use of BeanNameAutoProxyCreator or suchlike. The recommended usage pattern is to use either just the <aop:config> style, or just the AutoProxyCreator style.

11.3.1 Declaring an aspect

Using the schema support, an aspect is simply a regular Java object defined as a bean in your Spring application context. The state and behavior is captured in the fields and methods of the object, and the pointcut and advice information is captured in the XML.

An aspect is declared using the <aop:aspect> element, and the backing bean is referenced using the ref attribute:

<aop:config>
    <aop:aspect id="myAspect" ref="aBean">
        ...
    </aop:aspect>
</aop:config>

<bean id="aBean" class="...">
    ...
</bean>

The bean backing the aspect (" `aBean`" in this case) can of course be configured and dependency injected just like any other Spring bean.

11.3.2 Declaring a pointcut

A named pointcut can be declared inside an <aop:config> element, enabling the pointcut definition to be shared across several aspects and advisors.

A pointcut representing the execution of any business service in the service layer could be defined as follows:

<aop:config>

    <aop:pointcut id="businessService"
        expression="execution(* com.xyz.myapp.service.*.*(..))"/>

</aop:config>

Note that the pointcut expression itself is using the same AspectJ pointcut expression language as described in Section 11.2, “@AspectJ support”. If you are using the schema based declaration style, you can refer to named pointcuts defined in types (@Aspects) within the pointcut expression. Another way of defining the above pointcut would be:

<aop:config>

    <aop:pointcut id="businessService"
        expression="com.xyz.myapp.SystemArchitecture.businessService()"/>

</aop:config>

Assuming you have a SystemArchitecture aspect as described in the section called “Sharing common pointcut definitions”.

Declaring a pointcut inside an aspect is very similar to declaring a top-level pointcut:

<aop:config>

    <aop:aspect id="myAspect" ref="aBean">

        <aop:pointcut id="businessService"
            expression="execution(* com.xyz.myapp.service.*.*(..))"/>

        ...

    </aop:aspect>

</aop:config>

Much the same way in an @AspectJ aspect, pointcuts declared using the schema based definition style may collect join point context. For example, the following pointcut collects the 'this' object as the join point context and passes it to advice:

<aop:config>

    <aop:aspect id="myAspect" ref="aBean">

        <aop:pointcut id="businessService"
            expression="execution(* com.xyz.myapp.service.*.*(..)) &amp;&amp; this(service)"/>

        <aop:before pointcut-ref="businessService" method="monitor"/>

        ...

    </aop:aspect>

</aop:config>

The advice must be declared to receive the collected join point context by including parameters of the matching names:

public void monitor(Object service) {
    ...
}

When combining pointcut sub-expressions, '&&' is awkward within an XML document, and so the keywords 'and', 'or' and 'not' can be used in place of '&&', '||' and '!' respectively. For example, the previous pointcut may be better written as:

<aop:config>

    <aop:aspect id="myAspect" ref="aBean">

        <aop:pointcut id="businessService"
            expression="execution(* com.xyz.myapp.service.*.*(..)) **and** this(service)"/>

        <aop:before pointcut-ref="businessService" method="monitor"/>

        ...
    </aop:aspect>
</aop:config>

Note that pointcuts defined in this way are referred to by their XML id and cannot be used as named pointcuts to form composite pointcuts. The named pointcut support in the schema based definition style is thus more limited than that offered by the @AspectJ style.

11.3.3 Declaring advice

The same five advice kinds are supported as for the @AspectJ style, and they have exactly the same semantics.

Before advice

Before advice runs before a matched method execution. It is declared inside an <aop:aspect> using the <aop:before> element.

<aop:aspect id="beforeExample" ref="aBean">

    <aop:before
        pointcut-ref="dataAccessOperation"
        method="doAccessCheck"/>

    ...

</aop:aspect>

Here dataAccessOperation is the id of a pointcut defined at the top ( <aop:config>) level. To define the pointcut inline instead, replace the pointcut-ref attribute with a pointcut attribute:

<aop:aspect id="beforeExample" ref="aBean">

    <aop:before
        pointcut="execution(* com.xyz.myapp.dao.*.*(..))"
        method="doAccessCheck"/>

    ...

</aop:aspect>

As we noted in the discussion of the @AspectJ style, using named pointcuts can significantly improve the readability of your code.

The method attribute identifies a method ( doAccessCheck) that provides the body of the advice. This method must be defined for the bean referenced by the aspect element containing the advice. Before a data access operation is executed (a method execution join point matched by the pointcut expression), the "doAccessCheck" method on the aspect bean will be invoked.

After returning advice

After returning advice runs when a matched method execution completes normally. It is declared inside an <aop:aspect> in the same way as before advice. For example:

<aop:aspect id="afterReturningExample" ref="aBean">

    <aop:after-returning
        pointcut-ref="dataAccessOperation"
        method="doAccessCheck"/>

    ...

</aop:aspect>

Just as in the @AspectJ style, it is possible to get hold of the return value within the advice body. Use the returning attribute to specify the name of the parameter to which the return value should be passed:

<aop:aspect id="afterReturningExample" ref="aBean">

    <aop:after-returning
        pointcut-ref="dataAccessOperation"
        returning="retVal"
        method="doAccessCheck"/>

    ...

</aop:aspect>

The doAccessCheck method must declare a parameter named retVal. The type of this parameter constrains matching in the same way as described for @AfterReturning. For example, the method signature may be declared as:

public void doAccessCheck(Object retVal) {...

After throwing advice

After throwing advice executes when a matched method execution exits by throwing an exception. It is declared inside an <aop:aspect> using the after-throwing element:

<aop:aspect id="afterThrowingExample" ref="aBean">

    <aop:after-throwing
        pointcut-ref="dataAccessOperation"
        method="doRecoveryActions"/>

    ...

</aop:aspect>

Just as in the @AspectJ style, it is possible to get hold of the thrown exception within the advice body. Use the throwing attribute to specify the name of the parameter to which the exception should be passed:

<aop:aspect id="afterThrowingExample" ref="aBean">

    <aop:after-throwing
        pointcut-ref="dataAccessOperation"
        throwing="dataAccessEx"
        method="doRecoveryActions"/>

    ...

</aop:aspect>

The doRecoveryActions method must declare a parameter named dataAccessEx. The type of this parameter constrains matching in the same way as described for @AfterThrowing. For example, the method signature may be declared as:

public void doRecoveryActions(DataAccessException dataAccessEx) {...

After (finally) advice

After (finally) advice runs however a matched method execution exits. It is declared using the after element:

<aop:aspect id="afterFinallyExample" ref="aBean">

    <aop:after
        pointcut-ref="dataAccessOperation"
        method="doReleaseLock"/>

    ...

</aop:aspect>

Around advice

The final kind of advice is around advice. Around advice runs "around" a matched method execution. It has the opportunity to do work both before and after the method executes, and to determine when, how, and even if, the method actually gets to execute at all. Around advice is often used if you need to share state before and after a method execution in a thread-safe manner (starting and stopping a timer for example). Always use the least powerful form of advice that meets your requirements; don’t use around advice if simple before advice would do.

Around advice is declared using the aop:around element. The first parameter of the advice method must be of type ProceedingJoinPoint. Within the body of the advice, calling proceed() on the ProceedingJoinPoint causes the underlying method to execute. The proceed method may also be calling passing in anObject[] - the values in the array will be used as the arguments to the method execution when it proceeds. See the section called “Around advice” for notes on calling proceed with an Object[].

<aop:aspect id="aroundExample" ref="aBean">

    <aop:around
        pointcut-ref="businessService"
        method="doBasicProfiling"/>

    ...

</aop:aspect>

The implementation of the doBasicProfiling advice would be exactly the same as in the @AspectJ example (minus the annotation of course):

public Object doBasicProfiling(ProceedingJoinPoint pjp) throws Throwable {
    // start stopwatch
    Object retVal = pjp.proceed();
    // stop stopwatch
    return retVal;
}

Advice parameters

The schema based declaration style supports fully typed advice in the same way as described for the @AspectJ support - by matching pointcut parameters by name against advice method parameters. See the section called “Advice parameters” for details. If you wish to explicitly specify argument names for the advice methods (not relying on the detection strategies previously described) then this is done using the arg-names attribute of the advice element, which is treated in the same manner to the "argNames" attribute in an advice annotation as described in the section called “Determining argument names”. For example:

<aop:before
    pointcut="com.xyz.lib.Pointcuts.anyPublicMethod() and @annotation(auditable)"
    method="audit"
    arg-names="auditable"/>

The arg-names attribute accepts a comma-delimited list of parameter names.

Find below a slightly more involved example of the XSD-based approach that illustrates some around advice used in conjunction with a number of strongly typed parameters.

package x.y.service;

public interface FooService {

    Foo getFoo(String fooName, int age);
}

public class DefaultFooService implements FooService {

    public Foo getFoo(String name, int age) {
        return new Foo(name, age);
    }
}

Next up is the aspect. Notice the fact that the profile(..) method accepts a number of strongly-typed parameters, the first of which happens to be the join point used to proceed with the method call: the presence of this parameter is an indication that the profile(..) is to be used as around advice:

package x.y;

import org.aspectj.lang.ProceedingJoinPoint;
import org.springframework.util.StopWatch;

public class SimpleProfiler {

    public Object profile(ProceedingJoinPoint call, String name, int age) throws Throwable {
        StopWatch clock = new StopWatch("Profiling for '" + name + "' and '" + age + "'");
        try {
            clock.start(call.toShortString());
            return call.proceed();
        } finally {
            clock.stop();
            System.out.println(clock.prettyPrint());
        }
    }
}

Finally, here is the XML configuration that is required to effect the execution of the above advice for a particular join point:

<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:aop="http://www.springframework.org/schema/aop"
    xsi:schemaLocation="
        http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/aop http://www.springframework.org/schema/aop/spring-aop.xsd">

    <!-- this is the object that will be proxied by Spring's AOP infrastructure -->
    <bean id="fooService" class="x.y.service.DefaultFooService"/>

    <!-- this is the actual advice itself -->
    <bean id="profiler" class="x.y.SimpleProfiler"/>

    <aop:config>
        <aop:aspect ref="profiler">

            <aop:pointcut id="theExecutionOfSomeFooServiceMethod"
                expression="execution(* x.y.service.FooService.getFoo(String,int))
                and args(name, age)"/>

            <aop:around pointcut-ref="theExecutionOfSomeFooServiceMethod"
                method="profile"/>

        </aop:aspect>
    </aop:config>

</beans>

If we had the following driver script, we would get output something like this on standard output:

import org.springframework.beans.factory.BeanFactory;
import org.springframework.context.support.ClassPathXmlApplicationContext;
import x.y.service.FooService;

public final class Boot {

    public static void main(final String[] args) throws Exception {
        BeanFactory ctx = new ClassPathXmlApplicationContext("x/y/plain.xml");
        FooService foo = (FooService) ctx.getBean("fooService");
        foo.getFoo("Pengo", 12);
    }
}
StopWatch 'Profiling for 'Pengo' and '12'': running time (millis) = 0
-----------------------------------------
ms     %     Task name
-----------------------------------------
00000  ?  execution(getFoo)

Advice ordering

When multiple advice needs to execute at the same join point (executing method) the ordering rules are as described in the section called “Advice ordering”. The precedence between aspects is determined by either adding the Order annotation to the bean backing the aspect or by having the bean implement the Orderedinterface.

11.3.4 Introductions

Introductions (known as inter-type declarations in AspectJ) enable an aspect to declare that advised objects implement a given interface, and to provide an implementation of that interface on behalf of those objects.

An introduction is made using the aop:declare-parents element inside an aop:aspect This element is used to declare that matching types have a new parent (hence the name). For example, given an interface UsageTracked, and an implementation of that interface DefaultUsageTracked, the following aspect declares that all implementors of service interfaces also implement the UsageTracked interface. (In order to expose statistics via JMX for example.)

<aop:aspect id="usageTrackerAspect" ref="usageTracking">

    <aop:declare-parents
        types-matching="com.xzy.myapp.service.*+"
        implement-interface="com.xyz.myapp.service.tracking.UsageTracked"
        default-impl="com.xyz.myapp.service.tracking.DefaultUsageTracked"/>

    <aop:before
        pointcut="com.xyz.myapp.SystemArchitecture.businessService()
            and this(usageTracked)"
            method="recordUsage"/>

</aop:aspect>

The class backing the usageTracking bean would contain the method:

public void recordUsage(UsageTracked usageTracked) {
    usageTracked.incrementUseCount();
}

The interface to be implemented is determined by implement-interface attribute. The value of the types-matching attribute is an AspectJ type pattern :- any bean of a matching type will implement the UsageTracked interface. Note that in the before advice of the above example, service beans can be directly used as implementations of the UsageTracked interface. If accessing a bean programmatically you would write the following:

UsageTracked usageTracked = (UsageTracked) context.getBean("myService");

11.3.5 Aspect instantiation models

The only supported instantiation model for schema-defined aspects is the singleton model. Other instantiation models may be supported in future releases.

11.3.6 Advisors

The concept of "advisors" is brought forward from the AOP support defined in Spring 1.2 and does not have a direct equivalent in AspectJ. An advisor is like a small self-contained aspect that has a single piece of advice. The advice itself is represented by a bean, and must implement one of the advice interfaces described inSection 12.3.2, “Advice types in Spring”. Advisors can take advantage of AspectJ pointcut expressions though.

Spring supports the advisor concept with the <aop:advisor> element. You will most commonly see it used in conjunction with transactional advice, which also has its own namespace support in Spring. Here’s how it looks:

<aop:config>

    <aop:pointcut id="businessService"
        expression="execution(* com.xyz.myapp.service.*.*(..))"/>

    <aop:advisor
        pointcut-ref="businessService"
        advice-ref="tx-advice"/>

</aop:config>

<tx:advice id="tx-advice">
    <tx:attributes>
        <tx:method name="*" propagation="REQUIRED"/>
    </tx:attributes>
</tx:advice>

As well as the pointcut-ref attribute used in the above example, you can also use the pointcut attribute to define a pointcut expression inline.

To define the precedence of an advisor so that the advice can participate in ordering, use the order attribute to define the Ordered value of the advisor.

11.3.7 Example

Let’s see how the concurrent locking failure retry example from Section 11.2.7, “Example” looks when rewritten using the schema support.

The execution of business services can sometimes fail due to concurrency issues (for example, deadlock loser). If the operation is retried, it is quite likely it will succeed next time round. For business services where it is appropriate to retry in such conditions (idempotent operations that don’t need to go back to the user for conflict resolution), we’d like to transparently retry the operation to avoid the client seeing a PessimisticLockingFailureException. This is a requirement that clearly cuts across multiple services in the service layer, and hence is ideal for implementing via an aspect.

Because we want to retry the operation, we’ll need to use around advice so that we can call proceed multiple times. Here’s how the basic aspect implementation looks (it’s just a regular Java class using the schema support):

public class ConcurrentOperationExecutor implements Ordered {

    private static final int DEFAULT_MAX_RETRIES = 2;

    private int maxRetries = DEFAULT_MAX_RETRIES;
    private int order = 1;

    public void setMaxRetries(int maxRetries) {
        this.maxRetries = maxRetries;
    }

    public int getOrder() {
        return this.order;
    }

    public void setOrder(int order) {
        this.order = order;
    }

    public Object doConcurrentOperation(ProceedingJoinPoint pjp) throws Throwable {
        int numAttempts = 0;
        PessimisticLockingFailureException lockFailureException;
        do {
            numAttempts++;
            try {
                return pjp.proceed();
            }
            catch(PessimisticLockingFailureException ex) {
                lockFailureException = ex;
            }
        } while(numAttempts <= this.maxRetries);
        throw lockFailureException;
    }

}

Note that the aspect implements the Ordered interface so we can set the precedence of the aspect higher than the transaction advice (we want a fresh transaction each time we retry). The maxRetries and order properties will both be configured by Spring. The main action happens in the doConcurrentOperation around advice method. We try to proceed, and if we fail with a PessimisticLockingFailureException we simply try again unless we have exhausted all of our retry attempts.

Spring框架参考文档 Spring Framework Reference Documentation

This class is identical to the one used in the @AspectJ example, but with the annotations removed.

The corresponding Spring configuration is:

<aop:config>

    <aop:aspect id="concurrentOperationRetry" ref="concurrentOperationExecutor">

        <aop:pointcut id="idempotentOperation"
            expression="execution(* com.xyz.myapp.service.*.*(..))"/>

        <aop:around
            pointcut-ref="idempotentOperation"
            method="doConcurrentOperation"/>

    </aop:aspect>

</aop:config>

<bean id="concurrentOperationExecutor"
    class="com.xyz.myapp.service.impl.ConcurrentOperationExecutor">
        <property name="maxRetries" value="3"/>
        <property name="order" value="100"/>
</bean>

Notice that for the time being we assume that all business services are idempotent. If this is not the case we can refine the aspect so that it only retries genuinely idempotent operations, by introducing an Idempotent annotation:

@Retention(RetentionPolicy.RUNTIME)
public @interface Idempotent {
    // marker annotation
}

and using the annotation to annotate the implementation of service operations. The change to the aspect to retry only idempotent operations simply involves refining the pointcut expression so that only @Idempotent operations match:

<aop:pointcut id="idempotentOperation"
        expression="execution(* com.xyz.myapp.service.*.*(..)) and
        @annotation(com.xyz.myapp.service.Idempotent)"/>

11.4 Choosing which AOP declaration style to use

Once you have decided that an aspect is the best approach for implementing a given requirement, how do you decide between using Spring AOP or AspectJ, and between the Aspect language (code) style, @AspectJ annotation style, or the Spring XML style? These decisions are influenced by a number of factors including application requirements, development tools, and team familiarity with AOP.

11.4.1 Spring AOP or full AspectJ?

Use the simplest thing that can work. Spring AOP is simpler than using full AspectJ as there is no requirement to introduce the AspectJ compiler / weaver into your development and build processes. If you only need to advise the execution of operations on Spring beans, then Spring AOP is the right choice. If you need to advise objects not managed by the Spring container (such as domain objects typically), then you will need to use AspectJ. You will also need to use AspectJ if you wish to advise join points other than simple method executions (for example, field get or set join points, and so on).

When using AspectJ, you have the choice of the AspectJ language syntax (also known as the "code style") or the @AspectJ annotation style. Clearly, if you are not using Java 5+ then the choice has been made for you…​ use the code style. If aspects play a large role in your design, and you are able to use the AspectJ Development Tools (AJDT) plugin for Eclipse, then the AspectJ language syntax is the preferred option: it is cleaner and simpler because the language was purposefully designed for writing aspects. If you are not using Eclipse, or have only a few aspects that do not play a major role in your application, then you may want to consider using the @AspectJ style and sticking with a regular Java compilation in your IDE, and adding an aspect weaving phase to your build script.

11.4.2 @AspectJ or XML for Spring AOP?

If you have chosen to use Spring AOP, then you have a choice of @AspectJ or XML style. There are various tradeoffs to consider.

The XML style will be most familiar to existing Spring users and it is backed by genuine POJOs. When using AOP as a tool to configure enterprise services then XML can be a good choice (a good test is whether you consider the pointcut expression to be a part of your configuration you might want to change independently). With the XML style arguably it is clearer from your configuration what aspects are present in the system.

The XML style has two disadvantages. Firstly it does not fully encapsulate the implementation of the requirement it addresses in a single place. The DRY principle says that there should be a single, unambiguous, authoritative representation of any piece of knowledge within a system. When using the XML style, the knowledge of how a requirement is implemented is split across the declaration of the backing bean class, and the XML in the configuration file. When using the @AspectJ style there is a single module - the aspect - in which this information is encapsulated. Secondly, the XML style is slightly more limited in what it can express than the @AspectJ style: only the "singleton" aspect instantiation model is supported, and it is not possible to combine named pointcuts declared in XML. For example, in the @AspectJ style you can write something like:

@Pointcut(execution(* get*()))
public void propertyAccess() {}

@Pointcut(execution(org.xyz.Account+ *(..))
public void operationReturningAnAccount() {}

@Pointcut(propertyAccess() && operationReturningAnAccount())
public void accountPropertyAccess() {}

In the XML style I can declare the first two pointcuts:

<aop:pointcut id="propertyAccess"
        expression="execution(* get*())"/>

<aop:pointcut id="operationReturningAnAccount"
        expression="execution(org.xyz.Account+ *(..))"/>

The downside of the XML approach is that you cannot define the accountPropertyAccess pointcut by combining these definitions.

The @AspectJ style supports additional instantiation models, and richer pointcut composition. It has the advantage of keeping the aspect as a modular unit. It also has the advantage the @AspectJ aspects can be understood (and thus consumed) both by Spring AOP and by AspectJ - so if you later decide you need the capabilities of AspectJ to implement additional requirements then it is very easy to migrate to an AspectJ-based approach. On balance the Spring team prefer the @AspectJ style whenever you have aspects that do more than simple "configuration" of enterprise services.

11.5 Mixing aspect types

It is perfectly possible to mix @AspectJ style aspects using the autoproxying support, schema-defined <aop:aspect> aspects, <aop:advisor> declared advisors and even proxies and interceptors defined using the Spring 1.2 style in the same configuration. All of these are implemented using the same underlying support mechanism and will co-exist without any difficulty.

11.6 Proxying mechanisms

Spring AOP uses either JDK dynamic proxies or CGLIB to create the proxy for a given target object. (JDK dynamic proxies are preferred whenever you have a choice).

If the target object to be proxied implements at least one interface then a JDK dynamic proxy will be used. All of the interfaces implemented by the target type will be proxied. If the target object does not implement any interfaces then a CGLIB proxy will be created.

If you want to force the use of CGLIB proxying (for example, to proxy every method defined for the target object, not just those implemented by its interfaces) you can do so. However, there are some issues to consider:

  • final methods cannot be advised, as they cannot be overridden.
  • As of Spring 3.2, it is no longer necessary to add CGLIB to your project classpath, as CGLIB classes are repackaged under org.springframework and included directly in the spring-core JAR. This means that CGLIB-based proxy support 'just works' in the same way that JDK dynamic proxies always have.
  • As of Spring 4.0, the constructor of your proxied object will NOT be called twice anymore since the CGLIB proxy instance will be created via Objenesis. Only if your JVM does not allow for constructor bypassing, you might see double invocations and corresponding debug log entries from Spring’s AOP support.

To force the use of CGLIB proxies set the value of the proxy-target-class attribute of the <aop:config> element to true:

<aop:config proxy-target-class="true">
    <!-- other beans defined here... -->
</aop:config>

To force CGLIB proxying when using the @AspectJ autoproxy support, set the 'proxy-target-class' attribute of the <aop:aspectj-autoproxy> element to true:

<aop:aspectj-autoproxy proxy-target-class="true"/>
Spring框架参考文档 Spring Framework Reference Documentation

Multiple <aop:config/> sections are collapsed into a single unified auto-proxy creator at runtime, which applies the strongest proxy settings that any of the <aop:config/> sections (typically from different XML bean definition files) specified. This also applies to the <tx:annotation-driven/> and<aop:aspectj-autoproxy/> elements.

To be clear: using proxy-target-class="true" on <tx:annotation-driven/>, <aop:aspectj-autoproxy/> or <aop:config/> elements will force the use of CGLIB proxies for all three of them.

11.6.1 Understanding AOP proxies

Spring AOP is proxy-based. It is vitally important that you grasp the semantics of what that last statement actually means before you write your own aspects or use any of the Spring AOP-based aspects supplied with the Spring Framework.

Consider first the scenario where you have a plain-vanilla, un-proxied, nothing-special-about-it, straight object reference, as illustrated by the following code snippet.

public class SimplePojo implements Pojo {

    public void foo() {
        // this next method invocation is a direct call on the 'this' reference
        this.bar();
    }

    public void bar() {
        // some logic...
    }
}

If you invoke a method on an object reference, the method is invoked directly on that object reference, as can be seen below.

Spring框架参考文档 Spring Framework Reference Documentation
public class Main {

    public static void main(String[] args) {

        Pojo pojo = new SimplePojo();

        // this is a direct method call on the 'pojo' reference
        pojo.foo();
    }
}

Things change slightly when the reference that client code has is a proxy. Consider the following diagram and code snippet.

Spring框架参考文档 Spring Framework Reference Documentation
public class Main {

    public static void main(String[] args) {

        ProxyFactory factory = new ProxyFactory(new SimplePojo());
        factory.addInterface(Pojo.class);
        factory.addAdvice(new RetryAdvice());

        Pojo pojo = (Pojo) factory.getProxy();

        // this is a method call on the proxy!
        pojo.foo();
    }
}

The key thing to understand here is that the client code inside the main(..) of the Main class has a reference to the proxy. This means that method calls on that object reference will be calls on the proxy, and as such the proxy will be able to delegate to all of the interceptors (advice) that are relevant to that particular method call. However, once the call has finally reached the target object, the SimplePojo reference in this case, any method calls that it may make on itself, such as this.bar() orthis.foo(), are going to be invoked against the this reference, and not the proxy. This has important implications. It means that self-invocation is not going to result in the advice associated with a method invocation getting a chance to execute.

Okay, so what is to be done about this? The best approach (the term best is used loosely here) is to refactor your code such that the self-invocation does not happen. For sure, this does entail some work on your part, but it is the best, least-invasive approach. The next approach is absolutely horrendous, and I am almost reticent to point it out precisely because it is so horrendous. You can (choke!) totally tie the logic within your class to Spring AOP by doing this:

public class SimplePojo implements Pojo {

    public void foo() {
        // this works, but... gah!
        ((Pojo) AopContext.currentProxy()).bar();
    }

    public void bar() {
        // some logic...
    }
}

This totally couples your code to Spring AOP, and it makes the class itself aware of the fact that it is being used in an AOP context, which flies in the face of AOP. It also requires some additional configuration when the proxy is being created:

public class Main {

    public static void main(String[] args) {

        ProxyFactory factory = new ProxyFactory(new SimplePojo());
        factory.adddInterface(Pojo.class);
        factory.addAdvice(new RetryAdvice());
        factory.setExposeProxy(true);

        Pojo pojo = (Pojo) factory.getProxy();

        // this is a method call on the proxy!
        pojo.foo();
    }
}

Finally, it must be noted that AspectJ does not have this self-invocation issue because it is not a proxy-based AOP framework.

11.7 Programmatic creation of @AspectJ Proxies

In addition to declaring aspects in your configuration using either <aop:config> or <aop:aspectj-autoproxy>, it is also possible programmatically to create proxies that advise target objects. For the full details of Spring’s AOP API, see the next chapter. Here we want to focus on the ability to automatically create proxies using @AspectJ aspects.

The class org.springframework.aop.aspectj.annotation.AspectJProxyFactory can be used to create a proxy for a target object that is advised by one or more @AspectJ aspects. Basic usage for this class is very simple, as illustrated below. See the javadocs for full information.

// create a factory that can generate a proxy for the given target object
AspectJProxyFactory factory = new AspectJProxyFactory(targetObject);

// add an aspect, the class must be an @AspectJ aspect
// you can call this as many times as you need with different aspects
factory.addAspect(SecurityManager.class);

// you can also add existing aspect instances, the type of the object supplied must be an @AspectJ aspect
factory.addAspect(usageTracker);

// now get the proxy object...
MyInterfaceType proxy = factory.getProxy();

11.8 Using AspectJ with Spring applications

Everything we’ve covered so far in this chapter is pure Spring AOP. In this section, we’re going to look at how you can use the AspectJ compiler/weaver instead of, or in addition to, Spring AOP if your needs go beyond the facilities offered by Spring AOP alone.

Spring ships with a small AspectJ aspect library, which is available standalone in your distribution as spring-aspects.jar; you’ll need to add this to your classpath in order to use the aspects in it. Section 11.8.1, “Using AspectJ to dependency inject domain objects with Spring” and Section 11.8.2, “Other Spring aspects for AspectJ”discuss the content of this library and how you can use it. Section 11.8.3, “Configuring AspectJ aspects using Spring IoC” discusses how to dependency inject AspectJ aspects that are woven using the AspectJ compiler. Finally, Section 11.8.4, “Load-time weaving with AspectJ in the Spring Framework” provides an introduction to load-time weaving for Spring applications using AspectJ.

11.8.1 Using AspectJ to dependency inject domain objects with Spring

The Spring container instantiates and configures beans defined in your application context. It is also possible to ask a bean factory to configure a pre-existing object given the name of a bean definition containing the configuration to be applied. The spring-aspects.jar contains an annotation-driven aspect that exploits this capability to allow dependency injection of any object. The support is intended to be used for objects created outside of the control of any container. Domain objects often fall into this category because they are often created programmatically using the new operator, or by an ORM tool as a result of a database query.

The @Configurable annotation marks a class as eligible for Spring-driven configuration. In the simplest case it can be used just as a marker annotation:

package com.xyz.myapp.domain;

import org.springframework.beans.factory.annotation.Configurable;

@Configurable
public class Account {
    // ...
}

When used as a marker interface in this way, Spring will configure new instances of the annotated type ( Account in this case) using a bean definition (typically prototype-scoped) with the same name as the fully-qualified type name ( com.xyz.myapp.domain.Account). Since the default name for a bean is the fully-qualified name of its type, a convenient way to declare the prototype definition is simply to omit the id attribute:

<bean class="com.xyz.myapp.domain.Account" scope="prototype">
    <property name="fundsTransferService" ref="fundsTransferService"/>
</bean>

If you want to explicitly specify the name of the prototype bean definition to use, you can do so directly in the annotation:

package com.xyz.myapp.domain;

import org.springframework.beans.factory.annotation.Configurable;

@Configurable("account")
public class Account {
    // ...
}

Spring will now look for a bean definition named "account" and use that as the definition to configure new Account instances.

You can also use autowiring to avoid having to specify a dedicated bean definition at all. To have Spring apply autowiring use the autowire property of the@Configurable annotation: specify either @Configurable(autowire=Autowire.BY_TYPE) or @Configurable(autowire=Autowire.BY_NAME for autowiring by type or by name respectively. As an alternative, as of Spring 2.5 it is preferable to specify explicit, annotation-driven dependency injection for your @Configurable beans by using @Autowired or @Inject at the field or method level (see Section 7.9, “Annotation-based container configuration” for further details).

Finally you can enable Spring dependency checking for the object references in the newly created and configured object by using the dependencyCheck attribute (for example: @Configurable(autowire=Autowire.BY_NAME,dependencyCheck=true)). If this attribute is set to true, then Spring will validate after configuration that all properties (which are not primitives or collections) have been set.

Using the annotation on its own does nothing of course. It is the AnnotationBeanConfigurerAspect in spring-aspects.jar that acts on the presence of the annotation. In essence the aspect says "after returning from the initialization of a new object of a type annotated with @Configurable, configure the newly created object using Spring in accordance with the properties of the annotation". In this context, initialization refers to newly instantiated objects (e.g., objects instantiated with thenew operator) as well as to Serializable objects that are undergoing deserialization (e.g., via readResolve()).

Spring框架参考文档 Spring Framework Reference Documentation

One of the key phrases in the above paragraph is 'in essence'. For most cases, the exact semantics of 'after returning from the initialization of a new object' will be fine…​ in this context, 'after initialization' means that the dependencies will be injected after the object has been constructed - this means that the dependencies will not be available for use in the constructor bodies of the class. If you want the dependencies to be injected before the constructor bodies execute, and thus be available for use in the body of the constructors, then you need to define this on the @Configurable declaration like so:

@Configurable(preConstruction=true)

You can find out more information about the language semantics of the various pointcut types in AspectJ in this appendix of the AspectJ Programming Guide.

For this to work the annotated types must be woven with the AspectJ weaver - you can either use a build-time Ant or Maven task to do this (see for example the AspectJ Development Environment Guide) or load-time weaving (see Section 11.8.4, “Load-time weaving with AspectJ in the Spring Framework”). TheAnnotationBeanConfigurerAspect itself needs configuring by Spring (in order to obtain a reference to the bean factory that is to be used to configure new objects). If you are using Java based configuration simply add @EnableSpringConfigured to any @Configuration class.

@Configuration
@EnableSpringConfigured
public class AppConfig {

}

If you prefer XML based configuration, the Spring context namespace defines a convenient context:spring-configured element:

<context:spring-configured/>

Instances of @Configurable objects created before the aspect has been configured will result in a message being issued to the debug log and no configuration of the object taking place. An example might be a bean in the Spring configuration that creates domain objects when it is initialized by Spring. In this case you can use the "depends-on" bean attribute to manually specify that the bean depends on the configuration aspect.

<bean id="myService"
        class="com.xzy.myapp.service.MyService"
        depends-on="org.springframework.beans.factory.aspectj.AnnotationBeanConfigurerAspect">

    <!-- ... -->

</bean>
Spring框架参考文档 Spring Framework Reference Documentation

Do not activate @Configurable processing through the bean configurer aspect unless you really mean to rely on its semantics at runtime. In particular, make sure that you do not use @Configurable on bean classes which are registered as regular Spring beans with the container: You would get double initialization otherwise, once through the container and once through the aspect.

Unit testing @Configurable objects

One of the goals of the @Configurable support is to enable independent unit testing of domain objects without the difficulties associated with hard-coded lookups. If@Configurable types have not been woven by AspectJ then the annotation has no affect during unit testing, and you can simply set mock or stub property references in the object under test and proceed as normal. If @Configurable types have been woven by AspectJ then you can still unit test outside of the container as normal, but you will see a warning message each time that you construct an @Configurable object indicating that it has not been configured by Spring.

Working with multiple application contexts

The AnnotationBeanConfigurerAspect used to implement the @Configurable support is an AspectJ singleton aspect. The scope of a singleton aspect is the same as the scope of static members, that is to say there is one aspect instance per classloader that defines the type. This means that if you define multiple application contexts within the same classloader hierarchy you need to consider where to define the @EnableSpringConfigured bean and where to place spring-aspects.jaron the classpath.

Consider a typical Spring web-app configuration with a shared parent application context defining common business services and everything needed to support them, and one child application context per servlet containing definitions particular to that servlet. All of these contexts will co-exist within the same classloader hierarchy, and so the AnnotationBeanConfigurerAspect can only hold a reference to one of them. In this case we recommend defining the @EnableSpringConfigured bean in the shared (parent) application context: this defines the services that you are likely to want to inject into domain objects. A consequence is that you cannot configure domain objects with references to beans defined in the child (servlet-specific) contexts using the @Configurable mechanism (probably not something you want to do anyway!).

When deploying multiple web-apps within the same container, ensure that each web-application loads the types in spring-aspects.jar using its own classloader (for example, by placing spring-aspects.jar in 'WEB-INF/lib'). If spring-aspects.jar is only added to the container wide classpath (and hence loaded by the shared parent classloader), all web applications will share the same aspect instance which is probably not what you want.

11.8.2 Other Spring aspects for AspectJ

In addition to the @Configurable aspect, spring-aspects.jar contains an AspectJ aspect that can be used to drive Spring’s transaction management for types and methods annotated with the @Transactional annotation. This is primarily intended for users who want to use the Spring Framework’s transaction support outside of the Spring container.

The aspect that interprets @Transactional annotations is the AnnotationTransactionAspect. When using this aspect, you must annotate the implementation class (and/or methods within that class), not the interface (if any) that the class implements. AspectJ follows Java’s rule that annotations on interfaces are not inherited.

A @Transactional annotation on a class specifies the default transaction semantics for the execution of any public operation in the class.

A @Transactional annotation on a method within the class overrides the default transaction semantics given by the class annotation (if present). Methods of any visibility may be annotated, including private methods. Annotating non-public methods directly is the only way to get transaction demarcation for the execution of such methods.

Spring框架参考文档 Spring Framework Reference Documentation

Since Spring Framework 4.2, spring-aspects provides a similar aspect that offers the exact same features for the standardjavax.transaction.Transactional annotation. Check JtaAnnotationTransactionAspect for more details.

For AspectJ programmers that want to use the Spring configuration and transaction management support but don’t want to (or cannot) use annotations,spring-aspects.jar also contains abstract aspects you can extend to provide your own pointcut definitions. See the sources for theAbstractBeanConfigurerAspect and AbstractTransactionAspect aspects for more information. As an example, the following excerpt shows how you could write an aspect to configure all instances of objects defined in the domain model using prototype bean definitions that match the fully-qualified class names:

public aspect DomainObjectConfiguration extends AbstractBeanConfigurerAspect {

    public DomainObjectConfiguration() {
        setBeanWiringInfoResolver(new ClassNameBeanWiringInfoResolver());
    }

    // the creation of a new bean (any object in the domain model)
    protected pointcut beanCreation(Object beanInstance) :
        initialization(new(..)) &&
        SystemArchitecture.inDomainModel() &&
        this(beanInstance);

}

11.8.3 Configuring AspectJ aspects using Spring IoC

When using AspectJ aspects with Spring applications, it is natural to both want and expect to be able to configure such aspects using Spring. The AspectJ runtime itself is responsible for aspect creation, and the means of configuring the AspectJ created aspects via Spring depends on the AspectJ instantiation model (the per-xxxclause) used by the aspect.

The majority of AspectJ aspects are singleton aspects. Configuration of these aspects is very easy: simply create a bean definition referencing the aspect type as normal, and include the bean attribute 'factory-method="aspectOf"'. This ensures that Spring obtains the aspect instance by asking AspectJ for it rather than trying to create an instance itself. For example:

<bean id="profiler" class="com.xyz.profiler.Profiler"
        factory-method="aspectOf">

    <property name="profilingStrategy" ref="jamonProfilingStrategy"/>
</bean>

Non-singleton aspects are harder to configure: however it is possible to do so by creating prototype bean definitions and using the @Configurable support fromspring-aspects.jar to configure the aspect instances once they have bean created by the AspectJ runtime.

If you have some @AspectJ aspects that you want to weave with AspectJ (for example, using load-time weaving for domain model types) and other @AspectJ aspects that you want to use with Spring AOP, and these aspects are all configured using Spring, then you will need to tell the Spring AOP @AspectJ autoproxying support which exact subset of the @AspectJ aspects defined in the configuration should be used for autoproxying. You can do this by using one or more <include/> elements inside the <aop:aspectj-autoproxy/> declaration. Each <include/> element specifies a name pattern, and only beans with names matched by at least one of the patterns will be used for Spring AOP autoproxy configuration:

<aop:aspectj-autoproxy>
    <aop:include name="thisBean"/>
    <aop:include name="thatBean"/>
</aop:aspectj-autoproxy>
Spring框架参考文档 Spring Framework Reference Documentation

Do not be misled by the name of the <aop:aspectj-autoproxy/> element: using it will result in the creation of Spring AOP proxies. The @AspectJ style of aspect declaration is just being used here, but the AspectJ runtime is not involved.

11.8.4 Load-time weaving with AspectJ in the Spring Framework

Load-time weaving (LTW) refers to the process of weaving AspectJ aspects into an application’s class files as they are being loaded into the Java virtual machine (JVM). The focus of this section is on configuring and using LTW in the specific context of the Spring Framework: this section is not an introduction to LTW though. For full details on the specifics of LTW and configuring LTW with just AspectJ (with Spring not being involved at all), see the LTW section of the AspectJ Development Environment Guide.

The value-add that the Spring Framework brings to AspectJ LTW is in enabling much finer-grained control over the weaving process. 'Vanilla' AspectJ LTW is effected using a Java (5+) agent, which is switched on by specifying a VM argument when starting up a JVM. It is thus a JVM-wide setting, which may be fine in some situations, but often is a little too coarse. Spring-enabled LTW enables you to switch on LTW on a per-ClassLoader basis, which obviously is more fine-grained and which can make more sense in a 'single-JVM-multiple-application' environment (such as is found in a typical application server environment).

Further, in certain environments, this support enables load-time weaving without making any modifications to the application server’s launch script that will be needed to add -javaagent:path/to/aspectjweaver.jar or (as we describe later in this section) -javaagent:path/to/org.springframework.instrument-{version}.jar(previously named spring-agent.jar). Developers simply modify one or more files that form the application context to enable load-time weaving instead of relying on administrators who typically are in charge of the deployment configuration such as the launch script.

Now that the sales pitch is over, let us first walk through a quick example of AspectJ LTW using Spring, followed by detailed specifics about elements introduced in the following example. For a complete example, please see the Petclinic sample application.

A first example

Let us assume that you are an application developer who has been tasked with diagnosing the cause of some performance problems in a system. Rather than break out a profiling tool, what we are going to do is switch on a simple profiling aspect that will enable us to very quickly get some performance metrics, so that we can then apply a finer-grained profiling tool to that specific area immediately afterwards.

Spring框架参考文档 Spring Framework Reference Documentation

The example presented here uses XML style configuration, it is also possible to configure and use @AspectJ with Java Configuration. Specifically the@EnableLoadTimeWeaving annotation can be used as an alternative to <context:load-time-weaver/> (see below for details).

Here is the profiling aspect. Nothing too fancy, just a quick-and-dirty time-based profiler, using the @AspectJ-style of aspect declaration.

package foo;

import org.aspectj.lang.ProceedingJoinPoint;
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.annotation.Pointcut;
import org.springframework.util.StopWatch;
import org.springframework.core.annotation.Order;

@Aspect
public class ProfilingAspect {

    @Around("methodsToBeProfiled()")
    public Object profile(ProceedingJoinPoint pjp) throws Throwable {
        StopWatch sw = new StopWatch(getClass().getSimpleName());
        try {
            sw.start(pjp.getSignature().getName());
            return pjp.proceed();
        } finally {
            sw.stop();
            System.out.println(sw.prettyPrint());
        }
    }

    @Pointcut("execution(public * foo..*.*(..))")
    public void methodsToBeProfiled(){}
}

We will also need to create an META-INF/aop.xml file, to inform the AspectJ weaver that we want to weave our ProfilingAspect into our classes. This file convention, namely the presence of a file (or files) on the Java classpath called META-INF/aop.xml is standard AspectJ.

<!DOCTYPE aspectj PUBLIC "-//AspectJ//DTD//EN" "http://www.eclipse.org/aspectj/dtd/aspectj.dtd">
<aspectj>

    <weaver>
        <!-- only weave classes in our application-specific packages -->
        <include within="foo.*"/>
    </weaver>

    <aspects>
        <!-- weave in just this aspect -->
        <aspect name="foo.ProfilingAspect"/>
    </aspects>

</aspectj>

Now to the Spring-specific portion of the configuration. We need to configure a LoadTimeWeaver (all explained later, just take it on trust for now). This load-time weaver is the essential component responsible for weaving the aspect configuration in one or more META-INF/aop.xml files into the classes in your application. The good thing is that it does not require a lot of configuration, as can be seen below (there are some more options that you can specify, but these are detailed later).

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="
        http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <!-- a service object; we will be profiling its methods -->
    <bean id="entitlementCalculationService"
            class="foo.StubEntitlementCalculationService"/>

    <!-- this switches on the load-time weaving -->
    <context:load-time-weaver/>
</beans>

Now that all the required artifacts are in place - the aspect, the META-INF/aop.xml file, and the Spring configuration -, let us create a simple driver class with amain(..) method to demonstrate the LTW in action.

package foo;

import org.springframework.context.support.ClassPathXmlApplicationContext;

public final class Main {

    public static void main(String[] args) {

        ApplicationContext ctx = new ClassPathXmlApplicationContext("beans.xml", Main.class);

        EntitlementCalculationService entitlementCalculationService
            = (EntitlementCalculationService) ctx.getBean("entitlementCalculationService");

        // the profiling aspect is 'woven' around this method execution
        entitlementCalculationService.calculateEntitlement();
    }
}

There is one last thing to do. The introduction to this section did say that one could switch on LTW selectively on a per- ClassLoader basis with Spring, and this is true. However, just for this example, we are going to use a Java agent (supplied with Spring) to switch on the LTW. This is the command line we will use to run the aboveMain class:

java -javaagent:C:/projects/foo/lib/global/spring-instrument.jar foo.Main

The -javaagent is a flag for specifying and enabling agents to instrument programs running on the JVM. The Spring Framework ships with such an agent, theInstrumentationSavingAgent, which is packaged in the spring-instrument.jar that was supplied as the value of the -javaagent argument in the above example.

The output from the execution of the Main program will look something like that below. (I have introduced a Thread.sleep(..) statement into thecalculateEntitlement() implementation so that the profiler actually captures something other than 0 milliseconds - the 01234 milliseconds is not an overhead introduced by the AOP :) )

Calculating entitlement

StopWatch 'ProfilingAspect': running time (millis) = 1234
------ ----- ----------------------------
ms     %     Task name
------ ----- ----------------------------
01234  100%  calculateEntitlement

Since this LTW is effected using full-blown AspectJ, we are not just limited to advising Spring beans; the following slight variation on the Main program will yield the same result.

package foo;

import org.springframework.context.support.ClassPathXmlApplicationContext;

public final class Main {

    public static void main(String[] args) {

        new ClassPathXmlApplicationContext("beans.xml", Main.class);

        EntitlementCalculationService entitlementCalculationService =
            new StubEntitlementCalculationService();

        // the profiling aspect will be 'woven' around this method execution
        entitlementCalculationService.calculateEntitlement();
    }
}

Notice how in the above program we are simply bootstrapping the Spring container, and then creating a new instance of the StubEntitlementCalculationServicetotally outside the context of Spring…​ the profiling advice still gets woven in.

The example admittedly is simplistic…​ however the basics of the LTW support in Spring have all been introduced in the above example, and the rest of this section will explain the 'why' behind each bit of configuration and usage in detail.

Spring框架参考文档 Spring Framework Reference Documentation

The ProfilingAspect used in this example may be basic, but it is quite useful. It is a nice example of a development-time aspect that developers can use during development (of course), and then quite easily exclude from builds of the application being deployed into UAT or production.

Aspects

The aspects that you use in LTW have to be AspectJ aspects. They can be written in either the AspectJ language itself or you can write your aspects in the @AspectJ-style. It means that your aspects are then both valid AspectJ and Spring AOP aspects. Furthermore, the compiled aspect classes need to be available on the classpath.

'META-INF/aop.xml'

The AspectJ LTW infrastructure is configured using one or more META-INF/aop.xml files, that are on the Java classpath (either directly, or more typically in jar files).

The structure and contents of this file is detailed in the main AspectJ reference documentation, and the interested reader is referred to that resource. (I appreciate that this section is brief, but the aop.xml file is 100% AspectJ - there is no Spring-specific information or semantics that apply to it, and so there is no extra value that I can contribute either as a result), so rather than rehash the quite satisfactory section that the AspectJ developers wrote, I am just directing you there.)

Required libraries (JARS)

At a minimum you will need the following libraries to use the Spring Framework’s support for AspectJ LTW:

  • spring-aop.jar (version 2.5 or later, plus all mandatory dependencies)
  • aspectjweaver.jar (version 1.6.8 or later)

If you are using the Spring-provided agent to enable instrumentation, you will also need:

  • spring-instrument.jar

Spring configuration

The key component in Spring’s LTW support is the LoadTimeWeaver interface (in the org.springframework.instrument.classloading package), and the numerous implementations of it that ship with the Spring distribution. A LoadTimeWeaver is responsible for adding one or morejava.lang.instrument.ClassFileTransformers to a ClassLoader at runtime, which opens the door to all manner of interesting applications, one of which happens to be the LTW of aspects.

Spring框架参考文档 Spring Framework Reference Documentation

If you are unfamiliar with the idea of runtime class file transformation, you are encouraged to read the javadoc API documentation for thejava.lang.instrument package before continuing. This is not a huge chore because there is - rather annoyingly - precious little documentation there…​ the key interfaces and classes will at least be laid out in front of you for reference as you read through this section.

Configuring a LoadTimeWeaver for a particular ApplicationContext can be as easy as adding one line. (Please note that you almost certainly will need to be using an ApplicationContext as your Spring container - typically a BeanFactory will not be enough because the LTW support makes use ofBeanFactoryPostProcessors.)

To enable the Spring Framework’s LTW support, you need to configure a LoadTimeWeaver, which typically is done using the @EnableLoadTimeWeaving annotation.

@Configuration
@EnableLoadTimeWeaving
public class AppConfig {

}

Alternatively, if you prefer XML based configuration, use the <context:load-time-weaver/> element. Note that the element is defined in the context namespace.

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="
        http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:load-time-weaver/>

</beans>

The above configuration will define and register a number of LTW-specific infrastructure beans for you automatically, such as a LoadTimeWeaver and anAspectJWeavingEnabler. The default LoadTimeWeaver is the DefaultContextLoadTimeWeaver class, which attempts to decorate an automatically detectedLoadTimeWeaver: the exact type of LoadTimeWeaver that will be 'automatically detected' is dependent upon your runtime environment (summarized in the following table).

Table 11.1. DefaultContextLoadTimeWeaver LoadTimeWeavers

Runtime Environment LoadTimeWeaverimplementation

Running in Oracle’s WebLogic

WebLogicLoadTimeWeaver

Running in Oracle’s GlassFish

GlassFishLoadTimeWeaver

Running in Apache Tomcat

TomcatLoadTimeWeaver

Running in Red Hat’s JBoss AS or WildFly

JBossLoadTimeWeaver

Running in IBM’s WebSphere

WebSphereLoadTimeWeaver

JVM started with Spring InstrumentationSavingAgent (java -javaagent:path/to/spring-instrument.jar)

InstrumentationLoadTimeWeaver

Fallback, expecting the underlying ClassLoader to follow common conventions (e.g. applicable toTomcatInstrumentableClassLoader and Resin)

ReflectiveLoadTimeWeaver

Note that these are just the LoadTimeWeavers that are autodetected when using the DefaultContextLoadTimeWeaver: it is of course possible to specify exactly which LoadTimeWeaver implementation that you wish to use.

To specify a specific LoadTimeWeaver with Java configuration implement the LoadTimeWeavingConfigurer interface and override the getLoadTimeWeaver()method:

@Configuration
@EnableLoadTimeWeaving
public class AppConfig implements LoadTimeWeavingConfigurer {

    @Override
    public LoadTimeWeaver getLoadTimeWeaver() {
        return new ReflectiveLoadTimeWeaver();
    }
}

If you are using XML based configuration you can specify the fully-qualified classname as the value of the weaver-class attribute on the<context:load-time-weaver/> element:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:context="http://www.springframework.org/schema/context"
    xsi:schemaLocation="
        http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd
        http://www.springframework.org/schema/context
        http://www.springframework.org/schema/context/spring-context.xsd">

    <context:load-time-weaver
            weaver-class="org.springframework.instrument.classloading.ReflectiveLoadTimeWeaver"/>

</beans>

The LoadTimeWeaver that is defined and registered by the configuration can be later retrieved from the Spring container using the well-known name loadTimeWeaver. Remember that the LoadTimeWeaver exists just as a mechanism for Spring’s LTW infrastructure to add one or more ClassFileTransformers. The actualClassFileTransformer that does the LTW is the ClassPreProcessorAgentAdapter (from the org.aspectj.weaver.loadtime package) class. See the class-level javadocs of the ClassPreProcessorAgentAdapter class for further details, because the specifics of how the weaving is actually effected is beyond the scope of this section.

There is one final attribute of the configuration left to discuss: the aspectjWeaving attribute (or aspectj-weaving if you are using XML). This is a simple attribute that controls whether LTW is enabled or not; it is as simple as that. It accepts one of three possible values, summarized below, with the default value being autodetect if the attribute is not present.

Table 11.2. AspectJ weaving attribute values

Annotation Value XML Value Explanation

ENABLED

on

AspectJ weaving is on, and aspects will be woven at load-time as appropriate.

DISABLED

off

LTW is off…​ no aspect will be woven at load-time.

AUTODETECT

autodetect

If the Spring LTW infrastructure can find at least one META-INF/aop.xml file, then AspectJ weaving is on, else it is off. This is the default value.

Environment-specific configuration

This last section contains any additional settings and configuration that you will need when using Spring’s LTW support in environments such as application servers and web containers.

Tomcat

Historically, Apache Tomcat's default class loader did not support class transformation which is why Spring provides an enhanced implementation that addresses this need. Named TomcatInstrumentableClassLoader, the loader works on Tomcat 6.0 and above.

Spring框架参考文档 Spring Framework Reference Documentation

Do not define TomcatInstrumentableClassLoader anymore on Tomcat 8.0 and higher. Instead, let Spring automatically use Tomcat’s new nativeInstrumentableClassLoader facility through the TomcatLoadTimeWeaver strategy.

If you still need to use TomcatInstrumentableClassLoader, it can be registered individually for each web application as follows:

  • Copy org.springframework.instrument.tomcat.jar into $CATALINA_HOME/lib, where $CATALINA_HOME represents the root of the Tomcat installation)
  • Instruct Tomcat to use the custom class loader (instead of the default) by editing the web application context file:
<Context path="/myWebApp" docBase="/my/webApp/location">
    <Loader
        loaderClass="org.springframework.instrument.classloading.tomcat.TomcatInstrumentableClassLoader"/>
</Context>

Apache Tomcat (6.0+) supports several context locations:

  • server configuration file - $CATALINA_HOME/conf/server.xml
  • default context configuration - $CATALINA_HOME/conf/context.xml - that affects all deployed web applications
  • per-web application configuration which can be deployed either on the server-side at $CATALINA_HOME/conf/[enginename]/[hostname]/[webapp]-context.xml or embedded inside the web-app archive at META-INF/context.xml

For efficiency, the embedded per-web-app configuration style is recommended because it will impact only applications that use the custom class loader and does not require any changes to the server configuration. See the Tomcat 6.0.x documentation for more details about available context locations.

Alternatively, consider the use of the Spring-provided generic VM agent, to be specified in Tomcat’s launch script (see above). This will make instrumentation available to all deployed web applications, no matter what ClassLoader they happen to run on.

WebLogic, WebSphere, Resin, GlassFish, JBoss

Recent versions of WebLogic Server (version 10 and above), IBM WebSphere Application Server (version 7 and above), Resin (3.1 and above) and JBoss (6.x or above) provide a ClassLoader that is capable of local instrumentation. Spring’s native LTW leverages such ClassLoaders to enable AspectJ weaving. You can enable LTW by simply activating load-time weaving as described earlier. Specifically, you do not need to modify the launch script to add-javaagent:path/to/spring-instrument.jar.

Note that GlassFish instrumentation-capable ClassLoader is available only in its EAR environment. For GlassFish web applications, follow the Tomcat setup instructions as outlined above.

Note that on JBoss 6.x, the app server scanning needs to be disabled to prevent it from loading the classes before the application actually starts. A quick workaround is to add to your artifact a file named WEB-INF/jboss-scanning.xml with the following content:

<scanning xmlns="urn:jboss:scanning:1.0"/>
Generic Java applications

When class instrumentation is required in environments that do not support or are not supported by the existing LoadTimeWeaver implementations, a JDK agent can be the only solution. For such cases, Spring provides InstrumentationLoadTimeWeaver, which requires a Spring-specific (but very general) VM agent,org.springframework.instrument-{version}.jar (previously named spring-agent.jar).

To use it, you must start the virtual machine with the Spring agent, by supplying the following JVM options:

-javaagent:/path/to/org.springframework.instrument-{version}.jar

Note that this requires modification of the VM launch script which may prevent you from using this in application server environments (depending on your operation policies). Additionally, the JDK agent will instrument the entire VM which can prove expensive.

For performance reasons, it is recommended to use this configuration only if your target environment (such as Jetty) does not have (or does not support) a dedicated LTW.

11.9 Further Resources

More information on AspectJ can be found on the AspectJ website.

The book Eclipse AspectJ by Adrian Colyer et. al. (Addison-Wesley, 2005) provides a comprehensive introduction and reference for the AspectJ language.

The book AspectJ in Action, Second Edition by Ramnivas Laddad (Manning, 2009) comes highly recommended; the focus of the book is on AspectJ, but a lot of general AOP themes are explored (in some depth).

12. Spring AOP APIs

12.1 Introduction

The previous chapter described the Spring’s support for AOP using @AspectJ and schema-based aspect definitions. In this chapter we discuss the lower-level Spring AOP APIs and the AOP support used in Spring 1.2 applications. For new applications, we recommend the use of the Spring 2.0 and later AOP support described in the previous chapter, but when working with existing applications, or when reading books and articles, you may come across Spring 1.2 style examples. Spring 4.0 is backwards compatible with Spring 1.2 and everything described in this chapter is fully supported in Spring 4.0.

12.2 Pointcut API in Spring

Let’s look at how Spring handles the crucial pointcut concept.

12.2.1 Concepts

Spring’s pointcut model enables pointcut reuse independent of advice types. It’s possible to target different advice using the same pointcut.

The org.springframework.aop.Pointcut interface is the central interface, used to target advices to particular classes and methods. The complete interface is shown below:

public interface Pointcut {

    ClassFilter getClassFilter();

    MethodMatcher getMethodMatcher();

}

Splitting the Pointcut interface into two parts allows reuse of class and method matching parts, and fine-grained composition operations (such as performing a "union" with another method matcher).

The ClassFilter interface is used to restrict the pointcut to a given set of target classes. If the matches() method always returns true, all target classes will be matched:

public interface ClassFilter {

    boolean matches(Class clazz);
}

The MethodMatcher interface is normally more important. The complete interface is shown below:

public interface MethodMatcher {

    boolean matches(Method m, Class targetClass);

    boolean isRuntime();

    boolean matches(Method m, Class targetClass, Object[] args);
}

The matches(Method, Class) method is used to test whether this pointcut will ever match a given method on a target class. This evaluation can be performed when an AOP proxy is created, to avoid the need for a test on every method invocation. If the 2-argument matches method returns true for a given method, and theisRuntime() method for the MethodMatcher returns true, the 3-argument matches method will be invoked on every method invocation. This enables a pointcut to look at the arguments passed to the method invocation immediately before the target advice is to execute.

Most MethodMatchers are static, meaning that their isRuntime() method returns false. In this case, the 3-argument matches method will never be invoked.

Spring框架参考文档 Spring Framework Reference Documentation

If possible, try to make pointcuts static, allowing the AOP framework to cache the results of pointcut evaluation when an AOP proxy is created.

12.2.2 Operations on pointcuts

Spring supports operations on pointcuts: notably, union and intersection.

  • Union means the methods that either pointcut matches.
  • Intersection means the methods that both pointcuts match.
  • Union is usually more useful.
  • Pointcuts can be composed using the static methods in the org.springframework.aop.support.Pointcuts class, or using the ComposablePointcut class in the same package. However, using AspectJ pointcut expressions is usually a simpler approach.

12.2.3 AspectJ expression pointcuts

Since 2.0, the most important type of pointcut used by Spring is org.springframework.aop.aspectj.AspectJExpressionPointcut. This is a pointcut that uses an AspectJ supplied library to parse an AspectJ pointcut expression string.

See the previous chapter for a discussion of supported AspectJ pointcut primitives.

12.2.4 Convenience pointcut implementations

Spring provides several convenient pointcut implementations. Some can be used out of the box; others are intended to be subclassed in application-specific pointcuts.

Static pointcuts

Static pointcuts are based on method and target class, and cannot take into account the method’s arguments. Static pointcuts are sufficient - and best - for most usages. It’s possible for Spring to evaluate a static pointcut only once, when a method is first invoked: after that, there is no need to evaluate the pointcut again with each method invocation.

Let’s consider some static pointcut implementations included with Spring.

Regular expression pointcuts

One obvious way to specify static pointcuts is regular expressions. Several AOP frameworks besides Spring make this possible.org.springframework.aop.support.JdkRegexpMethodPointcut is a generic regular expression pointcut, using the regular expression support in JDK 1.4+.

Using the JdkRegexpMethodPointcut class, you can provide a list of pattern Strings. If any of these is a match, the pointcut will evaluate to true. (So the result is effectively the union of these pointcuts.)

The usage is shown below:

<bean id="settersAndAbsquatulatePointcut"
        class="org.springframework.aop.support.JdkRegexpMethodPointcut">
    <property name="patterns">
        <list>
            <value>.*set.*</value>
            <value>.*absquatulate</value>
        </list>
    </property>
</bean>

Spring provides a convenience class, RegexpMethodPointcutAdvisor, that allows us to also reference an Advice (remember that an Advice can be an interceptor, before advice, throws advice etc.). Behind the scenes, Spring will use a JdkRegexpMethodPointcut. Using RegexpMethodPointcutAdvisor simplifies wiring, as the one bean encapsulates both pointcut and advice, as shown below:

<bean id="settersAndAbsquatulateAdvisor"
        class="org.springframework.aop.support.RegexpMethodPointcutAdvisor">
    <property name="advice">
        <ref bean="beanNameOfAopAllianceInterceptor"/>
    </property>
    <property name="patterns">
        <list>
            <value>.*set.*</value>
            <value>.*absquatulate</value>
        </list>
    </property>
</bean>

RegexpMethodPointcutAdvisor can be used with any Advice type.

Attribute-driven pointcuts

An important type of static pointcut is a metadata-driven pointcut. This uses the values of metadata attributes: typically, source-level metadata.

Dynamic pointcuts

Dynamic pointcuts are costlier to evaluate than static pointcuts. They take into account method arguments, as well as static information. This means that they must be evaluated with every method invocation; the result cannot be cached, as arguments will vary.

The main example is the control flow pointcut.

Control flow pointcuts

Spring control flow pointcuts are conceptually similar to AspectJ cflow pointcuts, although less powerful. (There is currently no way to specify that a pointcut executes below a join point matched by another pointcut.) A control flow pointcut matches the current call stack. For example, it might fire if the join point was invoked by a method in the com.mycompany.web package, or by the SomeCaller class. Control flow pointcuts are specified using theorg.springframework.aop.support.ControlFlowPointcut class.

Spring框架参考文档 Spring Framework Reference Documentation

Control flow pointcuts are significantly more expensive to evaluate at runtime than even other dynamic pointcuts. In Java 1.4, the cost is about 5 times that of other dynamic pointcuts.

12.2.5 Pointcut superclasses

Spring provides useful pointcut superclasses to help you to implement your own pointcuts.

Because static pointcuts are most useful, you’ll probably subclass StaticMethodMatcherPointcut, as shown below. This requires implementing just one abstract method (although it’s possible to override other methods to customize behavior):

class TestStaticPointcut extends StaticMethodMatcherPointcut {

    public boolean matches(Method m, Class targetClass) {
        // return true if custom criteria match
    }
}

There are also superclasses for dynamic pointcuts.

You can use custom pointcuts with any advice type in Spring 1.0 RC2 and above.

12.2.6 Custom pointcuts

Because pointcuts in Spring AOP are Java classes, rather than language features (as in AspectJ) it’s possible to declare custom pointcuts, whether static or dynamic. Custom pointcuts in Spring can be arbitrarily complex. However, using the AspectJ pointcut expression language is recommended if possible.

Spring框架参考文档 Spring Framework Reference Documentation

Later versions of Spring may offer support for "semantic pointcuts" as offered by JAC: for example, "all methods that change instance variables in the target object."

12.3 Advice API in Spring

Let’s now look at how Spring AOP handles advice.

12.3.1 Advice lifecycles

Each advice is a Spring bean. An advice instance can be shared across all advised objects, or unique to each advised object. This corresponds to per-class or per-instance advice.

Per-class advice is used most often. It is appropriate for generic advice such as transaction advisors. These do not depend on the state of the proxied object or add new state; they merely act on the method and arguments.

Per-instance advice is appropriate for introductions, to support mixins. In this case, the advice adds state to the proxied object.

It’s possible to use a mix of shared and per-instance advice in the same AOP proxy.

12.3.2 Advice types in Spring

Spring provides several advice types out of the box, and is extensible to support arbitrary advice types. Let us look at the basic concepts and standard advice types.

Interception around advice

The most fundamental advice type in Spring is interception around advice.

Spring is compliant with the AOP Alliance interface for around advice using method interception. MethodInterceptors implementing around advice should implement the following interface:

public interface MethodInterceptor extends Interceptor {

    Object invoke(MethodInvocation invocation) throws Throwable;
}

The MethodInvocation argument to the invoke() method exposes the method being invoked; the target join point; the AOP proxy; and the arguments to the method. The invoke() method should return the invocation’s result: the return value of the join point.

A simple MethodInterceptor implementation looks as follows:

public class DebugInterceptor implements MethodInterceptor {

    public Object invoke(MethodInvocation invocation) throws Throwable {
        System.out.println("Before: invocation=[" + invocation + "]");
        Object rval = invocation.proceed();
        System.out.println("Invocation returned");
        return rval;
    }
}

Note the call to the MethodInvocation’s proceed() method. This proceeds down the interceptor chain towards the join point. Most interceptors will invoke this method, and return its return value. However, a MethodInterceptor, like any around advice, can return a different value or throw an exception rather than invoke the proceed method. However, you don’t want to do this without good reason!

Spring框架参考文档 Spring Framework Reference Documentation

MethodInterceptors offer interoperability with other AOP Alliance-compliant AOP implementations. The other advice types discussed in the remainder of this section implement common AOP concepts, but in a Spring-specific way. While there is an advantage in using the most specific advice type, stick with MethodInterceptor around advice if you are likely to want to run the aspect in another AOP framework. Note that pointcuts are not currently interoperable between frameworks, and the AOP Alliance does not currently define pointcut interfaces.

Before advice

A simpler advice type is a before advice. This does not need a MethodInvocation object, since it will only be called before entering the method.

The main advantage of a before advice is that there is no need to invoke the proceed() method, and therefore no possibility of inadvertently failing to proceed down the interceptor chain.

The MethodBeforeAdvice interface is shown below. (Spring’s API design would allow for field before advice, although the usual objects apply to field interception and it’s unlikely that Spring will ever implement it).

public interface MethodBeforeAdvice extends BeforeAdvice {

    void before(Method m, Object[] args, Object target) throws Throwable;
}

Note the return type is void. Before advice can insert custom behavior before the join point executes, but cannot change the return value. If a before advice throws an exception, this will abort further execution of the interceptor chain. The exception will propagate back up the interceptor chain. If it is unchecked, or on the signature of the invoked method, it will be passed directly to the client; otherwise it will be wrapped in an unchecked exception by the AOP proxy.

An example of a before advice in Spring, which counts all method invocations:

public class CountingBeforeAdvice implements MethodBeforeAdvice {

    private int count;

    public void before(Method m, Object[] args, Object target) throws Throwable {
        ++count;
    }

    public int getCount() {
        return count;
    }
}
Spring框架参考文档 Spring Framework Reference Documentation

Before advice can be used with any pointcut.

Throws advice

Throws advice is invoked after the return of the join point if the join point threw an exception. Spring offers typed throws advice. Note that this means that theorg.springframework.aop.ThrowsAdvice interface does not contain any methods: It is a tag interface identifying that the given object implements one or more typed throws advice methods. These should be in the form of:

afterThrowing([Method, args, target], subclassOfThrowable)

Only the last argument is required. The method signatures may have either one or four arguments, depending on whether the advice method is interested in the method and arguments. The following classes are examples of throws advice.

The advice below is invoked if a RemoteException is thrown (including subclasses):

public class RemoteThrowsAdvice implements ThrowsAdvice {

    public void afterThrowing(RemoteException ex) throws Throwable {
        // Do something with remote exception
    }
}

The following advice is invoked if a ServletException is thrown. Unlike the above advice, it declares 4 arguments, so that it has access to the invoked method, method arguments and target object:

public class ServletThrowsAdviceWithArguments implements ThrowsAdvice {

    public void afterThrowing(Method m, Object[] args, Object target, ServletException ex) {
        // Do something with all arguments
    }
}

The final example illustrates how these two methods could be used in a single class, which handles both RemoteException and ServletException. Any number of throws advice methods can be combined in a single class.

public static class CombinedThrowsAdvice implements ThrowsAdvice {

    public void afterThrowing(RemoteException ex) throws Throwable {
        // Do something with remote exception
    }

    public void afterThrowing(Method m, Object[] args, Object target, ServletException ex) {
        // Do something with all arguments
    }
}
Spring框架参考文档 Spring Framework Reference Documentation

If a throws-advice method throws an exception itself, it will override the original exception (i.e. change the exception thrown to the user). The overriding exception will typically be a RuntimeException; this is compatible with any method signature. However, if a throws-advice method throws a checked exception, it will have to match the declared exceptions of the target method and is hence to some degree coupled to specific target method signatures. Do not throw an undeclared checked exception that is incompatible with the target method’s signature!
Spring框架参考文档 Spring Framework Reference Documentation

Throws advice can be used with any pointcut.

After Returning advice

An after returning advice in Spring must implement the org.springframework.aop.AfterReturningAdvice interface, shown below:

public interface AfterReturningAdvice extends Advice {

    void afterReturning(Object returnValue, Method m, Object[] args, Object target)
            throws Throwable;
}

An after returning advice has access to the return value (which it cannot modify), invoked method, methods arguments and target.

The following after returning advice counts all successful method invocations that have not thrown exceptions:

public class CountingAfterReturningAdvice implements AfterReturningAdvice {

    private int count;

    public void afterReturning(Object returnValue, Method m, Object[] args, Object target)
            throws Throwable {
        ++count;
    }

    public int getCount() {
        return count;
    }
}

This advice doesn’t change the execution path. If it throws an exception, this will be thrown up the interceptor chain instead of the return value.

Spring框架参考文档 Spring Framework Reference Documentation

After returning advice can be used with any pointcut.

Introduction advice

Spring treats introduction advice as a special kind of interception advice.

Introduction requires an IntroductionAdvisor, and an IntroductionInterceptor, implementing the following interface:

public interface IntroductionInterceptor extends MethodInterceptor {

    boolean implementsInterface(Class intf);
}

The invoke() method inherited from the AOP Alliance MethodInterceptor interface must implement the introduction: that is, if the invoked method is on an introduced interface, the introduction interceptor is responsible for handling the method call - it cannot invoke proceed().

Introduction advice cannot be used with any pointcut, as it applies only at class, rather than method, level. You can only use introduction advice with theIntroductionAdvisor, which has the following methods:

public interface IntroductionAdvisor extends Advisor, IntroductionInfo {

    ClassFilter getClassFilter();

    void validateInterfaces() throws IllegalArgumentException;
}

public interface IntroductionInfo {

    Class[] getInterfaces();
}

There is no MethodMatcher, and hence no Pointcut, associated with introduction advice. Only class filtering is logical.

The getInterfaces() method returns the interfaces introduced by this advisor.

The validateInterfaces() method is used internally to see whether or not the introduced interfaces can be implemented by the configuredIntroductionInterceptor.

Let’s look at a simple example from the Spring test suite. Let’s suppose we want to introduce the following interface to one or more objects:

public interface Lockable {
    void lock();
    void unlock();
    boolean locked();
}

This illustrates a mixin. We want to be able to cast advised objects to Lockable, whatever their type, and call lock and unlock methods. If we call the lock() method, we want all setter methods to throw a LockedException. Thus we can add an aspect that provides the ability to make objects immutable, without them having any knowledge of it: a good example of AOP.

Firstly, we’ll need an IntroductionInterceptor that does the heavy lifting. In this case, we extend theorg.springframework.aop.support.DelegatingIntroductionInterceptor convenience class. We could implement IntroductionInterceptor directly, but usingDelegatingIntroductionInterceptor is best for most cases.

The DelegatingIntroductionInterceptor is designed to delegate an introduction to an actual implementation of the introduced interface(s), concealing the use of interception to do so. The delegate can be set to any object using a constructor argument; the default delegate (when the no-arg constructor is used) is this. Thus in the example below, the delegate is the LockMixin subclass of DelegatingIntroductionInterceptor. Given a delegate (by default itself), aDelegatingIntroductionInterceptor instance looks for all interfaces implemented by the delegate (other than IntroductionInterceptor), and will support introductions against any of them. It’s possible for subclasses such as LockMixin to call the suppressInterface(Class intf) method to suppress interfaces that should not be exposed. However, no matter how many interfaces an IntroductionInterceptor is prepared to support, the IntroductionAdvisor used will control which interfaces are actually exposed. An introduced interface will conceal any implementation of the same interface by the target.

Thus LockMixin extends DelegatingIntroductionInterceptor and implements Lockable itself. The superclass automatically picks up that Lockable can be supported for introduction, so we don’t need to specify that. We could introduce any number of interfaces in this way.

Note the use of the locked instance variable. This effectively adds additional state to that held in the target object.

public class LockMixin extends DelegatingIntroductionInterceptor implements Lockable {

    private boolean locked;

    public void lock() {
        this.locked = true;
    }

    public void unlock() {
        this.locked = false;
    }

    public boolean locked() {
        return this.locked;
    }

    public Object invoke(MethodInvocation invocation) throws Throwable {
        if (locked() && invocation.getMethod().getName().indexOf("set") == 0) {
            throw new LockedException();
        }
        return super.invoke(invocation);
    }

}

Often it isn’t necessary to override the invoke() method: the DelegatingIntroductionInterceptor implementation - which calls the delegate method if the method is introduced, otherwise proceeds towards the join point - is usually sufficient. In the present case, we need to add a check: no setter method can be invoked if in locked mode.

The introduction advisor required is simple. All it needs to do is hold a distinct LockMixin instance, and specify the introduced interfaces - in this case, just Lockable. A more complex example might take a reference to the introduction interceptor (which would be defined as a prototype): in this case, there’s no configuration relevant for a LockMixin, so we simply create it using new.

public class LockMixinAdvisor extends DefaultIntroductionAdvisor {

    public LockMixinAdvisor() {
        super(new LockMixin(), Lockable.class);
    }
}

We can apply this advisor very simply: it requires no configuration. (However, it is necessary: It’s impossible to use an IntroductionInterceptor without anIntroductionAdvisor.) As usual with introductions, the advisor must be per-instance, as it is stateful. We need a different instance of LockMixinAdvisor, and henceLockMixin, for each advised object. The advisor comprises part of the advised object’s state.

We can apply this advisor programmatically, using the Advised.addAdvisor() method, or (the recommended way) in XML configuration, like any other advisor. All proxy creation choices discussed below, including "auto proxy creators," correctly handle introductions and stateful mixins.

12.4 Advisor API in Spring

In Spring, an Advisor is an aspect that contains just a single advice object associated with a pointcut expression.

Apart from the special case of introductions, any advisor can be used with any advice. org.springframework.aop.support.DefaultPointcutAdvisor is the most commonly used advisor class. For example, it can be used with a MethodInterceptor, BeforeAdvice or ThrowsAdvice.

It is possible to mix advisor and advice types in Spring in the same AOP proxy. For example, you could use a interception around advice, throws advice and before advice in one proxy configuration: Spring will automatically create the necessary interceptor chain.

12.5 Using the ProxyFactoryBean to create AOP proxies

If you’re using the Spring IoC container (an ApplicationContext or BeanFactory) for your business objects - and you should be! - you will want to use one of Spring’s AOP FactoryBeans. (Remember that a factory bean introduces a layer of indirection, enabling it to create objects of a different type.)

Spring框架参考文档 Spring Framework Reference Documentation

The Spring AOP support also uses factory beans under the covers.

The basic way to create an AOP proxy in Spring is to use the org.springframework.aop.framework.ProxyFactoryBean. This gives complete control over the pointcuts and advice that will apply, and their ordering. However, there are simpler options that are preferable if you don’t need such control.

12.5.1 Basics

The ProxyFactoryBean, like other Spring FactoryBean implementations, introduces a level of indirection. If you define a ProxyFactoryBean with name foo, what objects referencing foo see is not the ProxyFactoryBean instance itself, but an object created by theProxyFactoryBean’s implementation of the `getObject() method. This method will create an AOP proxy wrapping a target object.

One of the most important benefits of using a ProxyFactoryBean or another IoC-aware class to create AOP proxies, is that it means that advices and pointcuts can also be managed by IoC. This is a powerful feature, enabling certain approaches that are hard to achieve with other AOP frameworks. For example, an advice may itself reference application objects (besides the target, which should be available in any AOP framework), benefiting from all the pluggability provided by Dependency Injection.

12.5.2 JavaBean properties

In common with most FactoryBean implementations provided with Spring, the ProxyFactoryBean class is itself a JavaBean. Its properties are used to:

Some key properties are inherited from org.springframework.aop.framework.ProxyConfig (the superclass for all AOP proxy factories in Spring). These key properties include:

  • proxyTargetClass: true if the target class is to be proxied, rather than the target class' interfaces. If this property value is set to true, then CGLIB proxies will be created (but see also Section 12.5.3, “JDK- and CGLIB-based proxies”).
  • optimize: controls whether or not aggressive optimizations are applied to proxies created via CGLIB. One should not blithely use this setting unless one fully understands how the relevant AOP proxy handles optimization. This is currently used only for CGLIB proxies; it has no effect with JDK dynamic proxies.
  • frozen: if a proxy configuration is frozen, then changes to the configuration are no longer allowed. This is useful both as a slight optimization and for those cases when you don’t want callers to be able to manipulate the proxy (via the Advised interface) after the proxy has been created. The default value of this property isfalse, so changes such as adding additional advice are allowed.
  • exposeProxy: determines whether or not the current proxy should be exposed in a ThreadLocal so that it can be accessed by the target. If a target needs to obtain the proxy and the exposeProxy property is set to true, the target can use the AopContext.currentProxy() method.

Other properties specific to ProxyFactoryBean include:

  • proxyInterfaces: array of String interface names. If this isn’t supplied, a CGLIB proxy for the target class will be used (but see also Section 12.5.3, “JDK- and CGLIB-based proxies”).
  • interceptorNames: String array of Advisor, interceptor or other advice names to apply. Ordering is significant, on a first come-first served basis. That is to say that the first interceptor in the list will be the first to be able to intercept the invocation.

The names are bean names in the current factory, including bean names from ancestor factories. You can’t mention bean references here since doing so would result in the ProxyFactoryBean ignoring the singleton setting of the advice.

You can append an interceptor name with an asterisk ( *). This will result in the application of all advisor beans with names starting with the part before the asterisk to be applied. An example of using this feature can be found in Section 12.5.6, “Using 'global' advisors”.

  • singleton: whether or not the factory should return a single object, no matter how often the getObject() method is called. Several FactoryBean implementations offer such a method. The default value is true. If you want to use stateful advice - for example, for stateful mixins - use prototype advices along with a singleton value of false.

12.5.3 JDK- and CGLIB-based proxies

This section serves as the definitive documentation on how the ProxyFactoryBean chooses to create one of either a JDK- and CGLIB-based proxy for a particular target object (that is to be proxied).

Spring框架参考文档 Spring Framework Reference Documentation

The behavior of the ProxyFactoryBean with regard to creating JDK- or CGLIB-based proxies changed between versions 1.2.x and 2.0 of Spring. TheProxyFactoryBean now exhibits similar semantics with regard to auto-detecting interfaces as those of the TransactionProxyFactoryBean class.

If the class of a target object that is to be proxied (hereafter simply referred to as the target class) doesn’t implement any interfaces, then a CGLIB-based proxy will be created. This is the easiest scenario, because JDK proxies are interface based, and no interfaces means JDK proxying isn’t even possible. One simply plugs in the target bean, and specifies the list of interceptors via the interceptorNames property. Note that a CGLIB-based proxy will be created even if the proxyTargetClass property of the ProxyFactoryBean has been set to false. (Obviously this makes no sense, and is best removed from the bean definition because it is at best redundant, and at worst confusing.)

If the target class implements one (or more) interfaces, then the type of proxy that is created depends on the configuration of the ProxyFactoryBean.

If the proxyTargetClass property of the ProxyFactoryBean has been set to true, then a CGLIB-based proxy will be created. This makes sense, and is in keeping with the principle of least surprise. Even if the proxyInterfaces property of the ProxyFactoryBean has been set to one or more fully qualified interface names, the fact that the proxyTargetClass property is set to true will cause CGLIB-based proxying to be in effect.

If the proxyInterfaces property of the ProxyFactoryBean has been set to one or more fully qualified interface names, then a JDK-based proxy will be created. The created proxy will implement all of the interfaces that were specified in the proxyInterfaces property; if the target class happens to implement a whole lot more interfaces than those specified in the proxyInterfaces property, that is all well and good but those additional interfaces will not be implemented by the returned proxy.

If the proxyInterfaces property of the ProxyFactoryBean has not been set, but the target class does implement one (or more) interfaces, then theProxyFactoryBean will auto-detect the fact that the target class does actually implement at least one interface, and a JDK-based proxy will be created. The interfaces that are actually proxied will be all of the interfaces that the target class implements; in effect, this is the same as simply supplying a list of each and every interface that the target class implements to the proxyInterfaces property. However, it is significantly less work, and less prone to typos.

12.5.4 Proxying interfaces

Let’s look at a simple example of ProxyFactoryBean in action. This example involves:

  • A target bean that will be proxied. This is the "personTarget" bean definition in the example below.
  • An Advisor and an Interceptor used to provide advice.
  • An AOP proxy bean definition specifying the target object (the personTarget bean) and the interfaces to proxy, along with the advices to apply.
<bean id="personTarget" class="com.mycompany.PersonImpl">
    <property name="name" value="Tony"/>
    <property name="age" value="51"/>
</bean>

<bean id="myAdvisor" class="com.mycompany.MyAdvisor">
    <property name="someProperty" value="Custom string property value"/>
</bean>

<bean id="debugInterceptor" class="org.springframework.aop.interceptor.DebugInterceptor">
</bean>

<bean id="person"
    class="org.springframework.aop.framework.ProxyFactoryBean">
    <property name="proxyInterfaces" value="com.mycompany.Person"/>

    <property name="target" ref="personTarget"/>
    <property name="interceptorNames">
        <list>
            <value>myAdvisor</value>
            <value>debugInterceptor</value>
        </list>
    </property>
</bean>

Note that the interceptorNames property takes a list of String: the bean names of the interceptor or advisors in the current factory. Advisors, interceptors, before, after returning and throws advice objects can be used. The ordering of advisors is significant.

Spring框架参考文档 Spring Framework Reference Documentation

You might be wondering why the list doesn’t hold bean references. The reason for this is that if the ProxyFactoryBean’s singleton property is set to false, it must be able to return independent proxy instances. If any of the advisors is itself a prototype, an independent instance would need to be returned, so it’s necessary to be able to obtain an instance of the prototype from the factory; holding a reference isn’t sufficient.

The "person" bean definition above can be used in place of a Person implementation, as follows:

Person person = (Person) factory.getBean("person");

Other beans in the same IoC context can express a strongly typed dependency on it, as with an ordinary Java object:

<bean id="personUser" class="com.mycompany.PersonUser">
    <property name="person"><ref bean="person"/></property>
</bean>

The PersonUser class in this example would expose a property of type Person. As far as it’s concerned, the AOP proxy can be used transparently in place of a "real" person implementation. However, its class would be a dynamic proxy class. It would be possible to cast it to the Advised interface (discussed below).

It’s possible to conceal the distinction between target and proxy using an anonymous inner bean, as follows. Only the ProxyFactoryBean definition is different; the advice is included only for completeness:

<bean id="myAdvisor" class="com.mycompany.MyAdvisor">
    <property name="someProperty" value="Custom string property value"/>
</bean>

<bean id="debugInterceptor" class="org.springframework.aop.interceptor.DebugInterceptor"/>

<bean id="person" class="org.springframework.aop.framework.ProxyFactoryBean">
    <property name="proxyInterfaces" value="com.mycompany.Person"/>
    <!-- Use inner bean, not local reference to target -->
    <property name="target">
        <bean class="com.mycompany.PersonImpl">
            <property name="name" value="Tony"/>
            <property name="age" value="51"/>
        </bean>
    </property>
    <property name="interceptorNames">
        <list>
            <value>myAdvisor</value>
            <value>debugInterceptor</value>
        </list>
    </property>
</bean>

This has the advantage that there’s only one object of type Person: useful if we want to prevent users of the application context from obtaining a reference to the un-advised object, or need to avoid any ambiguity with Spring IoC autowiring. There’s also arguably an advantage in that the ProxyFactoryBean definition is self-contained. However, there are times when being able to obtain the un-advised target from the factory might actually be an advantage: for example, in certain test scenarios.

12.5.5 Proxying classes

What if you need to proxy a class, rather than one or more interfaces?

Imagine that in our example above, there was no Person interface: we needed to advise a class called Person that didn’t implement any business interface. In this case, you can configure Spring to use CGLIB proxying, rather than dynamic proxies. Simply set the proxyTargetClass property on the ProxyFactoryBean above to true. While it’s best to program to interfaces, rather than classes, the ability to advise classes that don’t implement interfaces can be useful when working with legacy code. (In general, Spring isn’t prescriptive. While it makes it easy to apply good practices, it avoids forcing a particular approach.)

If you want to, you can force the use of CGLIB in any case, even if you do have interfaces.

CGLIB proxying works by generating a subclass of the target class at runtime. Spring configures this generated subclass to delegate method calls to the original target: the subclass is used to implement the Decorator pattern, weaving in the advice.

CGLIB proxying should generally be transparent to users. However, there are some issues to consider:

  • Final methods can’t be advised, as they can’t be overridden.
  • There is no need to add CGLIB to your classpath. As of Spring 3.2, CGLIB is repackaged and included in the spring-core JAR. In other words, CGLIB-based AOP will work "out of the box" just as do JDK dynamic proxies.

There’s little performance difference between CGLIB proxying and dynamic proxies. As of Spring 1.0, dynamic proxies are slightly faster. However, this may change in the future. Performance should not be a decisive consideration in this case.

12.5.6 Using 'global' advisors

By appending an asterisk to an interceptor name, all advisors with bean names matching the part before the asterisk, will be added to the advisor chain. This can come in handy if you need to add a standard set of 'global' advisors:

<bean id="proxy" class="org.springframework.aop.framework.ProxyFactoryBean">
    <property name="target" ref="service"/>
    <property name="interceptorNames">
        <list>
            <value>global*</value>
        </list>
    </property>
</bean>

<bean id="global_debug" class="org.springframework.aop.interceptor.DebugInterceptor"/>
<bean id="global_performance" class="org.springframework.aop.interceptor.PerformanceMonitorInterceptor"/>

12.6 Concise proxy definitions

Especially when defining transactional proxies, you may end up with many similar proxy definitions. The use of parent and child bean definitions, along with inner bean definitions, can result in much cleaner and more concise proxy definitions.

First a parent, template, bean definition is created for the proxy:

<bean id="txProxyTemplate" abstract="true"
        class="org.springframework.transaction.interceptor.TransactionProxyFactoryBean">
    <property name="transactionManager" ref="transactionManager"/>
    <property name="transactionAttributes">
        <props>
            <prop key="*">PROPAGATION_REQUIRED</prop>
        </props>
    </property>
</bean>

This will never be instantiated itself, so may actually be incomplete. Then each proxy which needs to be created is just a child bean definition, which wraps the target of the proxy as an inner bean definition, since the target will never be used on its own anyway.

<bean id="myService" parent="txProxyTemplate">
    <property name="target">
        <bean class="org.springframework.samples.MyServiceImpl">
        </bean>
    </property>
</bean>

It is of course possible to override properties from the parent template, such as in this case, the transaction propagation settings:

<bean id="mySpecialService" parent="txProxyTemplate">
    <property name="target">
        <bean class="org.springframework.samples.MySpecialServiceImpl">
        </bean>
    </property>
    <property name="transactionAttributes">
        <props>
            <prop key="get*">PROPAGATION_REQUIRED,readOnly</prop>
            <prop key="find*">PROPAGATION_REQUIRED,readOnly</prop>
            <prop key="load*">PROPAGATION_REQUIRED,readOnly</prop>
            <prop key="store*">PROPAGATION_REQUIRED</prop>
        </props>
    </property>
</bean>

Note that in the example above, we have explicitly marked the parent bean definition as abstract by using the abstract attribute, as described previously, so that it may not actually ever be instantiated. Application contexts (but not simple bean factories) will by default pre-instantiate all singletons. It is therefore important (at least for singleton beans) that if you have a (parent) bean definition which you intend to use only as a template, and this definition specifies a class, you must make sure to set theabstract attribute to true, otherwise the application context will actually try to pre-instantiate it.

12.7 Creating AOP proxies programmatically with the ProxyFactory

It’s easy to create AOP proxies programmatically using Spring. This enables you to use Spring AOP without dependency on Spring IoC.

The following listing shows creation of a proxy for a target object, with one interceptor and one advisor. The interfaces implemented by the target object will automatically be proxied:

ProxyFactory factory = new ProxyFactory(myBusinessInterfaceImpl);
factory.addAdvice(myMethodInterceptor);
factory.addAdvisor(myAdvisor);
MyBusinessInterface tb = (MyBusinessInterface) factory.getProxy();

The first step is to construct an object of type org.springframework.aop.framework.ProxyFactory. You can create this with a target object, as in the above example, or specify the interfaces to be proxied in an alternate constructor.

You can add advices (with interceptors as a specialized kind of advice) and/or advisors, and manipulate them for the life of the ProxyFactory. If you add an IntroductionInterceptionAroundAdvisor, you can cause the proxy to implement additional interfaces.

There are also convenience methods on ProxyFactory (inherited from AdvisedSupport) which allow you to add other advice types such as before and throws advice. AdvisedSupport is the superclass of both ProxyFactory and ProxyFactoryBean.

Spring框架参考文档 Spring Framework Reference Documentation

Integrating AOP proxy creation with the IoC framework is best practice in most applications. We recommend that you externalize configuration from Java code with AOP, as in general.

12.8 Manipulating advised objects

However you create AOP proxies, you can manipulate them using the org.springframework.aop.framework.Advised interface. Any AOP proxy can be cast to this interface, whichever other interfaces it implements. This interface includes the following methods:

Advisor[] getAdvisors();

void addAdvice(Advice advice) throws AopConfigException;

void addAdvice(int pos, Advice advice) throws AopConfigException;

void addAdvisor(Advisor advisor) throws AopConfigException;

void addAdvisor(int pos, Advisor advisor) throws AopConfigException;

int indexOf(Advisor advisor);

boolean removeAdvisor(Advisor advisor) throws AopConfigException;

void removeAdvisor(int index) throws AopConfigException;

boolean replaceAdvisor(Advisor a, Advisor b) throws AopConfigException;

boolean isFrozen();

The getAdvisors() method will return an Advisor for every advisor, interceptor or other advice type that has been added to the factory. If you added an Advisor, the returned advisor at this index will be the object that you added. If you added an interceptor or other advice type, Spring will have wrapped this in an advisor with a pointcut that always returns true. Thus if you added a MethodInterceptor, the advisor returned for this index will be an DefaultPointcutAdvisor returning yourMethodInterceptor and a pointcut that matches all classes and methods.

The addAdvisor() methods can be used to add any Advisor. Usually the advisor holding pointcut and advice will be the generic DefaultPointcutAdvisor, which can be used with any advice or pointcut (but not for introductions).

By default, it’s possible to add or remove advisors or interceptors even once a proxy has been created. The only restriction is that it’s impossible to add or remove an introduction advisor, as existing proxies from the factory will not show the interface change. (You can obtain a new proxy from the factory to avoid this problem.)

A simple example of casting an AOP proxy to the Advised interface and examining and manipulating its advice:

Advised advised = (Advised) myObject;
Advisor[] advisors = advised.getAdvisors();
int oldAdvisorCount = advisors.length;
System.out.println(oldAdvisorCount + " advisors");

// Add an advice like an interceptor without a pointcut
// Will match all proxied methods
// Can use for interceptors, before, after returning or throws advice
advised.addAdvice(new DebugInterceptor());

// Add selective advice using a pointcut
advised.addAdvisor(new DefaultPointcutAdvisor(mySpecialPointcut, myAdvice));

assertEquals("Added two advisors", oldAdvisorCount + 2, advised.getAdvisors().length);
Spring框架参考文档 Spring Framework Reference Documentation

It’s questionable whether it’s advisable (no pun intended) to modify advice on a business object in production, although there are no doubt legitimate usage cases. However, it can be very useful in development: for example, in tests. I have sometimes found it very useful to be able to add test code in the form of an interceptor or other advice, getting inside a method invocation I want to test. (For example, the advice can get inside a transaction created for that method: for example, to run SQL to check that a database was correctly updated, before marking the transaction for roll back.)

Depending on how you created the proxy, you can usually set a frozen flag, in which case the Advised isFrozen() method will return true, and any attempts to modify advice through addition or removal will result in an AopConfigException. The ability to freeze the state of an advised object is useful in some cases, for example, to prevent calling code removing a security interceptor. It may also be used in Spring 1.1 to allow aggressive optimization if runtime advice modification is known not to be required.

12.9 Using the "auto-proxy" facility

So far we’ve considered explicit creation of AOP proxies using a ProxyFactoryBean or similar factory bean.

Spring also allows us to use "auto-proxy" bean definitions, which can automatically proxy selected bean definitions. This is built on Spring "bean post processor" infrastructure, which enables modification of any bean definition as the container loads.

In this model, you set up some special bean definitions in your XML bean definition file to configure the auto proxy infrastructure. This allows you just to declare the targets eligible for auto-proxying: you don’t need to use ProxyFactoryBean.

There are two ways to do this:

  • Using an auto-proxy creator that refers to specific beans in the current context.
  • A special case of auto-proxy creation that deserves to be considered separately; auto-proxy creation driven by source-level metadata attributes.

12.9.1 Autoproxy bean definitions

The org.springframework.aop.framework.autoproxy package provides the following standard auto-proxy creators.

BeanNameAutoProxyCreator

The BeanNameAutoProxyCreator class is a BeanPostProcessor that automatically creates AOP proxies for beans with names matching literal values or wildcards.

<bean class="org.springframework.aop.framework.autoproxy.BeanNameAutoProxyCreator">
    <property name="beanNames" value="jdk*,onlyJdk"/>
    <property name="interceptorNames">
        <list>
            <value>myInterceptor</value>
        </list>
    </property>
</bean>

As with ProxyFactoryBean, there is an interceptorNames property rather than a list of interceptors, to allow correct behavior for prototype advisors. Named "interceptors" can be advisors or any advice type.

As with auto proxying in general, the main point of using BeanNameAutoProxyCreator is to apply the same configuration consistently to multiple objects, with minimal volume of configuration. It is a popular choice for applying declarative transactions to multiple objects.

Bean definitions whose names match, such as "jdkMyBean" and "onlyJdk" in the above example, are plain old bean definitions with the target class. An AOP proxy will be created automatically by the BeanNameAutoProxyCreator. The same advice will be applied to all matching beans. Note that if advisors are used (rather than the interceptor in the above example), the pointcuts may apply differently to different beans.

DefaultAdvisorAutoProxyCreator

A more general and extremely powerful auto proxy creator is DefaultAdvisorAutoProxyCreator. This will automagically apply eligible advisors in the current context, without the need to include specific bean names in the auto-proxy advisor’s bean definition. It offers the same merit of consistent configuration and avoidance of duplication as BeanNameAutoProxyCreator.

Using this mechanism involves:

  • Specifying a DefaultAdvisorAutoProxyCreator bean definition.
  • Specifying any number of Advisors in the same or related contexts. Note that these must be Advisors, not just interceptors or other advices. This is necessary because there must be a pointcut to evaluate, to check the eligibility of each advice to candidate bean definitions.

The DefaultAdvisorAutoProxyCreator will automatically evaluate the pointcut contained in each advisor, to see what (if any) advice it should apply to each business object (such as "businessObject1" and "businessObject2" in the example).

This means that any number of advisors can be applied automatically to each business object. If no pointcut in any of the advisors matches any method in a business object, the object will not be proxied. As bean definitions are added for new business objects, they will automatically be proxied if necessary.

Autoproxying in general has the advantage of making it impossible for callers or dependencies to obtain an un-advised object. Calling getBean("businessObject1") on this ApplicationContext will return an AOP proxy, not the target business object. (The "inner bean" idiom shown earlier also offers this benefit.)

<bean class="org.springframework.aop.framework.autoproxy.DefaultAdvisorAutoProxyCreator"/>

<bean class="org.springframework.transaction.interceptor.TransactionAttributeSourceAdvisor">
    <property name="transactionInterceptor" ref="transactionInterceptor"/>
</bean>

<bean id="customAdvisor" class="com.mycompany.MyAdvisor"/>

<bean id="businessObject1" class="com.mycompany.BusinessObject1">
    <!-- Properties omitted -->
</bean>

<bean id="businessObject2" class="com.mycompany.BusinessObject2"/>

The DefaultAdvisorAutoProxyCreator is very useful if you want to apply the same advice consistently to many business objects. Once the infrastructure definitions are in place, you can simply add new business objects without including specific proxy configuration. You can also drop in additional aspects very easily - for example, tracing or performance monitoring aspects - with minimal change to configuration.

The DefaultAdvisorAutoProxyCreator offers support for filtering (using a naming convention so that only certain advisors are evaluated, allowing use of multiple, differently configured, AdvisorAutoProxyCreators in the same factory) and ordering. Advisors can implement the org.springframework.core.Ordered interface to ensure correct ordering if this is an issue. The TransactionAttributeSourceAdvisor used in the above example has a configurable order value; the default setting is unordered.

AbstractAdvisorAutoProxyCreator

This is the superclass of DefaultAdvisorAutoProxyCreator. You can create your own auto-proxy creators by subclassing this class, in the unlikely event that advisor definitions offer insufficient customization to the behavior of the framework DefaultAdvisorAutoProxyCreator.

12.9.2 Using metadata-driven auto-proxying

A particularly important type of auto-proxying is driven by metadata. This produces a similar programming model to .NET ServicedComponents. Instead of defining metadata in XML descriptors, configuration for transaction management and other enterprise services is held in source-level attributes.

In this case, you use the DefaultAdvisorAutoProxyCreator, in combination with Advisors that understand metadata attributes. The metadata specifics are held in the pointcut part of the candidate advisors, rather than in the auto-proxy creation class itself.

This is really a special case of the DefaultAdvisorAutoProxyCreator, but deserves consideration on its own. (The metadata-aware code is in the pointcuts contained in the advisors, not the AOP framework itself.)

The /attributes directory of the JPetStore sample application shows the use of attribute-driven auto-proxying. In this case, there’s no need to use theTransactionProxyFactoryBean. Simply defining transactional attributes on business objects is sufficient, because of the use of metadata-aware pointcuts. The bean definitions include the following code, in /WEB-INF/declarativeServices.xml. Note that this is generic, and can be used outside the JPetStore:

<bean class="org.springframework.aop.framework.autoproxy.DefaultAdvisorAutoProxyCreator"/>

<bean class="org.springframework.transaction.interceptor.TransactionAttributeSourceAdvisor">
    <property name="transactionInterceptor" ref="transactionInterceptor"/>
</bean>

<bean id="transactionInterceptor"
        class="org.springframework.transaction.interceptor.TransactionInterceptor">
    <property name="transactionManager" ref="transactionManager"/>
    <property name="transactionAttributeSource">
        <bean class="org.springframework.transaction.interceptor.AttributesTransactionAttributeSource">
            <property name="attributes" ref="attributes"/>
        </bean>
    </property>
</bean>

<bean id="attributes" class="org.springframework.metadata.commons.CommonsAttributes"/>

The DefaultAdvisorAutoProxyCreator bean definition (the name is not significant, hence it can even be omitted) will pick up all eligible pointcuts in the current application context. In this case, the "transactionAdvisor" bean definition, of type TransactionAttributeSourceAdvisor, will apply to classes or methods carrying a transaction attribute. The TransactionAttributeSourceAdvisor depends on a TransactionInterceptor, via constructor dependency. The example resolves this via autowiring. The AttributesTransactionAttributeSource depends on an implementation of the org.springframework.metadata.Attributes interface. In this fragment, the "attributes" bean satisfies this, using the Jakarta Commons Attributes API to obtain attribute information. (The application code must have been compiled using the Commons Attributes compilation task.)

The /annotation directory of the JPetStore sample application contains an analogous example for auto-proxying driven by JDK 1.5+ annotations. The following configuration enables automatic detection of Spring’s Transactional annotation, leading to implicit proxies for beans containing that annotation:

<bean class="org.springframework.aop.framework.autoproxy.DefaultAdvisorAutoProxyCreator"/>

<bean class="org.springframework.transaction.interceptor.TransactionAttributeSourceAdvisor">
    <property name="transactionInterceptor" ref="transactionInterceptor"/>
</bean>

<bean id="transactionInterceptor"
        class="org.springframework.transaction.interceptor.TransactionInterceptor">
    <property name="transactionManager" ref="transactionManager"/>
    <property name="transactionAttributeSource">
        <bean class="org.springframework.transaction.annotation.AnnotationTransactionAttributeSource"/>
    </property>
</bean>

The TransactionInterceptor defined here depends on a PlatformTransactionManager definition, which is not included in this generic file (although it could be) because it will be specific to the application’s transaction requirements (typically JTA, as in this example, or Hibernate, JDO or JDBC):

<bean id="transactionManager"
        class="org.springframework.transaction.jta.JtaTransactionManager"/>
Spring框架参考文档 Spring Framework Reference Documentation

If you require only declarative transaction management, using these generic XML definitions will result in Spring automatically proxying all classes or methods with transaction attributes. You won’t need to work directly with AOP, and the programming model is similar to that of .NET ServicedComponents.

This mechanism is extensible. It’s possible to do auto-proxying based on custom attributes. You need to:

  • Define your custom attribute.
  • Specify an Advisor with the necessary advice, including a pointcut that is triggered by the presence of the custom attribute on a class or method. You may be able to use an existing advice, merely implementing a static pointcut that picks up the custom attribute.

It’s possible for such advisors to be unique to each advised class (for example, mixins): they simply need to be defined as prototype, rather than singleton, bean definitions. For example, the LockMixin introduction interceptor from the Spring test suite, shown above, could be used in conjunction with a genericDefaultIntroductionAdvisor:

<bean id="lockMixin" class="test.mixin.LockMixin" scope="prototype"/>

<bean id="lockableAdvisor" class="org.springframework.aop.support.DefaultIntroductionAdvisor"
        scope="prototype">
    <constructor-arg ref="lockMixin"/>
</bean>

Note that both lockMixin and lockableAdvisor are defined as prototypes.

12.10 Using TargetSources

Spring offers the concept of a TargetSource, expressed in the org.springframework.aop.TargetSource interface. This interface is responsible for returning the "target object" implementing the join point. The TargetSource implementation is asked for a target instance each time the AOP proxy handles a method invocation.

Developers using Spring AOP don’t normally need to work directly with TargetSources, but this provides a powerful means of supporting pooling, hot swappable and other sophisticated targets. For example, a pooling TargetSource can return a different target instance for each invocation, using a pool to manage instances.

If you do not specify a TargetSource, a default implementation is used that wraps a local object. The same target is returned for each invocation (as you would expect).

Let’s look at the standard target sources provided with Spring, and how you can use them.

Spring框架参考文档 Spring Framework Reference Documentation

When using a custom target source, your target will usually need to be a prototype rather than a singleton bean definition. This allows Spring to create a new target instance when required.

12.10.1 Hot swappable target sources

The org.springframework.aop.target.HotSwappableTargetSource exists to allow the target of an AOP proxy to be switched while allowing callers to keep their references to it.

Changing the target source’s target takes effect immediately. The HotSwappableTargetSource is threadsafe.

You can change the target via the swap() method on HotSwappableTargetSource as follows:

HotSwappableTargetSource swapper = (HotSwappableTargetSource) beanFactory.getBean("swapper");
Object oldTarget = swapper.swap(newTarget);

The XML definitions required look as follows:

<bean id="initialTarget" class="mycompany.OldTarget"/>

<bean id="swapper" class="org.springframework.aop.target.HotSwappableTargetSource">
    <constructor-arg ref="initialTarget"/>
</bean>

<bean id="swappable" class="org.springframework.aop.framework.ProxyFactoryBean">
    <property name="targetSource" ref="swapper"/>
</bean>

The above swap() call changes the target of the swappable bean. Clients who hold a reference to that bean will be unaware of the change, but will immediately start hitting the new target.

Although this example doesn’t add any advice - and it’s not necessary to add advice to use a TargetSource - of course any TargetSource can be used in conjunction with arbitrary advice.

12.10.2 Pooling target sources

Using a pooling target source provides a similar programming model to stateless session EJBs, in which a pool of identical instances is maintained, with method invocations going to free objects in the pool.

A crucial difference between Spring pooling and SLSB pooling is that Spring pooling can be applied to any POJO. As with Spring in general, this service can be applied in a non-invasive way.

Spring provides out-of-the-box support for Commons Pool 2.2, which provides a fairly efficient pooling implementation. You’ll need the commons-pool Jar on your application’s classpath to use this feature. It’s also possible to subclass org.springframework.aop.target.AbstractPoolingTargetSource to support any other pooling API.

Spring框架参考文档 Spring Framework Reference Documentation

Commons Pool 1.5+ is also supported but deprecated as of Spring Framework 4.2.

Sample configuration is shown below:

<bean id="businessObjectTarget" class="com.mycompany.MyBusinessObject"
        scope="prototype">
    ... properties omitted
</bean>

<bean id="poolTargetSource" class="org.springframework.aop.target.CommonsPool2TargetSource">
    <property name="targetBeanName" value="businessObjectTarget"/>
    <property name="maxSize" value="25"/>
</bean>

<bean id="businessObject" class="org.springframework.aop.framework.ProxyFactoryBean">
    <property name="targetSource" ref="poolTargetSource"/>
    <property name="interceptorNames" value="myInterceptor"/>
</bean>

Note that the target object - "businessObjectTarget" in the example - must be a prototype. This allows the PoolingTargetSource implementation to create new instances of the target to grow the pool as necessary. See the javadocs of AbstractPoolingTargetSource and the concrete subclass you wish to use for information about its properties: "maxSize" is the most basic, and always guaranteed to be present.

In this case, "myInterceptor" is the name of an interceptor that would need to be defined in the same IoC context. However, it isn’t necessary to specify interceptors to use pooling. If you want only pooling, and no other advice, don’t set the interceptorNames property at all.

It’s possible to configure Spring so as to be able to cast any pooled object to the org.springframework.aop.target.PoolingConfig interface, which exposes information about the configuration and current size of the pool through an introduction. You’ll need to define an advisor like this:

<bean id="poolConfigAdvisor" class="org.springframework.beans.factory.config.MethodInvokingFactoryBean">
    <property name="targetObject" ref="poolTargetSource"/>
    <property name="targetMethod" value="getPoolingConfigMixin"/>
</bean>

This advisor is obtained by calling a convenience method on the AbstractPoolingTargetSource class, hence the use of MethodInvokingFactoryBean. This advisor’s name ("poolConfigAdvisor" here) must be in the list of interceptors names in the ProxyFactoryBean exposing the pooled object.

The cast will look as follows:

PoolingConfig conf = (PoolingConfig) beanFactory.getBean("businessObject");
System.out.println("Max pool size is " + conf.getMaxSize());
Spring框架参考文档 Spring Framework Reference Documentation

Pooling stateless service objects is not usually necessary. We don’t believe it should be the default choice, as most stateless objects are naturally thread safe, and instance pooling is problematic if resources are cached.

Simpler pooling is available using auto-proxying. It’s possible to set the TargetSources used by any auto-proxy creator.

12.10.3 Prototype target sources

Setting up a "prototype" target source is similar to a pooling TargetSource. In this case, a new instance of the target will be created on every method invocation. Although the cost of creating a new object isn’t high in a modern JVM, the cost of wiring up the new object (satisfying its IoC dependencies) may be more expensive. Thus you shouldn’t use this approach without very good reason.

To do this, you could modify the poolTargetSource definition shown above as follows. (I’ve also changed the name, for clarity.)

<bean id="prototypeTargetSource" class="org.springframework.aop.target.PrototypeTargetSource">
    <property name="targetBeanName" ref="businessObjectTarget"/>
</bean>

There’s only one property: the name of the target bean. Inheritance is used in the TargetSource implementations to ensure consistent naming. As with the pooling target source, the target bean must be a prototype bean definition.

12.10.4 ThreadLocal target sources

ThreadLocal target sources are useful if you need an object to be created for each incoming request (per thread that is). The concept of a ThreadLocal provide a JDK-wide facility to transparently store resource alongside a thread. Setting up a ThreadLocalTargetSource is pretty much the same as was explained for the other types of target source:

<bean id="threadlocalTargetSource" class="org.springframework.aop.target.ThreadLocalTargetSource">
    <property name="targetBeanName" value="businessObjectTarget"/>
</bean>
Spring框架参考文档 Spring Framework Reference Documentation

ThreadLocals come with serious issues (potentially resulting in memory leaks) when incorrectly using them in a multi-threaded and multi-classloader environments. One should always consider wrapping a threadlocal in some other class and never directly use the ThreadLocal itself (except of course in the wrapper class). Also, one should always remember to correctly set and unset (where the latter simply involved a call to ThreadLocal.set(null)) the resource local to the thread. Unsetting should be done in any case since not unsetting it might result in problematic behavior. Spring’s ThreadLocal support does this for you and should always be considered in favor of using ThreadLocals without other proper handling code.

12.11 Defining new Advice types

Spring AOP is designed to be extensible. While the interception implementation strategy is presently used internally, it is possible to support arbitrary advice types in addition to the out-of-the-box interception around advice, before, throws advice and after returning advice.

The org.springframework.aop.framework.adapter package is an SPI package allowing support for new custom advice types to be added without changing the core framework. The only constraint on a custom Advice type is that it must implement the org.aopalliance.aop.Advice marker interface.

Please refer to the org.springframework.aop.framework.adapter javadocs for further information.

12.12 Further resources

Please refer to the Spring sample applications for further examples of Spring AOP:

  • The JPetStore’s default configuration illustrates the use of the TransactionProxyFactoryBean for declarative transaction management.
  • The /attributes directory of the JPetStore illustrates the use of attribute-driven declarative transaction management.

Part IV. Testing

 

The adoption of the test-driven-development (TDD) approach to software development is certainly advocated by the Spring team, and so coverage of Spring’s support for integration testing is covered (alongside best practices for unit testing). The Spring team has found that the correct use of IoC certainly does make both unit and integration testing easier (in that the presence of setter methods and appropriate constructors on classes makes them easier to wire together in a test without having to set up service locator registries and suchlike)…​ the chapter dedicated solely to testing will hopefully convince you of this as well.

13. Introduction to Spring Testing

Testing is an integral part of enterprise software development. This chapter focuses on the value-add of the IoC principle to unit testing and on the benefits of the Spring Framework’s support for integration testing. (A thorough treatment of testing in the enterprise is beyond the scope of this reference manual.)

14. Unit Testing

Dependency Injection should make your code less dependent on the container than it would be with traditional Java EE development. The POJOs that make up your application should be testable in JUnit or TestNG tests, with objects simply instantiated using the new operator, without Spring or any other container. You can use mock objects (in conjunction with other valuable testing techniques) to test your code in isolation. If you follow the architecture recommendations for Spring, the resulting clean layering and componentization of your codebase will facilitate easier unit testing. For example, you can test service layer objects by stubbing or mocking DAO or Repository interfaces, without needing to access persistent data while running unit tests.

True unit tests typically run extremely quickly, as there is no runtime infrastructure to set up. Emphasizing true unit tests as part of your development methodology will boost your productivity. You may not need this section of the testing chapter to help you write effective unit tests for your IoC-based applications. For certain unit testing scenarios, however, the Spring Framework provides the following mock objects and testing support classes.

14.1 Mock Objects

14.1.1 Environment

The org.springframework.mock.env package contains mock implementations of the Environment and PropertySource abstractions (see Section 7.13.1, “Bean definition profiles” and Section 7.13.3, “PropertySource abstraction”). MockEnvironment and MockPropertySource are useful for developing out-of-container tests for code that depends on environment-specific properties.

14.1.2 JNDI

The org.springframework.mock.jndi package contains an implementation of the JNDI SPI, which you can use to set up a simple JNDI environment for test suites or stand-alone applications. If, for example, JDBC DataSources get bound to the same JNDI names in test code as within a Java EE container, you can reuse both application code and configuration in testing scenarios without modification.

14.1.3 Servlet API

The org.springframework.mock.web package contains a comprehensive set of Servlet API mock objects, which are useful for testing web contexts, controllers, and filters. These mock objects are targeted at usage with Spring’s Web MVC framework and are generally more convenient to use than dynamic mock objects such asEasyMock or alternative Servlet API mock objects such as MockObjects. Since Spring Framework 4.0, the set of mocks in the org.springframework.mock.webpackage is based on the Servlet 3.0 API.

For thorough integration testing of your Spring MVC and REST Controllers in conjunction with your WebApplicationContext configuration for Spring MVC, see theSpring MVC Test Framework.

14.1.4 Portlet API

The org.springframework.mock.web.portlet package contains a set of Portlet API mock objects, targeted at usage with Spring’s Portlet MVC framework.

14.2 Unit Testing support Classes

14.2.1 General testing utilities

The org.springframework.test.util package contains several general purpose utilities for use in unit and integration testing.

ReflectionTestUtils is a collection of reflection-based utility methods. Developers use these methods in testing scenarios where they need to change the value of a constant, set a non-public field, invoke a non-public setter method, or invoke a non-public configuration or lifecycle callback method when testing application code involving use cases such as the following.

  • ORM frameworks such as JPA and Hibernate that condone private or protected field access as opposed to public setter methods for properties in a domain entity.
  • Spring’s support for annotations such as @Autowired, @Inject, and @Resource, which provides dependency injection for private or protected fields, setter methods, and configuration methods.
  • Use of annotations such as @PostConstruct and @PreDestroy for lifecycle callback methods.

AopTestUtils is a collection of AOP-related utility methods. These methods can be used to obtain a reference to the underlying target object hidden behind one or more Spring proxies. For example, if you have configured a bean as a dynamic mock using a library like EasyMock or Mockito and the mock is wrapped in a Spring proxy, you may need direct access to the underlying mock in order to configure expectations on it and perform verifications. For Spring’s core AOP utilities, seeAopUtils and AopProxyUtils.

14.2.2 Spring MVC

The org.springframework.test.web package contains ModelAndViewAssert, which you can use in combination with JUnit, TestNG, or any other testing framework for unit tests dealing with Spring MVC ModelAndView objects.

Spring框架参考文档 Spring Framework Reference Documentation

To unit test your Spring MVC Controllers as POJOs, use ModelAndViewAssert combined with MockHttpServletRequest, MockHttpSession, and so on from Spring’s Servlet API mocks. For thorough integration testing of your Spring MVC and REST Controllers in conjunction with yourWebApplicationContext configuration for Spring MVC, use the Spring MVC Test Framework instead.

15. Integration Testing

15.1 Overview

It is important to be able to perform some integration testing without requiring deployment to your application server or connecting to other enterprise infrastructure. This will enable you to test things such as:

  • The correct wiring of your Spring IoC container contexts.
  • Data access using JDBC or an ORM tool. This would include such things as the correctness of SQL statements, Hibernate queries, JPA entity mappings, etc.

The Spring Framework provides first-class support for integration testing in the spring-test module. The name of the actual JAR file might include the release version and might also be in the long org.springframework.test form, depending on where you get it from (see the section on Dependency Management for an explanation). This library includes the org.springframework.test package, which contains valuable classes for integration testing with a Spring container. This testing does not rely on an application server or other deployment environment. Such tests are slower to run than unit tests but much faster than the equivalent Selenium tests or remote tests that rely on deployment to an application server.

In Spring 2.5 and later, unit and integration testing support is provided in the form of the annotation-driven Spring TestContext Framework. The TestContext framework is agnostic of the actual testing framework in use, thus allowing instrumentation of tests in various environments including JUnit, TestNG, and so on.

15.2 Goals of Integration Testing

Spring’s integration testing support has the following primary goals:

The next few sections describe each goal and provide links to implementation and configuration details.

15.2.1 Context management and caching

The Spring TestContext Framework provides consistent loading of Spring ApplicationContexts and WebApplicationContexts as well as caching of those contexts. Support for the caching of loaded contexts is important, because startup time can become an issue — not because of the overhead of Spring itself, but because the objects instantiated by the Spring container take time to instantiate. For example, a project with 50 to 100 Hibernate mapping files might take 10 to 20 seconds to load the mapping files, and incurring that cost before running every test in every test fixture leads to slower overall test runs that reduce developer productivity.

Test classes typically declare either an array of resource locations for XML or Groovy configuration metadata — often in the classpath — or an array of annotated classesthat is used to configure the application. These locations or classes are the same as or similar to those specified in web.xml or other configuration files for production deployments.

By default, once loaded, the configured ApplicationContext is reused for each test. Thus the setup cost is incurred only once per test suite, and subsequent test execution is much faster. In this context, the term test suite means all tests run in the same JVM — for example, all tests run from an Ant, Maven, or Gradle build for a given project or module. In the unlikely case that a test corrupts the application context and requires reloading — for example, by modifying a bean definition or the state of an application object — the TestContext framework can be configured to reload the configuration and rebuild the application context before executing the next test.

See Section 15.5.4, “Context management” and the section called “Context caching” with the TestContext framework.

15.2.2 Dependency Injection of test fixtures

When the TestContext framework loads your application context, it can optionally configure instances of your test classes via Dependency Injection. This provides a convenient mechanism for setting up test fixtures using preconfigured beans from your application context. A strong benefit here is that you can reuse application contexts across various testing scenarios (e.g., for configuring Spring-managed object graphs, transactional proxies, DataSources, etc.), thus avoiding the need to duplicate complex test fixture setup for individual test cases.

As an example, consider the scenario where we have a class, HibernateTitleRepository, that implements data access logic for a Title domain entity. We want to write integration tests that test the following areas:

  • The Spring configuration: basically, is everything related to the configuration of the HibernateTitleRepository bean correct and present?
  • The Hibernate mapping file configuration: is everything mapped correctly, and are the correct lazy-loading settings in place?
  • The logic of the HibernateTitleRepository: does the configured instance of this class perform as anticipated?

See dependency injection of test fixtures with the TestContext framework.

15.2.3 Transaction management

One common issue in tests that access a real database is their effect on the state of the persistence store. Even when you’re using a development database, changes to the state may affect future tests. Also, many operations — such as inserting or modifying persistent data — cannot be performed (or verified) outside a transaction.

The TestContext framework addresses this issue. By default, the framework will create and roll back a transaction for each test. You simply write code that can assume the existence of a transaction. If you call transactionally proxied objects in your tests, they will behave correctly, according to their configured transactional semantics. In addition, if a test method deletes the contents of selected tables while running within the transaction managed for the test, the transaction will roll back by default, and the database will return to its state prior to execution of the test. Transactional support is provided to a test via a PlatformTransactionManager bean defined in the test’s application context.

If you want a transaction to commit — unusual, but occasionally useful when you want a particular test to populate or modify the database — the TestContext framework can be instructed to cause the transaction to commit instead of roll back via the @Commit annotation.

See transaction management with the TestContext framework.

15.2.4 Support classes for integration testing

The Spring TestContext Framework provides several abstract support classes that simplify the writing of integration tests. These base test classes provide well-defined hooks into the testing framework as well as convenient instance variables and methods, which enable you to access:

  • The ApplicationContext, for performing explicit bean lookups or testing the state of the context as a whole.
  • A JdbcTemplate, for executing SQL statements to query the database. Such queries can be used to confirm database state both prior to and after execution of database-related application code, and Spring ensures that such queries run in the scope of the same transaction as the application code. When used in conjunction with an ORM tool, be sure to avoid false positives.

In addition, you may want to create your own custom, application-wide superclass with instance variables and methods specific to your project.

See support classes for the TestContext framework.

15.3 JDBC Testing Support

The org.springframework.test.jdbc package contains JdbcTestUtils, which is a collection of JDBC related utility functions intended to simplify standard database testing scenarios. Specifically, JdbcTestUtils provides the following static utility methods.

  • countRowsInTable(..): counts the number of rows in the given table
  • countRowsInTableWhere(..): counts the number of rows in the given table, using the provided WHERE clause
  • deleteFromTables(..): deletes all rows from the specified tables
  • deleteFromTableWhere(..): deletes rows from the given table, using the provided WHERE clause
  • dropTables(..): drops the specified tables

Note that AbstractTransactionalJUnit4SpringContextTests and AbstractTransactionalTestNGSpringContextTests provide convenience methods which delegate to the aforementioned methods in JdbcTestUtils.

The spring-jdbc module provides support for configuring and launching an embedded database which can be used in integration tests that interact with a database. For details, see Section 19.8, “Embedded database support” and Section 19.8.5, “Testing data access logic with an embedded database”.

15.4 Annotations

15.4.1 Spring Testing Annotations

The Spring Framework provides the following set of Spring-specific annotations that you can use in your unit and integration tests in conjunction with the TestContext framework. Refer to the corresponding javadocs for further information, including default attribute values, attribute aliases, and so on.

@BootstrapWith

@BootstrapWith is a class-level annotation that is used to configure how the Spring TestContext Framework is bootstrapped. Specifically, @BootstrapWith is used to specify a custom TestContextBootstrapper. Consult the Bootstrapping the TestContext framework section for further details.

@ContextConfiguration

@ContextConfiguration defines class-level metadata that is used to determine how to load and configure an ApplicationContext for integration tests. Specifically,@ContextConfiguration declares the application context resource locations or the annotated classes that will be used to load the context.

Resource locations are typically XML configuration files or Groovy scripts located in the classpath; whereas, annotated classes are typically @Configuration classes. However, resource locations can also refer to files and scripts in the file system, and annotated classes can be component classes, etc.

@ContextConfiguration("/test-config.xml")
public class XmlApplicationContextTests {
    // class body...
}
@ContextConfiguration(classes = TestConfig.class)
public class ConfigClassApplicationContextTests {
    // class body...
}

As an alternative or in addition to declaring resource locations or annotated classes, @ContextConfiguration may be used to declareApplicationContextInitializer classes.

@ContextConfiguration(initializers = CustomContextIntializer.class)
public class ContextInitializerTests {
    // class body...
}

@ContextConfiguration may optionally be used to declare the ContextLoader strategy as well. Note, however, that you typically do not need to explicitly configure the loader since the default loader supports either resource locations or annotated classes as well as initializers.

@ContextConfiguration(locations = "/test-context.xml", loader = CustomContextLoader.class)
public class CustomLoaderXmlApplicationContextTests {
    // class body...
}
Spring框架参考文档 Spring Framework Reference Documentation

@ContextConfiguration provides support for inheriting resource locations or configuration classes as well as context initializers declared by superclasses by default.

See Section 15.5.4, “Context management” and the @ContextConfiguration javadocs for further details.

@WebAppConfiguration

@WebAppConfiguration is a class-level annotation that is used to declare that the ApplicationContext loaded for an integration test should be aWebApplicationContext. The mere presence of @WebAppConfiguration on a test class ensures that a WebApplicationContext will be loaded for the test, using the default value of "file:src/main/webapp" for the path to the root of the web application (i.e., the resource base path). The resource base path is used behind the scenes to create a MockServletContext which serves as the ServletContext for the test’s WebApplicationContext.

@ContextConfiguration
@WebAppConfiguration
public class WebAppTests {
    // class body...
}

To override the default, specify a different base resource path via the implicit value attribute. Both classpath: and file: resource prefixes are supported. If no resource prefix is supplied the path is assumed to be a file system resource.

@ContextConfiguration
@WebAppConfiguration("classpath:test-web-resources")
public class WebAppTests {
    // class body...
}

Note that @WebAppConfiguration must be used in conjunction with @ContextConfiguration, either within a single test class or within a test class hierarchy. See the@WebAppConfiguration javadocs for further details.

@ContextHierarchy

@ContextHierarchy is a class-level annotation that is used to define a hierarchy of ApplicationContexts for integration tests. @ContextHierarchy should be declared with a list of one or more @ContextConfiguration instances, each of which defines a level in the context hierarchy. The following examples demonstrate the use of @ContextHierarchy within a single test class; however, @ContextHierarchy can also be used within a test class hierarchy.

@ContextHierarchy({
    @ContextConfiguration("/parent-config.xml"),
    @ContextConfiguration("/child-config.xml")
})
public class ContextHierarchyTests {
    // class body...
}
@WebAppConfiguration
@ContextHierarchy({
    @ContextConfiguration(classes = AppConfig.class),
    @ContextConfiguration(classes = WebConfig.class)
})
public class WebIntegrationTests {
    // class body...
}

If you need to merge or override the configuration for a given level of the context hierarchy within a test class hierarchy, you must explicitly name that level by supplying the same value to the name attribute in @ContextConfiguration at each corresponding level in the class hierarchy. See the section called “Context hierarchies” and the @ContextHierarchy javadocs for further examples.

@ActiveProfiles

@ActiveProfiles is a class-level annotation that is used to declare which bean definition profiles should be active when loading an ApplicationContext for an integration test.

@ContextConfiguration
@ActiveProfiles("dev")
public class DeveloperTests {
    // class body...
}
@ContextConfiguration
@ActiveProfiles({"dev", "integration"})
public class DeveloperIntegrationTests {
    // class body...
}
Spring框架参考文档 Spring Framework Reference Documentation

@ActiveProfiles provides support for inheriting active bean definition profiles declared by superclasses by default. It is also possible to resolve active bean definition profiles programmatically by implementing a custom ActiveProfilesResolver and registering it via the resolver attribute of@ActiveProfiles.

See the section called “Context configuration with environment profiles” and the @ActiveProfiles javadocs for examples and further details.

@TestPropertySource

@TestPropertySource is a class-level annotation that is used to configure the locations of properties files and inlined properties to be added to the set ofPropertySources in the Environment for an ApplicationContext loaded for an integration test.

Test property sources have higher precedence than those loaded from the operating system’s environment or Java system properties as well as property sources added by the application declaratively via @PropertySource or programmatically. Thus, test property sources can be used to selectively override properties defined in system and application property sources. Furthermore, inlined properties have higher precedence than properties loaded from resource locations.

The following example demonstrates how to declare a properties file from the classpath.

@ContextConfiguration
@TestPropertySource("/test.properties")
public class MyIntegrationTests {
    // class body...
}

The following example demonstrates how to declare inlined properties.

@ContextConfiguration
@TestPropertySource(properties = { "timezone = GMT", "port: 4242" })
public class MyIntegrationTests {
    // class body...
}

@DirtiesContext

@DirtiesContext indicates that the underlying Spring ApplicationContext has been dirtied during the execution of a test (i.e., modified or corrupted in some manner — for example, by changing the state of a singleton bean) and should be closed. When an application context is marked dirty, it is removed from the testing framework’s cache and closed. As a consequence, the underlying Spring container will be rebuilt for any subsequent test that requires a context with the same configuration metadata.

@DirtiesContext can be used as both a class-level and method-level annotation within the same class or class hierarchy. In such scenarios, theApplicationContext is marked as dirty before or after any such annotated method as well as before or after the current test class, depending on the configuredmethodMode and classMode.

The following examples explain when the context would be dirtied for various configuration scenarios:

  • Before the current test class, when declared on a class with class mode set to BEFORE_CLASS.

    @DirtiesContext(classMode = BEFORE_CLASS)
public class FreshContextTests {
    // some tests that require a new Spring container
}

  • After the current test class, when declared on a class with class mode set to AFTER_CLASS (i.e., the default class mode).

    @DirtiesContext
public class ContextDirtyingTests {
    // some tests that result in the Spring container being dirtied
}

  • Before each test method in the current test class, when declared on a class with class mode set to BEFORE_EACH_TEST_METHOD.

    @DirtiesContext(classMode = BEFORE_EACH_TEST_METHOD)
public class FreshContextTests {
    // some tests that require a new Spring container
}

  • After each test method in the current test class, when declared on a class with class mode set to AFTER_EACH_TEST_METHOD.

    @DirtiesContext(classMode = AFTER_EACH_TEST_METHOD)
public class ContextDirtyingTests {
    // some tests that result in the Spring container being dirtied
}

  • Before the current test, when declared on a method with the method mode set to BEFORE_METHOD.

    @DirtiesContext(methodMode = BEFORE_METHOD)
@Test
public void testProcessWhichRequiresFreshAppCtx() {
    // some logic that requires a new Spring container
}

  • After the current test, when declared on a method with the method mode set to AFTER_METHOD (i.e., the default method mode).

    @DirtiesContext
@Test
public void testProcessWhichDirtiesAppCtx() {
    // some logic that results in the Spring container being dirtied
}

If @DirtiesContext is used in a test whose context is configured as part of a context hierarchy via @ContextHierarchy, the hierarchyMode flag can be used to control how the context cache is cleared. By default an exhaustive algorithm will be used that clears the context cache including not only the current level but also all other context hierarchies that share an ancestor context common to the current test; all ApplicationContexts that reside in a sub-hierarchy of the common ancestor context will be removed from the context cache and closed. If the exhaustive algorithm is overkill for a particular use case, the simpler current level algorithm can be specified instead, as seen below.

@ContextHierarchy({
    @ContextConfiguration("/parent-config.xml"),
    @ContextConfiguration("/child-config.xml")
})
public class BaseTests {
    // class body...
}

public class ExtendedTests extends BaseTests {

    @Test
    @DirtiesContext(hierarchyMode = CURRENT_LEVEL)
    public void test() {
        // some logic that results in the child context being dirtied
    }
}

For further details regarding the EXHAUSTIVE and CURRENT_LEVEL algorithms see the DirtiesContext.HierarchyMode javadocs.

@TestExecutionListeners

@TestExecutionListeners defines class-level metadata for configuring the TestExecutionListener implementations that should be registered with theTestContextManager. Typically, @TestExecutionListeners is used in conjunction with @ContextConfiguration.

@ContextConfiguration
@TestExecutionListeners({CustomTestExecutionListener.class, AnotherTestExecutionListener.class})
public class CustomTestExecutionListenerTests {
    // class body...
}

@TestExecutionListeners supports inherited listeners by default. See the javadocs for an example and further details.

@Commit

@Commit indicates that the transaction for a transactional test method should be committed after the test method has completed. @Commit can be used as a direct replacement for @Rollback(false) in order to more explicitly convey the intent of the code. Analogous to @Rollback, @Commit may also be declared as a class-level or method-level annotation.

@Commit
@Test
public void testProcessWithoutRollback() {
    // ...
}

@Rollback

@Rollback indicates whether the transaction for a transactional test method should be rolled back after the test method has completed. If true, the transaction is rolled back; otherwise, the transaction is committed (see also @Commit). Rollback semantics for integration tests in the Spring TestContext Framework default to true even if@Rollback is not explicitly declared.

When declared as a class-level annotation, @Rollback defines the default rollback semantics for all test methods within the test class hierarchy. When declared as a method-level annotation, @Rollback defines rollback semantics for the specific test method, potentially overriding class-level @Rollback or @Commit semantics.

@Rollback(false)
@Test
public void testProcessWithoutRollback() {
    // ...
}

@BeforeTransaction

@BeforeTransaction indicates that the annotated void method should be executed before a transaction is started for test methods configured to run within a transaction via Spring’s @Transactional annotation. As of Spring Framework 4.3, @BeforeTransaction methods are not required to be public and may be declared on Java 8 based interface default methods.

@BeforeTransaction
void beforeTransaction() {
    // logic to be executed before a transaction is started
}

@AfterTransaction

@AfterTransaction indicates that the annotated void method should be executed after a transaction is ended for test methods configured to run within a transaction via Spring’s @Transactional annotation. As of Spring Framework 4.3, @AfterTransaction methods are not required to be public and may be declared on Java 8 based interface default methods.

@AfterTransaction
void afterTransaction() {
    // logic to be executed after a transaction has ended
}

@Sql

@Sql is used to annotate a test class or test method to configure SQL scripts to be executed against a given database during integration tests.

@Test
@Sql({"/test-schema.sql", "/test-user-data.sql"})
public void userTest {
    // execute code that relies on the test schema and test data
}

See the section called “Executing SQL scripts declaratively with @Sql” for further details.

@SqlConfig

@SqlConfig defines metadata that is used to determine how to parse and execute SQL scripts configured via the @Sql annotation.

@Test
@Sql(
    scripts = "/test-user-data.sql",
    config = @SqlConfig(commentPrefix = "`", separator = "@@")
)
public void userTest {
    // execute code that relies on the test data
}

@SqlGroup

@SqlGroup is a container annotation that aggregates several @Sql annotations. @SqlGroup can be used natively, declaring several nested @Sql annotations, or it can be used in conjunction with Java 8’s support for repeatable annotations, where @Sql can simply be declared several times on the same class or method, implicitly generating this container annotation.

@Test
@SqlGroup({
    @Sql(scripts = "/test-schema.sql", config = @SqlConfig(commentPrefix = "`")),
    @Sql("/test-user-data.sql")
)}
public void userTest {
    // execute code that uses the test schema and test data
}

15.4.2 Standard Annotation Support

The following annotations are supported with standard semantics for all configurations of the Spring TestContext Framework. Note that these annotations are not specific to tests and can be used anywhere in the Spring Framework.

  • @Autowired
  • @Qualifier
  • @Resource (javax.annotation) if JSR-250 is present
  • @Inject (javax.inject) if JSR-330 is present
  • @Named (javax.inject) if JSR-330 is present
  • @PersistenceContext (javax.persistence) if JPA is present
  • @PersistenceUnit (javax.persistence) if JPA is present
  • @Required
  • @Transactional
Spring框架参考文档 Spring Framework Reference Documentation

In the Spring TestContext Framework @PostConstruct and @PreDestroy may be used with standard semantics on any application components configured in the ApplicationContext; however, these lifecycle annotations have limited usage within an actual test class.

If a method within a test class is annotated with @PostConstruct, that method will be executed before any before methods of the underlying test framework (e.g., methods annotated with JUnit 4’s @Before), and that will apply for every test method in the test class. On the other hand, if a method within a test class is annotated with @PreDestroy, that method will never be executed. Within a test class it is therefore recommended to use test lifecycle callbacks from the underlying test framework instead of @PostConstruct and @PreDestroy.

15.4.3 Spring JUnit 4 Testing Annotations

The following annotations are only supported when used in conjunction with the SpringRunner, Spring’s JUnit rules, or Spring’s JUnit 4 support classes.

@IfProfileValue

@IfProfileValue indicates that the annotated test is enabled for a specific testing environment. If the configured ProfileValueSource returns a matching value for the provided name, the test is enabled. Otherwise, the test will be disabled and effectively ignored.

@IfProfileValue can be applied at the class level, the method level, or both. Class-level usage of @IfProfileValue takes precedence over method-level usage for any methods within that class or its subclasses. Specifically, a test is enabled if it is enabled both at the class level and at the method level; the absence of@IfProfileValue means the test is implicitly enabled. This is analogous to the semantics of JUnit 4’s @Ignore annotation, except that the presence of @Ignorealways disables a test.

@IfProfileValue(name="java.vendor", value="Oracle Corporation")
@Test
public void testProcessWhichRunsOnlyOnOracleJvm() {
    // some logic that should run only on Java VMs from Oracle Corporation
}

Alternatively, you can configure @IfProfileValue with a list of values (with OR semantics) to achieve TestNG-like support for test groups in a JUnit 4 environment. Consider the following example:

@IfProfileValue(name="test-groups", values={"unit-tests", "integration-tests"})
@Test
public void testProcessWhichRunsForUnitOrIntegrationTestGroups() {
    // some logic that should run only for unit and integration test groups
}

@ProfileValueSourceConfiguration

@ProfileValueSourceConfiguration is a class-level annotation that specifies what type of ProfileValueSource to use when retrieving profile values configured through the @IfProfileValue annotation. If @ProfileValueSourceConfiguration is not declared for a test, SystemProfileValueSource is used by default.

@ProfileValueSourceConfiguration(CustomProfileValueSource.class)
public class CustomProfileValueSourceTests {
    // class body...
}

@Timed

@Timed indicates that the annotated test method must finish execution in a specified time period (in milliseconds). If the text execution time exceeds the specified time period, the test fails.

The time period includes execution of the test method itself, any repetitions of the test (see @Repeat), as well as any set up or tear down of the test fixture.

@Timed(millis=1000)
public void testProcessWithOneSecondTimeout() {
    // some logic that should not take longer than 1 second to execute
}

Spring’s @Timed annotation has different semantics than JUnit 4’s @Test(timeout=…​) support. Specifically, due to the manner in which JUnit 4 handles test execution timeouts (that is, by executing the test method in a separate Thread), @Test(timeout=…​) preemptively fails the test if the test takes too long. Spring’s @Timed, on the other hand, does not preemptively fail the test but rather waits for the test to complete before failing.

@Repeat

@Repeat indicates that the annotated test method must be executed repeatedly. The number of times that the test method is to be executed is specified in the annotation.

The scope of execution to be repeated includes execution of the test method itself as well as any set up or tear down of the test fixture.

@Repeat(10)
@Test
public void testProcessRepeatedly() {
    // ...
}

15.4.4 Meta-Annotation Support for Testing

It is possible to use most test-related annotations as meta-annotations in order to create custom composed annotations and reduce configuration duplication across a test suite.

Each of the following may be used as meta-annotations in conjunction with the TestContext framework.

  • @BootstrapWith
  • @ContextConfiguration
  • @ContextHierarchy
  • @ActiveProfiles
  • @TestPropertySource
  • @DirtiesContext
  • @WebAppConfiguration
  • @TestExecutionListeners
  • @Transactional
  • @BeforeTransaction
  • @AfterTransaction
  • @Commit
  • @Rollback
  • @Sql
  • @SqlConfig
  • @SqlGroup
  • @Repeat
  • @Timed
  • @IfProfileValue
  • @ProfileValueSourceConfiguration

For example, if we discover that we are repeating the following configuration across our JUnit 4 based test suite…​

@RunWith(SpringRunner.class)
@ContextConfiguration({"/app-config.xml", "/test-data-access-config.xml"})
@ActiveProfiles("dev")
@Transactional
public class OrderRepositoryTests { }

@RunWith(SpringRunner.class)
@ContextConfiguration({"/app-config.xml", "/test-data-access-config.xml"})
@ActiveProfiles("dev")
@Transactional
public class UserRepositoryTests { }

We can reduce the above duplication by introducing a custom composed annotation that centralizes the common test configuration like this:

@Target(ElementType.TYPE)
@Retention(RetentionPolicy.RUNTIME)
@ContextConfiguration({"/app-config.xml", "/test-data-access-config.xml"})
@ActiveProfiles("dev")
@Transactional
public @interface TransactionalDevTest { }

Then we can use our custom @TransactionalDevTest annotation to simplify the configuration of individual test classes as follows:

@RunWith(SpringRunner.class)
@TransactionalDevTest
public class OrderRepositoryTests { }

@RunWith(SpringRunner.class)
@TransactionalDevTest
public class UserRepositoryTests { }

For further details, consult the Spring Annotation Programming Model.

15.5 Spring TestContext Framework

The Spring TestContext Framework (located in the org.springframework.test.context package) provides generic, annotation-driven unit and integration testing support that is agnostic of the testing framework in use. The TestContext framework also places a great deal of importance on convention over configuration with reasonable defaults that can be overridden through annotation-based configuration.

In addition to generic testing infrastructure, the TestContext framework provides explicit support for JUnit 4 and TestNG in the form of abstract support classes. For JUnit 4, Spring also provides a custom JUnit Runner and custom JUnit Rules that allow one to write so-called POJO test classes. POJO test classes are not required to extend a particular class hierarchy.

The following section provides an overview of the internals of the TestContext framework. If you are only interested in using the framework and not necessarily interested in extending it with your own custom listeners or custom loaders, feel free to go directly to the configuration (context management, dependency injection, transaction management), support classes, and annotation support sections.

15.5.1 Key abstractions

The core of the framework consists of the TestContextManager class and the TestContext, TestExecutionListener, and SmartContextLoader interfaces. ATestContextManager is created per test class (e.g., for the execution of all test methods within a single test class in JUnit 4). The TestContextManager in turn manages a TestContext that holds the context of the current test. The TestContextManager also updates the state of the TestContext as the test progresses and delegates to TestExecutionListener implementations, which instrument the actual test execution by providing dependency injection, managing transactions, and so on. A SmartContextLoader is responsible for loading an ApplicationContext for a given test class. Consult the javadocs and the Spring test suite for further information and examples of various implementations.

TestContext

TestContext encapsulates the context in which a test is executed, agnostic of the actual testing framework in use, and provides context management and caching support for the test instance for which it is responsible. The TestContext also delegates to a SmartContextLoader to load an ApplicationContext if requested.

TestContextManager

TestContextManager is the main entry point into the Spring TestContext Framework, which manages a single TestContext and signals events to each registeredTestExecutionListener at well-defined test execution points:

  • prior to any before class or before all methods of a particular testing framework
  • test instance post-processing
  • prior to any before or before each methods of a particular testing framework
  • after any after or after each methods of a particular testing framework
  • after any after class or after all methods of a particular testing framework

TestExecutionListener

TestExecutionListener defines the API for reacting to test execution events published by the TestContextManager with which the listener is registered. SeeSection 15.5.3, “TestExecutionListener configuration”.

Context Loaders

ContextLoader is a strategy interface that was introduced in Spring 2.5 for loading an ApplicationContext for an integration test managed by the Spring TestContext Framework. Implement SmartContextLoader instead of this interface in order to provide support for annotated classes, active bean definition profiles, test property sources, context hierarchies, and WebApplicationContext support.

SmartContextLoader is an extension of the ContextLoader interface introduced in Spring 3.1. The SmartContextLoader SPI supersedes the ContextLoader SPI that was introduced in Spring 2.5. Specifically, a SmartContextLoader can choose to process resource locations, annotated classes, or context initializers. Furthermore, a SmartContextLoader can set active bean definition profiles and test property sources in the context that it loads.

Spring provides the following implementations:

  • DelegatingSmartContextLoader: one of two default loaders which delegates internally to an AnnotationConfigContextLoader, aGenericXmlContextLoader, or a GenericGroovyXmlContextLoader depending either on the configuration declared for the test class or on the presence of default locations or default configuration classes. Groovy support is only enabled if Groovy is on the classpath.
  • WebDelegatingSmartContextLoader: one of two default loaders which delegates internally to an AnnotationConfigWebContextLoader, aGenericXmlWebContextLoader, or a GenericGroovyXmlWebContextLoader depending either on the configuration declared for the test class or on the presence of default locations or default configuration classes. A web ContextLoader will only be used if @WebAppConfiguration is present on the test class. Groovy support is only enabled if Groovy is on the classpath.
  • AnnotationConfigContextLoader: loads a standard ApplicationContext from annotated classes.
  • AnnotationConfigWebContextLoader: loads a WebApplicationContext from annotated classes.
  • GenericGroovyXmlContextLoader: loads a standard ApplicationContext from resource locations that are either Groovy scripts or XML configuration files.
  • GenericGroovyXmlWebContextLoader: loads a WebApplicationContext from resource locations that are either Groovy scripts or XML configuration files.
  • GenericXmlContextLoader: loads a standard ApplicationContext from XML resource locations.
  • GenericXmlWebContextLoader: loads a WebApplicationContext from XML resource locations.
  • GenericPropertiesContextLoader: loads a standard ApplicationContext from Java Properties files.

15.5.2 Bootstrapping the TestContext framework

The default configuration for the internals of the Spring TestContext Framework is sufficient for all common use cases. However, there are times when a development team or third party framework would like to change the default ContextLoader, implement a custom TestContext or ContextCache, augment the default sets ofContextCustomizerFactory and TestExecutionListener implementations, etc. For such low level control over how the TestContext framework operates, Spring provides a bootstrapping strategy.

TestContextBootstrapper defines the SPI for bootstrapping the TestContext framework. A TestContextBootstrapper is used by the TestContextManager to load the TestExecutionListener implementations for the current test and to build the TestContext that it manages. A custom bootstrapping strategy can be configured for a test class (or test class hierarchy) via @BootstrapWith, either directly or as a meta-annotation. If a bootstrapper is not explicitly configured via@BootstrapWith, either the DefaultTestContextBootstrapper or the WebTestContextBootstrapper will be used, depending on the presence of@WebAppConfiguration.

Since the TestContextBootstrapper SPI is likely to change in the future in order to accommodate new requirements, implementers are strongly encouraged not to implement this interface directly but rather to extend AbstractTestContextBootstrapper or one of its concrete subclasses instead.

15.5.3 TestExecutionListener configuration

Spring provides the following TestExecutionListener implementations that are registered by default, exactly in this order.

  • ServletTestExecutionListener: configures Servlet API mocks for a WebApplicationContext
  • DirtiesContextBeforeModesTestExecutionListener: handles the @DirtiesContext annotation for before modes
  • DependencyInjectionTestExecutionListener: provides dependency injection for the test instance
  • DirtiesContextTestExecutionListener: handles the @DirtiesContext annotation for after modes
  • TransactionalTestExecutionListener: provides transactional test execution with default rollback semantics
  • SqlScriptsTestExecutionListener: executes SQL scripts configured via the @Sql annotation

Registering custom TestExecutionListeners

Custom TestExecutionListeners can be registered for a test class and its subclasses via the @TestExecutionListeners annotation. See annotation support and the javadocs for @TestExecutionListeners for details and examples.

Automatic discovery of default TestExecutionListeners

Registering custom TestExecutionListeners via @TestExecutionListeners is suitable for custom listeners that are used in limited testing scenarios; however, it can become cumbersome if a custom listener needs to be used across a test suite. Since Spring Framework 4.1, this issue is addressed via support for automatic discovery of default TestExecutionListener implementations via the SpringFactoriesLoader mechanism.

Specifically, the spring-test module declares all core default TestExecutionListeners under theorg.springframework.test.context.TestExecutionListener key in its META-INF/spring.factories properties file. Third-party frameworks and developers can contribute their own TestExecutionListeners to the list of default listeners in the same manner via their own META-INF/spring.factories properties file.

Ordering TestExecutionListeners

When the TestContext framework discovers default TestExecutionListeners via the aforementioned SpringFactoriesLoader mechanism, the instantiated listeners are sorted using Spring’s AnnotationAwareOrderComparator which honors Spring’s Ordered interface and @Order annotation for ordering.AbstractTestExecutionListener and all default TestExecutionListeners provided by Spring implement Ordered with appropriate values. Third-party frameworks and developers should therefore make sure that their default TestExecutionListeners are registered in the proper order by implementing Ordered or declaring @Order. Consult the javadocs for the getOrder() methods of the core default TestExecutionListeners for details on what values are assigned to each core listener.

Merging TestExecutionListeners

If a custom TestExecutionListener is registered via @TestExecutionListeners, the default listeners will not be registered. In most common testing scenarios, this effectively forces the developer to manually declare all default listeners in addition to any custom listeners. The following listing demonstrates this style of configuration.

@ContextConfiguration
@TestExecutionListeners({
    MyCustomTestExecutionListener.class,
    ServletTestExecutionListener.class,
    DirtiesContextBeforeModesTestExecutionListener.class,
    DependencyInjectionTestExecutionListener.class,
    DirtiesContextTestExecutionListener.class,
    TransactionalTestExecutionListener.class,
    SqlScriptsTestExecutionListener.class
})
public class MyTest {
    // class body...
}

The challenge with this approach is that it requires that the developer know exactly which listeners are registered by default. Moreover, the set of default listeners can change from release to release — for example, SqlScriptsTestExecutionListener was introduced in Spring Framework 4.1, andDirtiesContextBeforeModesTestExecutionListener was introduced in Spring Framework 4.2. Furthermore, third-party frameworks like Spring Security register their own default TestExecutionListeners via the aforementioned automatic discovery mechanism.

To avoid having to be aware of and re-declare all default listeners, the mergeMode attribute of @TestExecutionListeners can be set toMergeMode.MERGE_WITH_DEFAULTS. MERGE_WITH_DEFAULTS indicates that locally declared listeners should be merged with the default listeners. The merging algorithm ensures that duplicates are removed from the list and that the resulting set of merged listeners is sorted according to the semantics ofAnnotationAwareOrderComparator as described in the section called “Ordering TestExecutionListeners”. If a listener implements Ordered or is annotated with@Order it can influence the position in which it is merged with the defaults; otherwise, locally declared listeners will simply be appended to the list of default listeners when merged.

For example, if the MyCustomTestExecutionListener class in the previous example configures its order value (for example, 500) to be less than the order of theServletTestExecutionListener (which happens to be 1000), the MyCustomTestExecutionListener can then be automatically merged with the list of defaults in front of the ServletTestExecutionListener, and the previous example could be replaced with the following.

@ContextConfiguration
@TestExecutionListeners(
    listeners = MyCustomTestExecutionListener.class,
    mergeMode = MERGE_WITH_DEFAULTS
)
public class MyTest {
    // class body...
}

15.5.4 Context management

Each TestContext provides context management and caching support for the test instance it is responsible for. Test instances do not automatically receive access to the configured ApplicationContext. However, if a test class implements the ApplicationContextAware interface, a reference to the ApplicationContext is supplied to the test instance. Note that AbstractJUnit4SpringContextTests and AbstractTestNGSpringContextTests implement ApplicationContextAwareand therefore provide access to the ApplicationContext automatically.

Spring框架参考文档 Spring Framework Reference Documentation

As an alternative to implementing the ApplicationContextAware interface, you can inject the application context for your test class through the@Autowired annotation on either a field or setter method. For example:

@RunWith(SpringRunner.class)
@ContextConfiguration
public class MyTest {

    @Autowired
    private ApplicationContext applicationContext;

    // class body...
}

Similarly, if your test is configured to load a WebApplicationContext, you can inject the web application context into your test as follows:

@RunWith(SpringRunner.class)
@WebAppConfiguration
@ContextConfiguration
public class MyWebAppTest {
    @Autowired
    private WebApplicationContext wac;

    // class body...
}

Dependency injection via @Autowired is provided by the DependencyInjectionTestExecutionListener which is configured by default (seeSection 15.5.5, “Dependency injection of test fixtures”).

Test classes that use the TestContext framework do not need to extend any particular class or implement a specific interface to configure their application context. Instead, configuration is achieved simply by declaring the @ContextConfiguration annotation at the class level. If your test class does not explicitly declare application context resource locations or annotated classes, the configured ContextLoader determines how to load a context from a default location or default configuration classes. In addition to context resource locations and annotated classes, an application context can also be configured via application context initializers.

The following sections explain how to configure an ApplicationContext via XML configuration files, Groovy scripts, annotated classes (typically @Configurationclasses), or context initializers using Spring’s @ContextConfiguration annotation. Alternatively, you can implement and configure your own customSmartContextLoader for advanced use cases.

Context configuration with XML resources

To load an ApplicationContext for your tests using XML configuration files, annotate your test class with @ContextConfiguration and configure the locationsattribute with an array that contains the resource locations of XML configuration metadata. A plain or relative path — for example "context.xml" — will be treated as a classpath resource that is relative to the package in which the test class is defined. A path starting with a slash is treated as an absolute classpath location, for example"/org/example/config.xml". A path which represents a resource URL (i.e., a path prefixed with classpath:, file:, http:, etc.) will be used as is.

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from "/app-config.xml" and
// "/test-config.xml" in the root of the classpath
@ContextConfiguration(locations={"/app-config.xml", "/test-config.xml"})
public class MyTest {
    // class body...
}

@ContextConfiguration supports an alias for the locations attribute through the standard Java value attribute. Thus, if you do not need to declare additional attributes in @ContextConfiguration, you can omit the declaration of the locations attribute name and declare the resource locations by using the shorthand format demonstrated in the following example.

@RunWith(SpringRunner.class)
@ContextConfiguration({"/app-config.xml", "/test-config.xml"})
public class MyTest {
    // class body...
}

If you omit both the locations and value attributes from the @ContextConfiguration annotation, the TestContext framework will attempt to detect a default XML resource location. Specifically, GenericXmlContextLoader and GenericXmlWebContextLoader detect a default location based on the name of the test class. If your class is named com.example.MyTest, GenericXmlContextLoader loads your application context from "classpath:com/example/MyTest-context.xml".

package com.example;

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from
// "classpath:com/example/MyTest-context.xml"
@ContextConfiguration
public class MyTest {
    // class body...
}

Context configuration with Groovy scripts

To load an ApplicationContext for your tests using Groovy scripts that utilize the Groovy Bean Definition DSL, annotate your test class with@ContextConfiguration and configure the locations or value attribute with an array that contains the resource locations of Groovy scripts. Resource lookup semantics for Groovy scripts are the same as those described for XML configuration files.

Spring框架参考文档 Spring Framework Reference Documentation

Support for using Groovy scripts to load an ApplicationContext in the Spring TestContext Framework is enabled automatically if Groovy is on the classpath.

 
@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from "/AppConfig.groovy" and
// "/TestConfig.groovy" in the root of the classpath
@ContextConfiguration({"/AppConfig.groovy", "/TestConfig.Groovy"})
public class MyTest {
    // class body...
}

If you omit both the locations and value attributes from the @ContextConfiguration annotation, the TestContext framework will attempt to detect a default Groovy script. Specifically, GenericGroovyXmlContextLoader and GenericGroovyXmlWebContextLoader detect a default location based on the name of the test class. If your class is named com.example.MyTest, the Groovy context loader will load your application context from "classpath:com/example/MyTestContext.groovy".

package com.example;

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from
// "classpath:com/example/MyTestContext.groovy"
@ContextConfiguration
public class MyTest {
    // class body...
}
Spring框架参考文档 Spring Framework Reference Documentation

Both XML configuration files and Groovy scripts can be declared simultaneously via the locations or value attribute of @ContextConfiguration. If the path to a configured resource location ends with .xml it will be loaded using an XmlBeanDefinitionReader; otherwise it will be loaded using aGroovyBeanDefinitionReader.

The following listing demonstrates how to combine both in an integration test.

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from
// "/app-config.xml" and "/TestConfig.groovy"
@ContextConfiguration({ "/app-config.xml", "/TestConfig.groovy" })
public class MyTest {
    // class body...
}

Context configuration with annotated classes

To load an ApplicationContext for your tests using annotated classes (see Section 7.12, “Java-based container configuration”), annotate your test class with@ContextConfiguration and configure the classes attribute with an array that contains references to annotated classes.

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from AppConfig and TestConfig
@ContextConfiguration(classes = {AppConfig.class, TestConfig.class})
public class MyTest {
    // class body...
}
Spring框架参考文档 Spring Framework Reference Documentation

The term annotated class can refer to any of the following.

  • A class annotated with @Configuration
  • A component (i.e., a class annotated with @Component, @Service, @Repository, etc.)
  • A JSR-330 compliant class that is annotated with javax.inject annotations
  • Any other class that contains @Bean-methods

Consult the javadocs of @Configuration and @Bean for further information regarding the configuration and semantics of annotated classes, paying special attention to the discussion of `@Bean` Lite Mode.

If you omit the classes attribute from the @ContextConfiguration annotation, the TestContext framework will attempt to detect the presence of default configuration classes. Specifically, AnnotationConfigContextLoader and AnnotationConfigWebContextLoader will detect all static nested classes of the test class that meet the requirements for configuration class implementations as specified in the @Configuration javadocs. In the following example, the OrderServiceTest class declares a static nested configuration class named Config that will be automatically used to load the ApplicationContext for the test class. Note that the name of the configuration class is arbitrary. In addition, a test class can contain more than one static nested configuration class if desired.

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from the
// static nested Config class
@ContextConfiguration
public class OrderServiceTest {

    @Configuration
    static class Config {

        // this bean will be injected into the OrderServiceTest class
        @Bean
        public OrderService orderService() {
            OrderService orderService = new OrderServiceImpl();
            // set properties, etc.
            return orderService;
        }
    }

    @Autowired
    private OrderService orderService;

    @Test
    public void testOrderService() {
        // test the orderService
    }

}

Mixing XML, Groovy scripts, and annotated classes

It may sometimes be desirable to mix XML configuration files, Groovy scripts, and annotated classes (i.e., typically @Configuration classes) to configure anApplicationContext for your tests. For example, if you use XML configuration in production, you may decide that you want to use @Configuration classes to configure specific Spring-managed components for your tests, or vice versa.

Furthermore, some third-party frameworks (like Spring Boot) provide first-class support for loading an ApplicationContext from different types of resources simultaneously (e.g., XML configuration files, Groovy scripts, and @Configuration classes). The Spring Framework historically has not supported this for standard deployments. Consequently, most of the SmartContextLoader implementations that the Spring Framework delivers in the spring-test module support only one resource type per test context; however, this does not mean that you cannot use both. One exception to the general rule is that the GenericGroovyXmlContextLoaderand GenericGroovyXmlWebContextLoader support both XML configuration files and Groovy scripts simultaneously. Furthermore, third-party frameworks may choose to support the declaration of both locations and classes via @ContextConfiguration, and with the standard testing support in the TestContext framework, you have the following options.

If you want to use resource locations (e.g., XML or Groovy) and @Configuration classes to configure your tests, you will have to pick one as the entry point, and that one will have to include or import the other. For example, in XML or Groovy scripts you can include @Configuration classes via component scanning or define them as normal Spring beans; whereas, in a @Configuration class you can use @ImportResource to import XML configuration files or Groovy scripts. Note that this behavior is semantically equivalent to how you configure your application in production: in production configuration you will define either a set of XML or Groovy resource locations or a set of @Configuration classes that your production ApplicationContext will be loaded from, but you still have the freedom to include or import the other type of configuration.

Context configuration with context initializers

To configure an ApplicationContext for your tests using context initializers, annotate your test class with @ContextConfiguration and configure theinitializers attribute with an array that contains references to classes that implement ApplicationContextInitializer. The declared context initializers will then be used to initialize the ConfigurableApplicationContext that is loaded for your tests. Note that the concrete ConfigurableApplicationContext type supported by each declared initializer must be compatible with the type of ApplicationContext created by the SmartContextLoader in use (i.e., typically aGenericApplicationContext). Furthermore, the order in which the initializers are invoked depends on whether they implement Spring’s Ordered interface or are annotated with Spring’s @Order annotation or the standard @Priority annotation.

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from TestConfig
// and initialized by TestAppCtxInitializer
@ContextConfiguration(
    classes = TestConfig.class,
    initializers = TestAppCtxInitializer.class)
public class MyTest {
    // class body...
}

It is also possible to omit the declaration of XML configuration files, Groovy scripts, or annotated classes in @ContextConfiguration entirely and instead declare onlyApplicationContextInitializer classes which are then responsible for registering beans in the context — for example, by programmatically loading bean definitions from XML files or configuration classes.

@RunWith(SpringRunner.class)
// ApplicationContext will be initialized by EntireAppInitializer
// which presumably registers beans in the context
@ContextConfiguration(initializers = EntireAppInitializer.class)
public class MyTest {
    // class body...
}

Context configuration inheritance

@ContextConfiguration supports boolean inheritLocations and inheritInitializers attributes that denote whether resource locations or annotated classes and context initializers declared by superclasses should be inherited. The default value for both flags is true. This means that a test class inherits the resource locations or annotated classes as well as the context initializers declared by any superclasses. Specifically, the resource locations or annotated classes for a test class are appended to the list of resource locations or annotated classes declared by superclasses. Similarly, the initializers for a given test class will be added to the set of initializers defined by test superclasses. Thus, subclasses have the option of extending the resource locations, annotated classes, or context initializers.

If the inheritLocations or inheritInitializers attribute in @ContextConfiguration is set to false, the resource locations or annotated classes and the context initializers, respectively, for the test class shadow and effectively replace the configuration defined by superclasses.

In the following example that uses XML resource locations, the ApplicationContext for ExtendedTest will be loaded from "base-config.xml" and "extended-config.xml", in that order. Beans defined in "extended-config.xml" may therefore override (i.e., replace) those defined in "base-config.xml".

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from "/base-config.xml"
// in the root of the classpath
@ContextConfiguration("/base-config.xml")
public class BaseTest {
    // class body...
}

// ApplicationContext will be loaded from "/base-config.xml" and
// "/extended-config.xml" in the root of the classpath
@ContextConfiguration("/extended-config.xml")
public class ExtendedTest extends BaseTest {
    // class body...
}

Similarly, in the following example that uses annotated classes, the ApplicationContext for ExtendedTest will be loaded from the BaseConfig andExtendedConfig classes, in that order. Beans defined in ExtendedConfig may therefore override (i.e., replace) those defined in BaseConfig.

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from BaseConfig
@ContextConfiguration(classes = BaseConfig.class)
public class BaseTest {
    // class body...
}

// ApplicationContext will be loaded from BaseConfig and ExtendedConfig
@ContextConfiguration(classes = ExtendedConfig.class)
public class ExtendedTest extends BaseTest {
    // class body...
}

In the following example that uses context initializers, the ApplicationContext for ExtendedTest will be initialized using BaseInitializer andExtendedInitializer. Note, however, that the order in which the initializers are invoked depends on whether they implement Spring’s Ordered interface or are annotated with Spring’s @Order annotation or the standard @Priority annotation.

@RunWith(SpringRunner.class)
// ApplicationContext will be initialized by BaseInitializer
@ContextConfiguration(initializers = BaseInitializer.class)
public class BaseTest {
    // class body...
}

// ApplicationContext will be initialized by BaseInitializer
// and ExtendedInitializer
@ContextConfiguration(initializers = ExtendedInitializer.class)
public class ExtendedTest extends BaseTest {
    // class body...
}

Context configuration with environment profiles

Spring 3.1 introduced first-class support in the framework for the notion of environments and profiles (a.k.a., bean definition profiles), and integration tests can be configured to activate particular bean definition profiles for various testing scenarios. This is achieved by annotating a test class with the @ActiveProfiles annotation and supplying a list of profiles that should be activated when loading the ApplicationContext for the test.

Spring框架参考文档 Spring Framework Reference Documentation

@ActiveProfiles may be used with any implementation of the new SmartContextLoader SPI, but @ActiveProfiles is not supported with implementations of the older ContextLoader SPI.

Let’s take a look at some examples with XML configuration and @Configuration classes.

<!-- app-config.xml -->
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xmlns:jdbc="http://www.springframework.org/schema/jdbc"
    xmlns:jee="http://www.springframework.org/schema/jee"
    xsi:schemaLocation="...">

    <bean id="transferService"
            class="com.bank.service.internal.DefaultTransferService">
        <constructor-arg ref="accountRepository"/>
        <constructor-arg ref="feePolicy"/>
    </bean>

    <bean id="accountRepository"
            class="com.bank.repository.internal.JdbcAccountRepository">
        <constructor-arg ref="dataSource"/>
    </bean>

    <bean id="feePolicy"
        class="com.bank.service.internal.ZeroFeePolicy"/>

    <beans profile="dev">
        <jdbc:embedded-database id="dataSource">
            <jdbc:script
                location="classpath:com/bank/config/sql/schema.sql"/>
            <jdbc:script
                location="classpath:com/bank/config/sql/test-data.sql"/>
        </jdbc:embedded-database>
    </beans>

    <beans profile="production">
        <jee:jndi-lookup id="dataSource" jndi-name="java:comp/env/jdbc/datasource"/>
    </beans>

    <beans profile="default">
        <jdbc:embedded-database id="dataSource">
            <jdbc:script
                location="classpath:com/bank/config/sql/schema.sql"/>
        </jdbc:embedded-database>
    </beans>

</beans>
package com.bank.service;

@RunWith(SpringRunner.class)
// ApplicationContext will be loaded from "classpath:/app-config.xml"
@ContextConfiguration("/app-config.xml")
@ActiveProfiles("dev")
public class TransferServiceTest {

    @Autowired
    private TransferService transferService;

    @Test
    public void testTransferService() {
        // test the transferService
    }
}

When TransferServiceTest is run, its ApplicationContext will be loaded from the app-config.xml configuration file in the root of the classpath. If you inspectapp-config.xml you’ll notice that the accountRepository bean has a dependency on a dataSource bean; however, dataSource is not defined as a top-level bean. Instead, dataSource is defined three times: in the production profile, the dev profile, and the default profile.

By annotating TransferServiceTest with @ActiveProfiles("dev") we instruct the Spring TestContext Framework to load the ApplicationContext with the active profiles set to {"dev"}. As a result, an embedded database will be created and populated with test data, and the accountRepository bean will be wired with a reference to the development DataSource. And that’s likely what we want in an integration test.

It is sometimes useful to assign beans to a default profile. Beans within the default profile are only included when no other profile is specifically activated. This can be used to define fallback beans to be used in the application’s default state. For example, you may explicitly provide a data source for dev and production profiles, but define an in-memory data source as a default when neither of these is active.

The following code listings demonstrate how to implement the same configuration and integration test but using @Configuration classes instead of XML.

@Configuration
@Profile("dev")
public class StandaloneDataConfig {

    @Bean
    public DataSource dataSource() {
        return new EmbeddedDatabaseBuilder()
            .setType(EmbeddedDatabaseType.HSQL)
            .addScript("classpath:com/bank/config/sql/schema.sql")
            .addScript("classpath:com/bank/config/sql/test-data.sql")
            .build();
    }
}
@Configuration
@Profile("production")
public class JndiDataConfig {

    @Bean(destroyMethod="")
    public DataSource dataSource() throws Exception {
        Context ctx = new InitialContext();
        return (DataSource) ctx.lookup("java:comp/env/jdbc/datasource");
    }
}
@Configuration
@Profile("default")
public class DefaultDataConfig {

    @Bean
    public DataSource dataSource() {
        return new EmbeddedDatabaseBuilder()
            .setType(EmbeddedDatabaseType.HSQL)
            .addScript("classpath:com/bank/config/sql/schema.sql")
            .build();
    }
}
@Configuration
public class TransferServiceConfig {

    @Autowired DataSource dataSource;

    @Bean
    public TransferService transferService() {
        return new DefaultTransferService(accountRepository(), feePolicy());
    }

    @Bean
    public AccountRepository accountRepository() {
        return new JdbcAccountRepository(dataSource);
    }

    @Bean
    public FeePolicy feePolicy() {
        return new ZeroFeePolicy();
    }

}
package com.bank.service;

@RunWith(SpringRunner.class)
@ContextConfiguration(classes = {
        TransferServiceConfig.class,
        StandaloneDataConfig.class,
        JndiDataConfig.class,
        DefaultDataConfig.class})
@ActiveProfiles("dev")
public class TransferServiceTest {

    @Autowired
    private TransferService transferService;

    @Test
    public void testTransferService() {
        // test the transferService
    }
}

In this variation, we have split the XML configuration into four independent @Configuration classes:

  • TransferServiceConfig: acquires a dataSource via dependency injection using @Autowired
  • StandaloneDataConfig: defines a dataSource for an embedded database suitable for developer tests
  • JndiDataConfig: defines a dataSource that is retrieved from JNDI in a production environment
  • DefaultDataConfig: defines a dataSource for a default embedded database in case no profile is active

As with the XML-based configuration example, we still annotate TransferServiceTest with @ActiveProfiles("dev"), but this time we specify all four configuration classes via the @ContextConfiguration annotation. The body of the test class itself remains completely unchanged.

It is often the case that a single set of profiles is used across multiple test classes within a given project. Thus, to avoid duplicate declarations of the @ActiveProfilesannotation it is possible to declare @ActiveProfiles once on a base class, and subclasses will automatically inherit the @ActiveProfiles configuration from the base class. In the following example, the declaration of @ActiveProfiles (as well as other annotations) has been moved to an abstract superclass,AbstractIntegrationTest.

package com.bank.service;

@RunWith(SpringRunner.class)
@ContextConfiguration(classes = {
        TransferServiceConfig.class,
        StandaloneDataConfig.class,
        JndiDataConfig.class,
        DefaultDataConfig.class})
@ActiveProfiles("dev")
public abstract class AbstractIntegrationTest {
}
package com.bank.service;

// "dev" profile inherited from superclass
public class TransferServiceTest extends AbstractIntegrationTest {

    @Autowired
    private TransferService transferService;

    @Test
    public void testTransferService() {
        // test the transferService
    }
}

@ActiveProfiles also supports an inheritProfiles attribute that can be used to disable the inheritance of active profiles.

package com.bank.service;

// "dev" profile overridden with "production"
@ActiveProfiles(profiles = "production", inheritProfiles = false)
public class ProductionTransferServiceTest extends AbstractIntegrationTest {
    // test body
}

Furthermore, it is sometimes necessary to resolve active profiles for tests programmatically instead of declaratively — for example, based on:

  • the current operating system
  • whether tests are being executed on a continuous integration build server
  • the presence of certain environment variables
  • the presence of custom class-level annotations
  • etc.

To resolve active bean definition profiles programmatically, simply implement a custom ActiveProfilesResolver and register it via the resolver attribute of@ActiveProfiles. The following example demonstrates how to implement and register a custom OperatingSystemActiveProfilesResolver. For further information, refer to the corresponding javadocs.

package com.bank.service;

// "dev" profile overridden programmatically via a custom resolver
@ActiveProfiles(
    resolver = OperatingSystemActiveProfilesResolver.class,
    inheritProfiles = false)
public class TransferServiceTest extends AbstractIntegrationTest {
    // test body
}
package com.bank.service.test;

public class OperatingSystemActiveProfilesResolver implements ActiveProfilesResolver {

    @Override
    String[] resolve(Class<?> testClass) {
        String profile = ...;
        // determine the value of profile based on the operating system
        return new String[] {profile};
    }
}

Context configuration with test property sources

Spring 3.1 introduced first-class support in the framework for the notion of an environment with a hierarchy of property sources, and since Spring 4.1 integration tests can be configured with test-specific property sources. In contrast to the @PropertySource annotation used on @Configuration classes, the @TestPropertySourceannotation can be declared on a test class to declare resource locations for test properties files or inlined properties. These test property sources will be added to the set of PropertySources in the Environment for the ApplicationContext loaded for the annotated integration test.

Spring框架参考文档 Spring Framework Reference Documentation

@TestPropertySource may be used with any implementation of the SmartContextLoader SPI, but @TestPropertySource is not supported with implementations of the older ContextLoader SPI.

Implementations of SmartContextLoader gain access to merged test property source values via the getPropertySourceLocations() andgetPropertySourceProperties() methods in MergedContextConfiguration.

Declaring test property sources

Test properties files can be configured via the locations or value attribute of @TestPropertySource as shown in the following example.

Both traditional and XML-based properties file formats are supported — for example, "classpath:/com/example/test.properties" or"file:///path/to/file.xml".

Each path will be interpreted as a Spring Resource. A plain path — for example, "test.properties" — will be treated as a classpath resource that is relative to the package in which the test class is defined. A path starting with a slash will be treated as an absolute classpath resource, for example: "/org/example/test.xml". A path which references a URL (e.g., a path prefixed with classpath:, file:, http:, etc.) will be loaded using the specified resource protocol. Resource location wildcards (e.g. */.properties) are not permitted: each location must evaluate to exactly one .properties or .xml resource.

@ContextConfiguration
@TestPropertySource("/test.properties")
public class MyIntegrationTests {
    // class body...
}

Inlined properties in the form of key-value pairs can be configured via the properties attribute of @TestPropertySource as shown in the following example. All key-value pairs will be added to the enclosing Environment as a single test PropertySource with the highest precedence.

The supported syntax for key-value pairs is the same as the syntax defined for entries in a Java properties file:

  • "key=value"
  • "key:value"
  • "key value"
@ContextConfiguration
@TestPropertySource(properties = {"timezone = GMT", "port: 4242"})
public class MyIntegrationTests {
    // class body...
}

Default properties file detection

If @TestPropertySource is declared as an empty annotation (i.e., without explicit values for the locations or properties attributes), an attempt will be made to detect a default properties file relative to the class that declared the annotation. For example, if the annotated test class is com.example.MyTest, the corresponding default properties file is "classpath:com/example/MyTest.properties". If the default cannot be detected, an IllegalStateException will be thrown.

Precedence

Test property sources have higher precedence than those loaded from the operating system’s environment or Java system properties as well as property sources added by the application declaratively via @PropertySource or programmatically. Thus, test property sources can be used to selectively override properties defined in system and application property sources. Furthermore, inlined properties have higher precedence than properties loaded from resource locations.

In the following example, the timezone and port properties as well as any properties defined in "/test.properties" will override any properties of the same name that are defined in system and application property sources. Furthermore, if the "/test.properties" file defines entries for the timezone and port properties those will be overridden by the inlined properties declared via the properties attribute.

@ContextConfiguration
@TestPropertySource(
    locations = "/test.properties",
    properties = {"timezone = GMT", "port: 4242"}
)
public class MyIntegrationTests {
    // class body...
}

Inheriting and overriding test property sources

@TestPropertySource supports boolean inheritLocations and inheritProperties attributes that denote whether resource locations for properties files and inlined properties declared by superclasses should be inherited. The default value for both flags is true. This means that a test class inherits the locations and inlined properties declared by any superclasses. Specifically, the locations and inlined properties for a test class are appended to the locations and inlined properties declared by superclasses. Thus, subclasses have the option of extending the locations and inlined properties. Note that properties that appear later will shadow (i.e.., override) properties of the same name that appear earlier. In addition, the aforementioned precedence rules apply for inherited test property sources as well.

If the inheritLocations or inheritProperties attribute in @TestPropertySource is set to false, the locations or inlined properties, respectively, for the test class shadow and effectively replace the configuration defined by superclasses.

In the following example, the ApplicationContext for BaseTest will be loaded using only the "base.properties" file as a test property source. In contrast, theApplicationContext for ExtendedTest will be loaded using the "base.properties" and "extended.properties" files as test property source locations.

@TestPropertySource("base.properties")
@ContextConfiguration
public class BaseTest {
    // ...
}

@TestPropertySource("extended.properties")
@ContextConfiguration
public class ExtendedTest extends BaseTest {
    // ...
}

In the following example, the ApplicationContext for BaseTest will be loaded using only the inlined key1 property. In contrast, the ApplicationContext forExtendedTest will be loaded using the inlined key1 and key2 properties.

@TestPropertySource(properties = "key1 = value1")
@ContextConfiguration
public class BaseTest {
    // ...
}

@TestPropertySource(properties = "key2 = value2")
@ContextConfiguration
public class ExtendedTest extends BaseTest {
    // ...
}

Loading a WebApplicationContext

Spring 3.2 introduced support for loading a WebApplicationContext in integration tests. To instruct the TestContext framework to load a WebApplicationContextinstead of a standard ApplicationContext, simply annotate the respective test class with @WebAppConfiguration.

The presence of @WebAppConfiguration on your test class instructs the TestContext framework (TCF) that a WebApplicationContext (WAC) should be loaded for your integration tests. In the background the TCF makes sure that a MockServletContext is created and supplied to your test’s WAC. By default the base resource path for your MockServletContext will be set to "src/main/webapp". This is interpreted as a path relative to the root of your JVM (i.e., normally the path to your project). If you’re familiar with the directory structure of a web application in a Maven project, you’ll know that "src/main/webapp" is the default location for the root of your WAR. If you need to override this default, simply provide an alternate path to the @WebAppConfiguration annotation (e.g., @WebAppConfiguration("src/test/webapp")). If you wish to reference a base resource path from the classpath instead of the file system, just use Spring’s classpath: prefix.

Please note that Spring’s testing support for WebApplicationContexts is on par with its support for standard ApplicationContexts. When testing with aWebApplicationContext you are free to declare XML configuration files, Groovy scripts, or @Configuration classes via @ContextConfiguration. You are of course also free to use any other test annotations such as @ActiveProfiles, @TestExecutionListeners, @Sql, @Rollback, etc.

The following examples demonstrate some of the various configuration options for loading a WebApplicationContext.

Conventions. 

@RunWith(SpringRunner.class)

// defaults to "file:src/main/webapp"
@WebAppConfiguration

// detects "WacTests-context.xml" in same package
// or static nested @Configuration class
@ContextConfiguration

public class WacTests {
    //...
}

 

The above example demonstrates the TestContext framework’s support for convention over configuration. If you annotate a test class with @WebAppConfigurationwithout specifying a resource base path, the resource path will effectively default to "file:src/main/webapp". Similarly, if you declare @ContextConfiguration without specifying resource locations, annotated classes, or context initializers, Spring will attempt to detect the presence of your configuration using conventions (i.e., "WacTests-context.xml" in the same package as the WacTests class or static nested @Configuration classes).

Default resource semantics. 

@RunWith(SpringRunner.class)

// file system resource
@WebAppConfiguration("webapp")

// classpath resource
@ContextConfiguration("/spring/test-servlet-config.xml")

public class WacTests {
    //...
}

 

This example demonstrates how to explicitly declare a resource base path with @WebAppConfiguration and an XML resource location with @ContextConfiguration. The important thing to note here is the different semantics for paths with these two annotations. By default, @WebAppConfiguration resource paths are file system based; whereas, @ContextConfiguration resource locations are classpath based.

Explicit resource semantics. 

@RunWith(SpringRunner.class)

// classpath resource
@WebAppConfiguration("classpath:test-web-resources")

// file system resource
@ContextConfiguration("file:src/main/webapp/WEB-INF/servlet-config.xml")

public class WacTests {
    //...
}

 

In this third example, we see that we can override the default resource semantics for both annotations by specifying a Spring resource prefix. Contrast the comments in this example with the previous example.

To provide comprehensive web testing support, Spring 3.2 introduced a ServletTestExecutionListener that is enabled by default. When testing against aWebApplicationContext this TestExecutionListener sets up default thread-local state via Spring Web’s RequestContextHolder before each test method and creates a MockHttpServletRequest, MockHttpServletResponse, and ServletWebRequest based on the base resource path configured via @WebAppConfiguration.ServletTestExecutionListener also ensures that the MockHttpServletResponse and ServletWebRequest can be injected into the test instance, and once the test is complete it cleans up thread-local state.

Once you have a WebApplicationContext loaded for your test you might find that you need to interact with the web mocks — for example, to set up your test fixture or to perform assertions after invoking your web component. The following example demonstrates which mocks can be autowired into your test instance. Note that theWebApplicationContext and MockServletContext are both cached across the test suite; whereas, the other mocks are managed per test method by theServletTestExecutionListener.

Injecting mocks. 

@WebAppConfiguration
@ContextConfiguration
public class WacTests {

    @Autowired
    WebApplicationContext wac; // cached

    @Autowired
    MockServletContext servletContext; // cached

    @Autowired
    MockHttpSession session;

    @Autowired
    MockHttpServletRequest request;

    @Autowired
    MockHttpServletResponse response;

    @Autowired
    ServletWebRequest webRequest;

    //...
}

 

Context caching

Once the TestContext framework loads an ApplicationContext (or WebApplicationContext) for a test, that context will be cached and reused for all subsequent tests that declare the same unique context configuration within the same test suite. To understand how caching works, it is important to understand what is meant byunique and test suite.

An ApplicationContext can be uniquely identified by the combination of configuration parameters that are used to load it. Consequently, the unique combination of configuration parameters are used to generate a key under which the context is cached. The TestContext framework uses the following configuration parameters to build the context cache key:

  • locations (from @ContextConfiguration)
  • classes (from @ContextConfiguration)
  • contextInitializerClasses (from @ContextConfiguration)
  • contextCustomizers (from ContextCustomizerFactory)
  • contextLoader (from @ContextConfiguration)
  • parent (from @ContextHierarchy)
  • activeProfiles (from @ActiveProfiles)
  • propertySourceLocations (from @TestPropertySource)
  • propertySourceProperties (from @TestPropertySource)
  • resourceBasePath (from @WebAppConfiguration)

For example, if TestClassA specifies {"app-config.xml", "test-config.xml"} for the locations (or value) attribute of @ContextConfiguration, the TestContext framework will load the corresponding ApplicationContext and store it in a static context cache under a key that is based solely on those locations. So if TestClassB also defines {"app-config.xml", "test-config.xml"} for its locations (either explicitly or implicitly through inheritance) but does not define@WebAppConfiguration, a different ContextLoader, different active profiles, different context initializers, different test property sources, or a different parent context, then the same ApplicationContext will be shared by both test classes. This means that the setup cost for loading an application context is incurred only once (per test suite), and subsequent test execution is much faster.

Spring框架参考文档 Spring Framework Reference Documentation

The Spring TestContext framework stores application contexts in a static cache. This means that the context is literally stored in a static variable. In other words, if tests execute in separate processes the static cache will be cleared between each test execution, and this will effectively disable the caching mechanism.

To benefit from the caching mechanism, all tests must run within the same process or test suite. This can be achieved by executing all tests as a group within an IDE. Similarly, when executing tests with a build framework such as Ant, Maven, or Gradle it is important to make sure that the build framework does not fork between tests. For example, if the forkMode for the Maven Surefire plug-in is set to always or pertest, the TestContext framework will not be able to cache application contexts between test classes and the build process will run significantly slower as a result.

Since Spring Framework 4.3, the size of the context cache is bounded with a default maximum size of 32. Whenever the maximum size is reached, a least recently used(LRU) eviction policy is used to evict and close stale contexts. The maximum size can be configured from the command line or a build script by setting a JVM system property named spring.test.context.cache.maxSize. As an alternative, the same property can be set programmatically via the SpringProperties API.

Since having a large number of application contexts loaded within a given test suite can cause the suite to take an unnecessarily long time to execute, it is often beneficial to know exactly how many contexts have been loaded and cached. To view the statistics for the underlying context cache, simply set the log level for theorg.springframework.test.context.cache logging category to DEBUG.

In the unlikely case that a test corrupts the application context and requires reloading — for example, by modifying a bean definition or the state of an application object — you can annotate your test class or test method with @DirtiesContext (see the discussion of @DirtiesContext in Section 15.4.1, “Spring Testing Annotations”). This instructs Spring to remove the context from the cache and rebuild the application context before executing the next test. Note that support for the@DirtiesContext annotation is provided by the DirtiesContextBeforeModesTestExecutionListener and the DirtiesContextTestExecutionListener which are enabled by default.

Context hierarchies

When writing integration tests that rely on a loaded Spring ApplicationContext, it is often sufficient to test against a single context; however, there are times when it is beneficial or even necessary to test against a hierarchy of ApplicationContexts. For example, if you are developing a Spring MVC web application you will typically have a root WebApplicationContext loaded via Spring’s ContextLoaderListener and a child WebApplicationContext loaded via Spring’s DispatcherServlet. This results in a parent-child context hierarchy where shared components and infrastructure configuration are declared in the root context and consumed in the child context by web-specific components. Another use case can be found in Spring Batch applications where you often have a parent context that provides configuration for shared batch infrastructure and a child context for the configuration of a specific batch job.

Since Spring Framework 3.2.2, it is possible to write integration tests that use context hierarchies by declaring context configuration via the @ContextHierarchyannotation, either on an individual test class or within a test class hierarchy. If a context hierarchy is declared on multiple classes within a test class hierarchy it is also possible to merge or override the context configuration for a specific, named level in the context hierarchy. When merging configuration for a given level in the hierarchy the configuration resource type (i.e., XML configuration files or annotated classes) must be consistent; otherwise, it is perfectly acceptable to have different levels in a context hierarchy configured using different resource types.

The following JUnit 4 based examples demonstrate common configuration scenarios for integration tests that require the use of context hierarchies.

ControllerIntegrationTests represents a typical integration testing scenario for a Spring MVC web application by declaring a context hierarchy consisting of two levels, one for the root WebApplicationContext (loaded using the TestAppConfig @Configuration class) and one for the dispatcher servletWebApplicationContext (loaded using the WebConfig @Configuration class). The WebApplicationContext that is autowired into the test instance is the one for the child context (i.e., the lowest context in the hierarchy).

@RunWith(SpringRunner.class)
@WebAppConfiguration
@ContextHierarchy({
    @ContextConfiguration(classes = TestAppConfig.class),
    @ContextConfiguration(classes = WebConfig.class)
})
public class ControllerIntegrationTests {

    @Autowired
    private WebApplicationContext wac;

    // ...
}

The following test classes define a context hierarchy within a test class hierarchy. AbstractWebTests declares the configuration for a root WebApplicationContext in a Spring-powered web application. Note, however, that AbstractWebTests does not declare @ContextHierarchy; consequently, subclasses of AbstractWebTestscan optionally participate in a context hierarchy or simply follow the standard semantics for @ContextConfiguration. SoapWebServiceTests andRestWebServiceTests both extend AbstractWebTests and define a context hierarchy via @ContextHierarchy. The result is that three application contexts will be loaded (one for each declaration of @ContextConfiguration), and the application context loaded based on the configuration in AbstractWebTests will be set as the parent context for each of the contexts loaded for the concrete subclasses.

@RunWith(SpringRunner.class)
@WebAppConfiguration
@ContextConfiguration("file:src/main/webapp/WEB-INF/applicationContext.xml")
public abstract class AbstractWebTests {}

@ContextHierarchy(@ContextConfiguration("/spring/soap-ws-config.xml")
public class SoapWebServiceTests extends AbstractWebTests {}

@ContextHierarchy(@ContextConfiguration("/spring/rest-ws-config.xml")
public class RestWebServiceTests extends AbstractWebTests {}

The following classes demonstrate the use of named hierarchy levels in order to merge the configuration for specific levels in a context hierarchy. BaseTests defines two levels in the hierarchy, parent and child. ExtendedTests extends BaseTests and instructs the Spring TestContext Framework to merge the context configuration for the child hierarchy level, simply by ensuring that the names declared via the name attribute in @ContextConfiguration are both "child". The result is that three application contexts will be loaded: one for "/app-config.xml", one for "/user-config.xml", and one for{"/user-config.xml", "/order-config.xml"}. As with the previous example, the application context loaded from "/app-config.xml" will be set as the parent context for the contexts loaded from "/user-config.xml" and {"/user-config.xml", "/order-config.xml"}.

@RunWith(SpringRunner.class)
@ContextHierarchy({
    @ContextConfiguration(name = "parent", locations = "/app-config.xml"),
    @ContextConfiguration(name = "child", locations = "/user-config.xml")
})
public class BaseTests {}

@ContextHierarchy(
    @ContextConfiguration(name = "child", locations = "/order-config.xml")
)
public class ExtendedTests extends BaseTests {}

In contrast to the previous example, this example demonstrates how to override the configuration for a given named level in a context hierarchy by setting theinheritLocations flag in @ContextConfiguration to false. Consequently, the application context for ExtendedTests will be loaded only from"/test-user-config.xml" and will have its parent set to the context loaded from "/app-config.xml".

@RunWith(SpringRunner.class)
@ContextHierarchy({
    @ContextConfiguration(name = "parent", locations = "/app-config.xml"),
    @ContextConfiguration(name = "child", locations = "/user-config.xml")
})
public class BaseTests {}

@ContextHierarchy(
    @ContextConfiguration(
        name = "child",
        locations = "/test-user-config.xml",
        inheritLocations = false
))
public class ExtendedTests extends BaseTests {}
Spring框架参考文档 Spring Framework Reference Documentation

If @DirtiesContext is used in a test whose context is configured as part of a context hierarchy, the hierarchyMode flag can be used to control how the context cache is cleared. For further details consult the discussion of @DirtiesContext in Spring Testing Annotations and the @DirtiesContextjavadocs.

15.5.5 Dependency injection of test fixtures

When you use the DependencyInjectionTestExecutionListener — which is configured by default — the dependencies of your test instances are injected from beans in the application context that you configured with @ContextConfiguration. You may use setter injection, field injection, or both, depending on which annotations you choose and whether you place them on setter methods or fields. For consistency with the annotation support introduced in Spring 2.5 and 3.0, you can use Spring’s @Autowired annotation or the @Inject annotation from JSR 330.

Spring框架参考文档 Spring Framework Reference Documentation

The TestContext framework does not instrument the manner in which a test instance is instantiated. Thus the use of @Autowired or @Inject for constructors has no effect for test classes.

Because @Autowired is used to perform autowiring by type, if you have multiple bean definitions of the same type, you cannot rely on this approach for those particular beans. In that case, you can use @Autowired in conjunction with @Qualifier. As of Spring 3.0 you may also choose to use @Inject in conjunction with @Named. Alternatively, if your test class has access to its ApplicationContext, you can perform an explicit lookup by using (for example) a call toapplicationContext.getBean("titleRepository").

If you do not want dependency injection applied to your test instances, simply do not annotate fields or setter methods with @Autowired or @Inject. Alternatively, you can disable dependency injection altogether by explicitly configuring your class with @TestExecutionListeners and omittingDependencyInjectionTestExecutionListener.class from the list of listeners.

Consider the scenario of testing a HibernateTitleRepository class, as outlined in the Goals section. The next two code listings demonstrate the use of @Autowiredon fields and setter methods. The application context configuration is presented after all sample code listings.

Spring框架参考文档 Spring Framework Reference Documentation

The dependency injection behavior in the following code listings is not specific to JUnit 4. The same DI techniques can be used in conjunction with any testing framework.

The following examples make calls to static assertion methods such as assertNotNull() but without prepending the call with Assert. In such cases, assume that the method was properly imported through an import static declaration that is not shown in the example.

The first code listing shows a JUnit 4 based implementation of the test class that uses @Autowired for field injection.

@RunWith(SpringRunner.class)
// specifies the Spring configuration to load for this test fixture
@ContextConfiguration("repository-config.xml")
public class HibernateTitleRepositoryTests {

    // this instance will be dependency injected by type
    @Autowired
    private HibernateTitleRepository titleRepository;

    @Test
    public void findById() {
        Title title = titleRepository.findById(new Long(10));
        assertNotNull(title);
    }
}

Alternatively, you can configure the class to use @Autowired for setter injection as seen below.

@RunWith(SpringRunner.class)
// specifies the Spring configuration to load for this test fixture
@ContextConfiguration("repository-config.xml")
public class HibernateTitleRepositoryTests {

    // this instance will be dependency injected by type
    private HibernateTitleRepository titleRepository;

    @Autowired
    public void setTitleRepository(HibernateTitleRepository titleRepository) {
        this.titleRepository = titleRepository;
    }

    @Test
    public void findById() {
        Title title = titleRepository.findById(new Long(10));
        assertNotNull(title);
    }
}

The preceding code listings use the same XML context file referenced by the @ContextConfiguration annotation (that is, repository-config.xml), which looks like this:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.springframework.org/schema/beans
        http://www.springframework.org/schema/beans/spring-beans.xsd">

    <!-- this bean will be injected into the HibernateTitleRepositoryTests class -->
    <bean id="titleRepository" class="com.foo.repository.hibernate.HibernateTitleRepository">
        <property name="sessionFactory" ref="sessionFactory"/>
    </bean>

    <bean id="sessionFactory" class="org.springframework.orm.hibernate5.LocalSessionFactoryBean">
        <!-- configuration elided for brevity -->
    </bean>

</beans>
Spring框架参考文档 Spring Framework Reference Documentation

If you are extending from a Spring-provided test base class that happens to use @Autowired on one of its setter methods, you might have multiple beans of the affected type defined in your application context: for example, multiple DataSource beans. In such a case, you can override the setter method and use the @Qualifier annotation to indicate a specific target bean as follows, but make sure to delegate to the overridden method in the superclass as well.

// ...

    @Autowired
    @Override
    public void setDataSource(@Qualifier("myDataSource") DataSource dataSource) {
        super.setDataSource(dataSource);
    }

// ...

The specified qualifier value indicates the specific DataSource bean to inject, narrowing the set of type matches to a specific bean. Its value is matched against <qualifier> declarations within the corresponding <bean> definitions. The bean name is used as a fallback qualifier value, so you may effectively also point to a specific bean by name there (as shown above, assuming that "myDataSource" is the bean id).

15.5.6 Testing request and session scoped beans

Request and session scoped beans have been supported by Spring since the early years, and since Spring 3.2 it’s a breeze to test your request-scoped and session-scoped beans by following these steps.

  • Ensure that a WebApplicationContext is loaded for your test by annotating your test class with @WebAppConfiguration.
  • Inject the mock request or session into your test instance and prepare your test fixture as appropriate.
  • Invoke your web component that you retrieved from the configured WebApplicationContext (i.e., via dependency injection).
  • Perform assertions against the mocks.

The following code snippet displays the XML configuration for a login use case. Note that the userService bean has a dependency on a request-scoped loginActionbean. Also, the LoginAction is instantiated using SpEL expressions that retrieve the username and password from the current HTTP request. In our test, we will want to configure these request parameters via the mock managed by the TestContext framework.

Request-scoped bean configuration. 

<beans>

    <bean id="userService"
            class="com.example.SimpleUserService"
            c:loginAction-ref="loginAction" />

    <bean id="loginAction" class="com.example.LoginAction"
            c:username="{request.getParameter('user')}"
            c:password="{request.getParameter('pswd')}"
            scope="request">
        <aop:scoped-proxy />
    </bean>

</beans>

 

In RequestScopedBeanTests we inject both the UserService (i.e., the subject under test) and the MockHttpServletRequest into our test instance. Within ourrequestScope() test method we set up our test fixture by setting request parameters in the provided MockHttpServletRequest. When the loginUser() method is invoked on our userService we are assured that the user service has access to the request-scoped loginAction for the current MockHttpServletRequest (i.e., the one we just set parameters in). We can then perform assertions against the results based on the known inputs for the username and password.

Request-scoped bean test. 

@RunWith(SpringRunner.class)
@ContextConfiguration
@WebAppConfiguration
public class RequestScopedBeanTests {

    @Autowired UserService userService;
    @Autowired MockHttpServletRequest request;

    @Test
    public void requestScope() {

        request.setParameter("user", "enigma");
        request.setParameter("pswd", "$pr!ng");

        LoginResults results = userService.loginUser();

        // assert results
    }
}

 

The following code snippet is similar to the one we saw above for a request-scoped bean; however, this time the userService bean has a dependency on a session-scoped userPreferences bean. Note that the UserPreferences bean is instantiated using a SpEL expression that retrieves the theme from the current HTTP session. In our test, we will need to configure a theme in the mock session managed by the TestContext framework.

Session-scoped bean configuration. 

<beans>

    <bean id="userService"
            class="com.example.SimpleUserService"
            c:userPreferences-ref="userPreferences" />

    <bean id="userPreferences"
            class="com.example.UserPreferences"
            c:theme="#{session.getAttribute('theme')}"
            scope="session">
        <aop:scoped-proxy />
    </bean>

</beans>

 

In SessionScopedBeanTests we inject the UserService and the MockHttpSession into our test instance. Within our sessionScope() test method we set up our test fixture by setting the expected "theme" attribute in the provided MockHttpSession. When the processUserPreferences() method is invoked on ouruserService we are assured that the user service has access to the session-scoped userPreferences for the current MockHttpSession, and we can perform assertions against the results based on the configured theme.

Session-scoped bean test. 

@RunWith(SpringRunner.class)
@ContextConfiguration
@WebAppConfiguration
public class SessionScopedBeanTests {

    @Autowired UserService userService;
    @Autowired MockHttpSession session;

    @Test
    public void sessionScope() throws Exception {

        session.setAttribute("theme", "blue");

        Results results = userService.processUserPreferences();

        // assert results
    }
}

 

15.5.7 Transaction management

In the TestContext framework, transactions are managed by the TransactionalTestExecutionListener which is configured by default, even if you do not explicitly declare @TestExecutionListeners on your test class. To enable support for transactions, however, you must configure a PlatformTransactionManager bean in theApplicationContext that is loaded via @ContextConfiguration semantics (further details are provided below). In addition, you must declare Spring’s@Transactional annotation either at the class or method level for your tests.

Test-managed transactions

Test-managed transactions are transactions that are managed declaratively via the TransactionalTestExecutionListener or programmatically viaTestTransaction (see below). Such transactions should not be confused with Spring-managed transactions (i.e., those managed directly by Spring within theApplicationContext loaded for tests) or application-managed transactions (i.e., those managed programmatically within application code that is invoked via tests). Spring-managed and application-managed transactions will typically participate in test-managed transactions; however, caution should be taken if Spring-managed or application-managed transactions are configured with any propagation type other than REQUIRED or SUPPORTS (see the discussion on transaction propagation for details).

Enabling and disabling transactions

Annotating a test method with @Transactional causes the test to be run within a transaction that will, by default, be automatically rolled back after completion of the test. If a test class is annotated with @Transactional, each test method within that class hierarchy will be run within a transaction. Test methods that are not annotated with @Transactional (at the class or method level) will not be run within a transaction. Furthermore, tests that are annotated with @Transactional but have thepropagation type set to NOT_SUPPORTED will not be run within a transaction.

Note that AbstractTransactionalJUnit4SpringContextTests and AbstractTransactionalTestNGSpringContextTests are preconfigured for transactional support at the class level.

The following example demonstrates a common scenario for writing an integration test for a Hibernate-based UserRepository. As explained in the section called “Transaction rollback and commit behavior”, there is no need to clean up the database after the createUser() method is executed since any changes made to the database will be automatically rolled back by the TransactionalTestExecutionListener. See Section 15.7, “PetClinic Example” for an additional example.

@RunWith(SpringRunner.class)
@ContextConfiguration(classes = TestConfig.class)
@Transactional
public class HibernateUserRepositoryTests {

    @Autowired
    HibernateUserRepository repository;

    @Autowired
    SessionFactory sessionFactory;

    JdbcTemplate jdbcTemplate;

    @Autowired
    public void setDataSource(DataSource dataSource) {
        this.jdbcTemplate = new JdbcTemplate(dataSource);
    }

    @Test
    public void createUser() {
        // track initial state in test database:
        final int count = countRowsInTable("user");

        User user = new User(...);
        repository.save(user);

        // Manual flush is required to avoid false positive in test
        sessionFactory.getCurrentSession().flush();
        assertNumUsers(count + 1);
    }

    protected int countRowsInTable(String tableName) {
        return JdbcTestUtils.countRowsInTable(this.jdbcTemplate, tableName);
    }

    protected void assertNumUsers(int expected) {
        assertEquals("Number of rows in the [user] table.", expected, countRowsInTable("user"));
    }
}

Transaction rollback and commit behavior

By default, test transactions will be automatically rolled back after completion of the test; however, transactional commit and rollback behavior can be configured declaratively via the @Commit and @Rollback annotations. See the corresponding entries in the annotation support section for further details.

Programmatic transaction management

Since Spring Framework 4.1, it is possible to interact with test-managed transactions programmatically via the static methods in TestTransaction. For example,TestTransaction may be used within test methods, before methods, and after methods to start or end the current test-managed transaction or to configure the current test-managed transaction for rollback or commit. Support for TestTransaction is automatically available whenever the TransactionalTestExecutionListener is enabled.

The following example demonstrates some of the features of TestTransaction. Consult the javadocs for TestTransaction for further details.

@ContextConfiguration(classes = TestConfig.class)
public class ProgrammaticTransactionManagementTests extends
        AbstractTransactionalJUnit4SpringContextTests {

    @Test
    public void transactionalTest() {
        // assert initial state in test database:
        assertNumUsers(2);

        deleteFromTables("user");

        // changes to the database will be committed!
        TestTransaction.flagForCommit();
        TestTransaction.end();
        assertFalse(TestTransaction.isActive());
        assertNumUsers(0);

        TestTransaction.start();
        // perform other actions against the database that will
        // be automatically rolled back after the test completes...
    }

    protected void assertNumUsers(int expected) {
        assertEquals("Number of rows in the [user] table.", expected, countRowsInTable("user"));
    }
}

Executing code outside of a transaction

Occasionally you need to execute certain code before or after a transactional test method but outside the transactional context — for example, to verify the initial database state prior to execution of your test or to verify expected transactional commit behavior after test execution (if the test was configured to commit the transaction). TransactionalTestExecutionListener supports the @BeforeTransaction and @AfterTransaction annotations exactly for such scenarios. Simply annotate any void method in a test class or any void default method in a test interface with one of these annotations, and theTransactionalTestExecutionListener ensures that your before transaction method or after transaction method is executed at the appropriate time.

Spring框架参考文档 Spring Framework Reference Documentation

Any before methods (such as methods annotated with JUnit 4’s @Before) and any after methods (such as methods annotated with JUnit 4’s @After) are executed within a transaction. In addition, methods annotated with @BeforeTransaction or @AfterTransaction are naturally not executed for test methods that are not configured to run within a transaction.

Configuring a transaction manager

TransactionalTestExecutionListener expects a PlatformTransactionManager bean to be defined in the Spring ApplicationContext for the test. In case there are multiple instances of PlatformTransactionManager within the test’s ApplicationContext, a qualifier may be declared via @Transactional("myTxMgr")or @Transactional(transactionManager = "myTxMgr"), or TransactionManagementConfigurer can be implemented by an @Configuration class. Consult the javadocs for TestContextTransactionUtils.retrieveTransactionManager() for details on the algorithm used to look up a transaction manager in the test’sApplicationContext.

Demonstration of all transaction-related annotations

The following JUnit 4 based example displays a fictitious integration testing scenario highlighting all transaction-related annotations. The example is not intended to demonstrate best practices but rather to demonstrate how these annotations can be used. Consult the annotation support section for further information and configuration examples. Transaction management for @Sql contains an additional example using @Sql for declarative SQL script execution with default transaction rollback semantics.

@RunWith(SpringRunner.class)
@ContextConfiguration
@Transactional(transactionManager = "txMgr")
@Commit
public class FictitiousTransactionalTest {

    @BeforeTransaction
    void verifyInitialDatabaseState() {
        // logic to verify the initial state before a transaction is started
    }

    @Before
    public void setUpTestDataWithinTransaction() {
        // set up test data within the transaction
    }

    @Test
    // overrides the class-level @Commit setting
    @Rollback
    public void modifyDatabaseWithinTransaction() {
        // logic which uses the test data and modifies database state
    }

    @After
    public void tearDownWithinTransaction() {
        // execute "tear down" logic within the transaction
    }

    @AfterTransaction
    void verifyFinalDatabaseState() {
        // logic to verify the final state after transaction has rolled back
    }

}
Spring框架参考文档 Spring Framework Reference Documentation

When you test application code that manipulates the state of a Hibernate session or JPA persistence context, make sure to flush the underlying unit of work within test methods that execute that code. Failing to flush the underlying unit of work can produce false positives: your test may pass, but the same code throws an exception in a live, production environment. In the following Hibernate-based example test case, one method demonstrates a false positive, and the other method correctly exposes the results of flushing the session. Note that this applies to any ORM frameworks that maintain an in-memory unit of work.

// ...

@Autowired
SessionFactory sessionFactory;

@Transactional
@Test // no expected exception!
public void falsePositive() {
    updateEntityInHibernateSession();
    // False positive: an exception will be thrown once the Hibernate
    // Session is finally flushed (i.e., in production code)
}

@Transactional
@Test(expected = ...)
public void updateWithSessionFlush() {
    updateEntityInHibernateSession();
    // Manual flush is required to avoid false positive in test
    sessionFactory.getCurrentSession().flush();
}

// ...

Or for JPA:

// ...

@PersistenceContext
EntityManager entityManager;

@Transactional
@Test // no expected exception!
public void falsePositive() {
    updateEntityInJpaPersistenceContext();
    // False positive: an exception will be thrown once the JPA
    // EntityManager is finally flushed (i.e., in production code)
}

@Transactional
@Test(expected = ...)
public void updateWithEntityManagerFlush() {
    updateEntityInJpaPersistenceContext();
    // Manual flush is required to avoid false positive in test
    entityManager.flush();
}

// ...

15.5.8 Executing SQL scripts

When writing integration tests against a relational database, it is often beneficial to execute SQL scripts to modify the database schema or insert test data into tables. Thespring-jdbc module provides support for initializing an embedded or existing database by executing SQL scripts when the Spring ApplicationContext is loaded. See Section 19.8, “Embedded database support” and Section 19.8.5, “Testing data access logic with an embedded database” for details.

Although it is very useful to initialize a database for testing once when the ApplicationContext is loaded, sometimes it is essential to be able to modify the databaseduring integration tests. The following sections explain how to execute SQL scripts programmatically and declaratively during integration tests.

Executing SQL scripts programmatically

Spring provides the following options for executing SQL scripts programmatically within integration test methods.

  • org.springframework.jdbc.datasource.init.ScriptUtils
  • org.springframework.jdbc.datasource.init.ResourceDatabasePopulator
  • org.springframework.test.context.junit4.AbstractTransactionalJUnit4SpringContextTests
  • org.springframework.test.context.testng.AbstractTransactionalTestNGSpringContextTests

ScriptUtils provides a collection of static utility methods for working with SQL scripts and is mainly intended for internal use within the framework. However, if you require full control over how SQL scripts are parsed and executed, ScriptUtils may suit your needs better than some of the other alternatives described below. Consult the javadocs for individual methods in ScriptUtils for further details.

ResourceDatabasePopulator provides a simple object-based API for programmatically populating, initializing, or cleaning up a database using SQL scripts defined in external resources. ResourceDatabasePopulator provides options for configuring the character encoding, statement separator, comment delimiters, and error handling flags used when parsing and executing the scripts, and each of the configuration options has a reasonable default value. Consult the javadocs for details on default values. To execute the scripts configured in a ResourceDatabasePopulator, you can invoke either the populate(Connection) method to execute the populator against a java.sql.Connection or the execute(DataSource) method to execute the populator against a javax.sql.DataSource. The following example specifies SQL scripts for a test schema and test data, sets the statement separator to "@@", and then executes the scripts against a DataSource.

@Test
public void databaseTest {
    ResourceDatabasePopulator populator = new ResourceDatabasePopulator();
    populator.addScripts(
        new ClassPathResource("test-schema.sql"),
        new ClassPathResource("test-data.sql"));
    populator.setSeparator("@@");
    populator.execute(this.dataSource);
    // execute code that uses the test schema and data
}

Note that ResourceDatabasePopulator internally delegates to ScriptUtils for parsing and executing SQL scripts. Similarly, the executeSqlScript(..) methods in AbstractTransactionalJUnit4SpringContextTests and AbstractTransactionalTestNGSpringContextTests internally use aResourceDatabasePopulator for executing SQL scripts. Consult the javadocs for the various executeSqlScript(..) methods for further details.

Executing SQL scripts declaratively with @Sql

In addition to the aforementioned mechanisms for executing SQL scripts programmatically, SQL scripts can also be configured declaratively in the Spring TestContext Framework. Specifically, the @Sql annotation can be declared on a test class or test method to configure the resource paths to SQL scripts that should be executed against a given database either before or after an integration test method. Note that method-level declarations override class-level declarations and that support for@Sql is provided by the SqlScriptsTestExecutionListener which is enabled by default.

Path resource semantics

Each path will be interpreted as a Spring Resource. A plain path — for example, "schema.sql" — will be treated as a classpath resource that is relative to the package in which the test class is defined. A path starting with a slash will be treated as an absolute classpath resource, for example: "/org/example/schema.sql". A path which references a URL (e.g., a path prefixed with classpath:, file:, http:, etc.) will be loaded using the specified resource protocol.

The following example demonstrates how to use @Sql at the class level and at the method level within a JUnit 4 based integration test class.

@RunWith(SpringRunner.class)
@ContextConfiguration
@Sql("/test-schema.sql")
public class DatabaseTests {

    @Test
    public void emptySchemaTest {
        // execute code that uses the test schema without any test data
    }

    @Test
    @Sql({"/test-schema.sql", "/test-user-data.sql"})
    public void userTest {
        // execute code that uses the test schema and test data
    }
}

Default script detection

If no SQL scripts are specified, an attempt will be made to detect a default script depending on where @Sql is declared. If a default cannot be detected, anIllegalStateException will be thrown.

  • class-level declaration: if the annotated test class is com.example.MyTest, the corresponding default script is "classpath:com/example/MyTest.sql".
  • method-level declaration: if the annotated test method is named testMethod() and is defined in the class com.example.MyTest, the corresponding default script is "classpath:com/example/MyTest.testMethod.sql".

Declaring multiple @Sql sets

If multiple sets of SQL scripts need to be configured for a given test class or test method but with different syntax configuration, different error handling rules, or different execution phases per set, it is possible to declare multiple instances of @Sql. With Java 8, @Sql can be used as a repeatable annotation. Otherwise, the @SqlGroupannotation can be used as an explicit container for declaring multiple instances of @Sql.

The following example demonstrates the use of @Sql as a repeatable annotation using Java 8. In this scenario the test-schema.sql script uses a different syntax for single-line comments.

@Test
@Sql(scripts = "/test-schema.sql", config = @SqlConfig(commentPrefix = "`"))
@Sql("/test-user-data.sql")
public void userTest {
    // execute code that uses the test schema and test data
}

The following example is identical to the above except that the @Sql declarations are grouped together within @SqlGroup for compatibility with Java 6 and Java 7.

@Test
@SqlGroup({
    @Sql(scripts = "/test-schema.sql", config = @SqlConfig(commentPrefix = "`")),
    @Sql("/test-user-data.sql")
)}
public void userTest {
    // execute code that uses the test schema and test data
}

Script execution phases

By default, SQL scripts will be executed before the corresponding test method. However, if a particular set of scripts needs to be executed after the test method — for example, to clean up database state — the executionPhase attribute in @Sql can be used as seen in the following example. Note that ISOLATED andAFTER_TEST_METHOD are statically imported from Sql.TransactionMode and Sql.ExecutionPhase respectively.

@Test
@Sql(
    scripts = "create-test-data.sql",
    config = @SqlConfig(transactionMode = ISOLATED)
)
@Sql(
    scripts = "delete-test-data.sql",
    config = @SqlConfig(transactionMode = ISOLATED),
    executionPhase = AFTER_TEST_METHOD
)
public void userTest {
    // execute code that needs the test data to be committed
    // to the database outside of the test's transaction
}

Script configuration with @SqlConfig

Configuration for script parsing and error handling can be configured via the @SqlConfig annotation. When declared as a class-level annotation on an integration test class, @SqlConfig serves as global configuration for all SQL scripts within the test class hierarchy. When declared directly via the config attribute of the @Sqlannotation, @SqlConfig serves as local configuration for the SQL scripts declared within the enclosing @Sql annotation. Every attribute in @SqlConfig has an implicit default value which is documented in the javadocs of the corresponding attribute. Due to the rules defined for annotation attributes in the Java Language Specification, it is unfortunately not possible to assign a value of null to an annotation attribute. Thus, in order to support overrides of inherited global configuration, @SqlConfigattributes have an explicit default value of either "" for Strings or DEFAULT for Enums. This approach allows local declarations of @SqlConfig to selectively override individual attributes from global declarations of @SqlConfig by providing a value other than "" or DEFAULT. Global @SqlConfig attributes are inherited whenever local @SqlConfig attributes do not supply an explicit value other than "" or DEFAULT. Explicit local configuration therefore overrides global configuration.

The configuration options provided by @Sql and @SqlConfig are equivalent to those supported by ScriptUtils and ResourceDatabasePopulator but are a superset of those provided by the <jdbc:initialize-database/> XML namespace element. Consult the javadocs of individual attributes in @Sql and @SqlConfigfor details.

Transaction management for @Sql

By default, the SqlScriptsTestExecutionListener will infer the desired transaction semantics for scripts configured via @Sql. Specifically, SQL scripts will be executed without a transaction, within an existing Spring-managed transaction — for example, a transaction managed by the TransactionalTestExecutionListenerfor a test annotated with @Transactional — or within an isolated transaction, depending on the configured value of the transactionMode attribute in @SqlConfigand the presence of a PlatformTransactionManager in the test’s ApplicationContext. As a bare minimum however, a javax.sql.DataSource must be present in the test’s ApplicationContext.

If the algorithms used by SqlScriptsTestExecutionListener to detect a DataSource and PlatformTransactionManager and infer the transaction semantics do not suit your needs, you may specify explicit names via the dataSource and transactionManager attributes of @SqlConfig. Furthermore, the transaction propagation behavior can be controlled via the transactionMode attribute of @SqlConfig — for example, if scripts should be executed in an isolated transaction. Although a thorough discussion of all supported options for transaction management with @Sql is beyond the scope of this reference manual, the javadocs for@SqlConfig and SqlScriptsTestExecutionListener provide detailed information, and the following example demonstrates a typical testing scenario using JUnit 4 and transactional tests with @Sql. Note that there is no need to clean up the database after the usersTest() method is executed since any changes made to the database (either within the test method or within the /test-data.sql script) will be automatically rolled back by the TransactionalTestExecutionListener (seetransaction management for details).

@RunWith(SpringRunner.class)
@ContextConfiguration(classes = TestDatabaseConfig.class)
@Transactional
public class TransactionalSqlScriptsTests {

    protected JdbcTemplate jdbcTemplate;

    @Autowired
    public void setDataSource(DataSource dataSource) {
        this.jdbcTemplate = new JdbcTemplate(dataSource);
    }

    @Test
    @Sql("/test-data.sql")
    public void usersTest() {
        // verify state in test database:
        assertNumUsers(2);
        // execute code that uses the test data...
    }

    protected int countRowsInTable(String tableName) {
        return JdbcTestUtils.countRowsInTable(this.jdbcTemplate, tableName);
    }

    protected void assertNumUsers(int expected) {
        assertEquals("Number of rows in the [user] table.", expected, countRowsInTable("user"));
    }
}

15.5.9 TestContext Framework support classes

Spring JUnit 4 Runner

The Spring TestContext Framework offers full integration with JUnit 4 through a custom runner (supported on JUnit 4.12 or higher). By annotating test classes with@RunWith(SpringJUnit4ClassRunner.class) or the shorter @RunWith(SpringRunner.class) variant, developers can implement standard JUnit 4 based unit and integration tests and simultaneously reap the benefits of the TestContext framework such as support for loading application contexts, dependency injection of test instances, transactional test method execution, and so on. If you would like to use the Spring TestContext Framework with an alternative runner such as JUnit 4’sParameterized or third-party runners such as the MockitoJUnitRunner, you may optionally use Spring’s support for JUnit rules instead.

The following code listing displays the minimal requirements for configuring a test class to run with the custom Spring Runner. @TestExecutionListeners is configured with an empty list in order to disable the default listeners, which otherwise would require an ApplicationContext to be configured through@ContextConfiguration.

@RunWith(SpringRunner.class)
@TestExecutionListeners({})
public class SimpleTest {

   @Test
   public void testMethod() {
      // execute test logic...
   }
}

Spring JUnit 4 Rules

The org.springframework.test.context.junit4.rules package provides the following JUnit 4 rules (supported on JUnit 4.12 or higher).

  • SpringClassRule
  • SpringMethodRule

SpringClassRule is a JUnit TestRule that supports class-level features of the Spring TestContext Framework; whereas, SpringMethodRule is a JUnit MethodRulethat supports instance-level and method-level features of the Spring TestContext Framework.

In contrast to the SpringRunner, Spring’s rule-based JUnit support has the advantage that it is independent of any org.junit.runner.Runner implementation and can therefore be combined with existing alternative runners like JUnit 4’s Parameterized or third-party runners such as the MockitoJUnitRunner.

In order to support the full functionality of the TestContext framework, a SpringClassRule must be combined with a SpringMethodRule. The following example demonstrates the proper way to declare these rules in an integration test.

// Optionally specify a non-Spring Runner via @RunWith(...)
@ContextConfiguration
public class IntegrationTest {

   @ClassRule
   public static final SpringClassRule SPRING_CLASS_RULE = new SpringClassRule();

   @Rule
   public final SpringMethodRule springMethodRule = new SpringMethodRule();

   @Test
   public void testMethod() {
      // execute test logic...
   }
}

JUnit 4 support classes

The org.springframework.test.context.junit4 package provides the following support classes for JUnit 4 based test cases (supported on JUnit 4.12 or higher).

  • AbstractJUnit4SpringContextTests
  • AbstractTransactionalJUnit4SpringContextTests

AbstractJUnit4SpringContextTests is an abstract base test class that integrates the Spring TestContext Framework with explicit ApplicationContext testing support in a JUnit 4 environment. When you extend AbstractJUnit4SpringContextTests, you can access a protected applicationContext instance variable that can be used to perform explicit bean lookups or to test the state of the context as a whole.

AbstractTransactionalJUnit4SpringContextTests is an abstract transactional extension of AbstractJUnit4SpringContextTests that adds some convenience functionality for JDBC access. This class expects a javax.sql.DataSource bean and a PlatformTransactionManager bean to be defined in theApplicationContext. When you extend AbstractTransactionalJUnit4SpringContextTests you can access a protected jdbcTemplate instance variable that can be used to execute SQL statements to query the database. Such queries can be used to confirm database state both prior to and after execution of database-related application code, and Spring ensures that such queries run in the scope of the same transaction as the application code. When used in conjunction with an ORM tool, be sure to avoid false positives. As mentioned in Section 15.3, “JDBC Testing Support”, AbstractTransactionalJUnit4SpringContextTests also provides convenience methods which delegate to methods in JdbcTestUtils using the aforementioned jdbcTemplate. Furthermore,AbstractTransactionalJUnit4SpringContextTests provides an executeSqlScript(..) method for executing SQL scripts against the configured DataSource.

Spring框架参考文档 Spring Framework Reference Documentation

These classes are a convenience for extension. If you do not want your test classes to be tied to a Spring-specific class hierarchy, you can configure your own custom test classes by using @RunWith(SpringRunner.class) or Spring’s JUnit rules.

TestNG support classes

The org.springframework.test.context.testng package provides the following support classes for TestNG based test cases.

  • AbstractTestNGSpringContextTests
  • AbstractTransactionalTestNGSpringContextTests

AbstractTestNGSpringContextTests is an abstract base test class that integrates the Spring TestContext Framework with explicit ApplicationContext testing support in a TestNG environment. When you extend AbstractTestNGSpringContextTests, you can access a protected applicationContext instance variable that can be used to perform explicit bean lookups or to test the state of the context as a whole.

AbstractTransactionalTestNGSpringContextTests is an abstract transactional extension of AbstractTestNGSpringContextTests that adds some convenience functionality for JDBC access. This class expects a javax.sql.DataSource bean and a PlatformTransactionManager bean to be defined in theApplicationContext. When you extend AbstractTransactionalTestNGSpringContextTests you can access a protected jdbcTemplate instance variable that can be used to execute SQL statements to query the database. Such queries can be used to confirm database state both prior to and after execution of database-related application code, and Spring ensures that such queries run in the scope of the same transaction as the application code. When used in conjunction with an ORM tool, be sure to avoid false positives. As mentioned in Section 15.3, “JDBC Testing Support”, AbstractTransactionalTestNGSpringContextTests also provides convenience methods which delegate to methods in JdbcTestUtils using the aforementioned jdbcTemplate. Furthermore,AbstractTransactionalTestNGSpringContextTests provides an executeSqlScript(..) method for executing SQL scripts against the configured DataSource.

Spring框架参考文档 Spring Framework Reference Documentation

These classes are a convenience for extension. If you do not want your test classes to be tied to a Spring-specific class hierarchy, you can configure your own custom test classes by using @ContextConfiguration, @TestExecutionListeners, and so on, and by manually instrumenting your test class with a TestContextManager. See the source code of AbstractTestNGSpringContextTests for an example of how to instrument your test class.

15.6 Spring MVC Test Framework

The Spring MVC Test framework provides first class support for testing Spring MVC code using a fluent API that can be used with JUnit, TestNG, or any other testing framework. It’s built on the Servlet API mock objects from the spring-test module and hence does not use a running Servlet container. It uses theDispatcherServlet to provide full Spring MVC runtime behavior and provides support for loading actual Spring configuration with the TestContext framework in addition to a standalone mode in which controllers may be instantiated manually and tested one at a time.

Spring MVC Test also provides client-side support for testing code that uses the RestTemplate. Client-side tests mock the server responses and also do not use a running server.

Spring框架参考文档 Spring Framework Reference Documentation

Spring Boot provides an option to write full, end-to-end integration tests that include a running server. If this is your goal please have a look at the Spring Boot reference page. For more information on the differences between out-of-container and end-to-end integration tests, see the section called “Differences between Out-of-Container and End-to-End Integration Tests”.

15.6.1 Server-Side Tests

It’s easy to write a plain unit test for a Spring MVC controller using JUnit or TestNG: simply instantiate the controller, inject it with mocked or stubbed dependencies, and call its methods passing MockHttpServletRequest, MockHttpServletResponse, etc., as necessary. However, when writing such a unit test, much remains untested: for example, request mappings, data binding, type conversion, validation, and much more. Furthermore, other controller methods such as @InitBinder,@ModelAttribute, and @ExceptionHandler may also be invoked as part of the request processing lifecycle.

The goal of Spring MVC Test is to provide an effective way for testing controllers by performing requests and generating responses through the actualDispatcherServlet.

Spring MVC Test builds on the familiar "mock" implementations of the Servlet API available in the spring-test module. This allows performing requests and generating responses without the need for running in a Servlet container. For the most part everything should work as it does at runtime with a few notable exceptions as explained in the section called “Differences between Out-of-Container and End-to-End Integration Tests”. Here is a JUnit 4 based example of using Spring MVC Test:

import static org.springframework.test.web.servlet.request.MockMvcRequestBuilders.*;
import static org.springframework.test.web.servlet.result.MockMvcResultMatchers.*;

@RunWith(SpringRunner.class)
@WebAppConfiguration
@ContextConfiguration("test-servlet-context.xml")
public class ExampleTests {

    @Autowired
    private WebApplicationContext wac;

    private MockMvc mockMvc;

    @Before
    public void setup() {
        this.mockMvc = MockMvcBuilders.webAppContextSetup(this.wac).build();
    }

    @Test
    public void getAccount() throws Exception {
        this.mockMvc.perform(get("/accounts/1").accept(MediaType.parseMediaType("application/json;charset=UTF-8")))
            .andExpect(status().isOk())
            .andExpect(content().contentType("application/json"))
            .andExpect(jsonPath("$.name").value("Lee"));
    }

}

The above test relies on the WebApplicationContext support of the TestContext framework for loading Spring configuration from an XML configuration file located in the same package as the test class, but Java-based and Groovy-based configuration are also supported. See these sample tests.

The MockMvc instance is used to perform a GET request to "/accounts/1" and verify that the resulting response has status 200, the content type is"application/json", and the response body has a JSON property called "name" with the value "Lee". The jsonPath syntax is supported through the JaywayJsonPath project. There are lots of other options for verifying the result of the performed request that will be discussed below.

Static Imports

The fluent API in the example above requires a few static imports such as MockMvcRequestBuilders.*, MockMvcResultMatchers.*, and MockMvcBuilders.*. An easy way to find these classes is to search for types matching "MockMvc*". If using Eclipse, be sure to add them as "favorite static members" in the Eclipse preferences under Java → Editor → Content Assist → Favorites. That will allow use of content assist after typing the first character of the static method name. Other IDEs (e.g. IntelliJ) may not require any additional configuration. Just check the support for code completion on static members.

Setup Options

There are two main options for creating an instance of MockMvc. The first is to load Spring MVC configuration through the TestContext framework, which loads the Spring configuration and injects a WebApplicationContext into the test to use to build a MockMvc instance:

@RunWith(SpringRunner.class)
@WebAppConfiguration
@ContextConfiguration("my-servlet-context.xml")
public class MyWebTests {

    @Autowired
    private WebApplicationContext wac;

    private MockMvc mockMvc;

    @Before
    public void setup() {
        this.mockMvc = MockMvcBuilders.webAppContextSetup(this.wac).build();
    }

    // ...

}

The second is to simply create a controller instance manually without loading Spring configuration. Instead basic default configuration, roughly comparable to that of the MVC JavaConfig or the MVC namespace, is automatically created and can be customized to a degree:

public class MyWebTests {

    private MockMvc mockMvc;

    @Before
    public void setup() {
        this.mockMvc = MockMvcBuilders.standaloneSetup(new AccountController()).build();
    }

    // ...

}

Which setup option should you use?

The "webAppContextSetup" loads your actual Spring MVC configuration resulting in a more complete integration test. Since the TestContext framework caches the loaded Spring configuration, it helps keep tests running fast, even as you introduce more tests in your test suite. Furthermore, you can inject mock services into controllers through Spring configuration in order to remain focused on testing the web layer. Here is an example of declaring a mock service with Mockito:

<bean id="accountService" class="org.mockito.Mockito" factory-method="mock">
    <constructor-arg value="org.example.AccountService"/>
</bean>

You can then inject the mock service into the test in order set up and verify expectations:

@RunWith(SpringRunner.class)
@WebAppConfiguration
@ContextConfiguration("test-servlet-context.xml")
public class AccountTests {

    @Autowired
    private WebApplicationContext wac;

    private MockMvc mockMvc;

    @Autowired
    private AccountService accountService;

    // ...

}

The "standaloneSetup" on the other hand is a little closer to a unit test. It tests one controller at a time: the controller can be injected with mock dependencies manually, and it doesn’t involve loading Spring configuration. Such tests are more focused on style and make it easier to see which controller is being tested, whether any specific Spring MVC configuration is required to work, and so on. The "standaloneSetup" is also a very convenient way to write ad-hoc tests to verify specific behavior or to debug an issue.

Just like with any "integration vs. unit testing" debate, there is no right or wrong answer. However, using the "standaloneSetup" does imply the need for additional "webAppContextSetup" tests in order to verify your Spring MVC configuration. Alternatively, you may choose to write all tests with "webAppContextSetup" in order to always test against your actual Spring MVC configuration.

Performing Requests

It’s easy to perform requests using any HTTP method:

mockMvc.perform(post("/hotels/{id}", 42).accept(MediaType.APPLICATION_JSON));

You can also perform file upload requests that internally use MockMultipartHttpServletRequest so that there is no actual parsing of a multipart request but rather you have to set it up:

mockMvc.perform(fileUpload("/doc").file("a1", "ABC".getBytes("UTF-8")));

You can specify query parameters in URI template style:

mockMvc.perform(get("/hotels?foo={foo}", "bar"));

Or you can add Servlet request parameters representing either query of form parameters:

mockMvc.perform(get("/hotels").param("foo", "bar"));

If application code relies on Servlet request parameters and doesn’t check the query string explicitly (as is most often the case) then it doesn’t matter which option you use. Keep in mind however that query params provided with the URI template will be decoded while request parameters provided through the param(…​) method are expected to already be decoded.

In most cases it’s preferable to leave out the context path and the Servlet path from the request URI. If you must test with the full request URI, be sure to set thecontextPath and servletPath accordingly so that request mappings will work:

mockMvc.perform(get("/app/main/hotels/{id}").contextPath("/app").servletPath("/main"))

Looking at the above example, it would be cumbersome to set the contextPath and servletPath with every performed request. Instead you can set up default request properties:

public class MyWebTests {

    private MockMvc mockMvc;

    @Before
    public void setup() {
        mockMvc = standaloneSetup(new AccountController())
            .defaultRequest(get("/")
            .contextPath("/app").servletPath("/main")
            .accept(MediaType.APPLICATION_JSON).build();
    }

The above properties will affect every request performed through the MockMvc instance. If the same property is also specified on a given request, it overrides the default value. That is why the HTTP method and URI in the default request don’t matter since they must be specified on every request.

Defining Expectations

Expectations can be defined by appending one or more .andExpect(..) calls after performing a request:

mockMvc.perform(get("/accounts/1")).andExpect(status().isOk());

MockMvcResultMatchers.* provides a number of expectations, some of which are further nested with more detailed expectations.

Expectations fall in two general categories. The first category of assertions verifies properties of the response: for example, the response status, headers, and content. These are the most important results to assert.

The second category of assertions goes beyond the response. These assertions allow one to inspect Spring MVC specific aspects such as which controller method processed the request, whether an exception was raised and handled, what the content of the model is, what view was selected, what flash attributes were added, and so on. They also allow one to inspect Servlet specific aspects such as request and session attributes.

The following test asserts that binding or validation failed:

mockMvc.perform(post("/persons"))
    .andExpect(status().isOk())
    .andExpect(model().attributeHasErrors("person"));

Many times when writing tests, it’s useful to dump the results of the performed request. This can be done as follows, where print() is a static import fromMockMvcResultHandlers:

mockMvc.perform(post("/persons"))
    .andDo(print())
    .andExpect(status().isOk())
    .andExpect(model().attributeHasErrors("person"));

As long as request processing does not cause an unhandled exception, the print() method will print all the available result data to System.out. Spring Framework 4.2 introduced a log() method and two additional variants of the print() method, one that accepts an OutputStream and one that accepts a Writer. For example, invoking print(System.err) will print the result data to System.err; while invoking print(myWriter) will print the result data to a custom writer. If you would like to have the result data logged instead of printed, simply invoke the log() method which will log the result data as a single DEBUG message under theorg.springframework.test.web.servlet.result logging category.

In some cases, you may want to get direct access to the result and verify something that cannot be verified otherwise. This can be achieved by appending.andReturn() after all other expectations:

MvcResult mvcResult = mockMvc.perform(post("/persons")).andExpect(status().isOk()).andReturn();
// ...

If all tests repeat the same expectations you can set up common expectations once when building the MockMvc instance:

standaloneSetup(new SimpleController())
    .alwaysExpect(status().isOk())
    .alwaysExpect(content().contentType("application/json;charset=UTF-8"))
    .build()

Note that common expectations are always applied and cannot be overridden without creating a separate MockMvc instance.

When JSON response content contains hypermedia links created with Spring HATEOAS, the resulting links can be verified using JsonPath expressions:

mockMvc.perform(get("/people").accept(MediaType.APPLICATION_JSON))
    .andExpect(jsonPath("$.links[?(@.rel == 'self')].href").value("http://localhost:8080/people"));

When XML response content contains hypermedia links created with Spring HATEOAS, the resulting links can be verified using XPath expressions:

Map<String, String> ns = Collections.singletonMap("ns", "http://www.w3.org/2005/Atom");
mockMvc.perform(get("/handle").accept(MediaType.APPLICATION_XML))
    .andExpect(xpath("/person/ns:link[@rel='self']/@href", ns).string("http://localhost:8080/people"));

Filter Registrations

When setting up a MockMvc instance, you can register one or more Servlet Filter instances:

mockMvc = standaloneSetup(new PersonController()).addFilters(new CharacterEncodingFilter()).build();

Registered filters will be invoked through via the MockFilterChain from spring-test, and the last filter will delegate to the DispatcherServlet.

Differences between Out-of-Container and End-to-End Integration Tests

As mentioned earlier Spring MVC Test is built on the Servlet API mock objects from the spring-test module and does not use a running Servlet container. Therefore there are some important differences compared to full end-to-end integration tests with an actual client and server running.

The easiest way to think about this is starting with a blank MockHttpServletRequest. Whatever you add to it is what the request will be. Things that may catch you by surprise are that there is no context path by default, no jsessionid cookie, no forwarding, error, or async dispatches, and therefore no actual JSP rendering. Instead, "forwarded" and "redirected" URLs are saved in the MockHttpServletResponse and can be asserted with expectations.

This means if you are using JSPs you can verify the JSP page to which the request was forwarded, but there won’t be any HTML rendered. In other words, the JSP will not be invoked. Note however that all other rendering technologies which don’t rely on forwarding such as Thymeleaf, Freemarker, and Velocity will render HTML to the response body as expected. The same is true for rendering JSON, XML, and other formats via @ResponseBody methods.

Alternatively you may consider the full end-to-end integration testing support from Spring Boot via @WebIntegrationTest. See the Spring Boot reference.

There are pros and cons for each approach. The options provided in Spring MVC Test are different stops on the scale from classic unit testing to full integration testing. To be certain, none of the options in Spring MVC Test fall under the category of classic unit testing, but they are a little closer to it. For example, you can isolate the web layer by injecting mocked services into controllers, in which case you’re testing the web layer only through the DispatcherServlet but with actual Spring configuration, just like you might test the data access layer in isolation from the layers above. Or you can use the standalone setup focusing on one controller at a time and manually providing the configuration required to make it work.

Another important distinction when using Spring MVC Test is that conceptually such tests are on the inside of the server-side so you can check what handler was used, if an exception was handled with a HandlerExceptionResolver, what the content of the model is, what binding errors there were, etc. That means it’s easier to write expectations since the server is not a black box as it is when testing it through an actual HTTP client. This is generally an advantage of classic unit testing, that it’s easier to write, reason about, and debug but does not replace the need for full integration tests. At the same time it’s important not to lose sight of the fact that the response is the most important thing to check. In short, there is room here for multiple styles and strategies of testing even within the same project.

Further Server-Side Test Examples

The framework’s own tests include many sample tests intended to demonstrate how to use Spring MVC Test. Browse these examples for further ideas. Also the spring-mvc-showcase has full test coverage based on Spring MVC Test.

15.6.2 HtmlUnit Integration

Spring provides integration between MockMvc and HtmlUnit. This simplifies performing end-to-end testing when using HTML based views. This integration enables developers to:

  • Easily test HTML pages using tools such as HtmlUnit, WebDriver, & Geb without the need to deploy to a Servlet container
  • Test JavaScript within pages
  • Optionally test using mock services to speed up testing
  • Share logic between in-container end-to-end tests and out-of-container integration tests
Spring框架参考文档 Spring Framework Reference Documentation

MockMvc works with templating technologies that do not rely on a Servlet Container (e.g., Thymeleaf, Freemarker, Velocity, etc.), but it does not work with JSPs since they rely on the Servlet Container.

Why HtmlUnit Integration?

The most obvious question that comes to mind is, "Why do I need this?". The answer is best found by exploring a very basic sample application. Assume you have a Spring MVC web application that supports CRUD operations on a Message object. The application also supports paging through all messages. How would you go about testing it?

With Spring MVC Test, we can easily test if we are able to create a Message.

MockHttpServletRequestBuilder createMessage = post("/messages/")
	.param("summary", "Spring Rocks")
	.param("text", "In case you didn't know, Spring Rocks!");

mockMvc.perform(createMessage)
	.andExpect(status().is3xxRedirection())
	.andExpect(redirectedUrl("/messages/123"));

What if we want to test our form view that allows us to create the message? For example, assume our form looks like the following snippet:

<form id="messageForm" action="/messages/" method="post">
  <div class="pull-right"><a href="/messages/">Messages</a></div>

  <label for="summary">Summary</label>
  <input type="text" class="required" id="summary" name="summary" value="" />

  <label for="text">Message</label>
  <textarea id="text" name="text"></textarea>

  <div class="form-actions">
	<input type="submit" value="Create" />
  </div>
</form>

How do we ensure that our form will produce the correct request to create a new message? A naive attempt would look like this:

mockMvc.perform(get("/messages/form"))
	.andExpect(xpath("//input[@name='summary']").exists())
	.andExpect(xpath("//textarea[@name='text']").exists());

This test has some obvious drawbacks. If we update our controller to use the parameter message instead of text, our form test would continue to pass even though the HTML form is out of synch with the controller. To resolve this we can combine our two tests.

String summaryParamName = "summary";
String textParamName = "text";
mockMvc.perform(get("/messages/form"))
		.andExpect(xpath("//input[@name='" + summaryParamName + "']").exists())
		.andExpect(xpath("//textarea[@name='" + textParamName + "']").exists());

MockHttpServletRequestBuilder createMessage = post("/messages/")
		.param(summaryParamName, "Spring Rocks")
		.param(textParamName, "In case you didn't know, Spring Rocks!");

mockMvc.perform(createMessage)
		.andExpect(status().is3xxRedirection())
		.andExpect(redirectedUrl("/messages/123"));

This would reduce the risk of our test incorrectly passing, but there are still some problems.

  • What if we have multiple forms on our page? Admittedly we could update our xpath expressions, but they get more complicated the more factors we take into account (Are the fields the correct type? Are the fields enabled? etc.).
  • Another issue is that we are doing double the work we would expect. We must first verify the view, and then we submit the view with the same parameters we just verified. Ideally this could be done all at once.
  • Finally, there are some things that we still cannot account for. For example, what if the form has JavaScript validation that we wish to test as well?

The overall problem is that testing a web page does not involve a single interaction. Instead, it is a combination of how the user interacts with a web page and how that web page interacts with other resources. For example, the result of a form view is used as the input to a user for creating a message. In addition, our form view may potentially utilize additional resources which impact the behavior of the page, such as JavaScript validation.

Integration testing to the rescue?

To resolve the issues above we could perform end-to-end integration testing, but this has some obvious drawbacks. Consider testing the view that allows us to page through the messages. We might need the following tests.

  • Does our page display a notification to the user indicating that no results are available when the messages are empty?
  • Does our page properly display a single message?
  • Does our page properly support paging?

To set up these tests, we would need to ensure our database contained the proper messages in it. This leads to a number of additional challenges.

  • Ensuring the proper messages are in the database can be tedious; consider foreign key constraints.
  • Testing can become slow since each test would need to ensure that the database is in the correct state.
  • Since our database needs to be in a specific state, we cannot run tests in parallel.
  • Performing assertions on things like auto-generated ids, timestamps, etc. can be difficult.

These challenges do not mean that we should abandon end-to-end integration testing altogether. Instead, we can reduce the number of end-to-end integration tests by refactoring our detailed tests to use mock services which will execute much faster, more reliably, and without side effects. We can then implement a small number of trueend-to-end integration tests that validate simple workflows to ensure that everything works together properly.

Enter HtmlUnit Integration

So how can we achieve a balance between testing the interactions of our pages and still retain good performance within our test suite? The answer is: "By integrating MockMvc with HtmlUnit."

HtmlUnit Integration Options

There are a number of ways to integrate MockMvc with HtmlUnit.

  • MockMvc and HtmlUnit: Use this option if you want to use the raw HtmlUnit libraries.
  • MockMvc and WebDriver: Use this option to ease development and reuse code between integration and end-to-end testing.
  • MockMvc and Geb: Use this option if you would like to use Groovy for testing, ease development, and reuse code between integration and end-to-end testing.

MockMvc and HtmlUnit

This section describes how to integrate MockMvc and HtmlUnit. Use this option if you want to use the raw HtmlUnit libraries.

MockMvc and HtmlUnit Setup

First, make sure that you have included a test dependency on net.sourceforge.htmlunit:htmlunit. In order to use HtmlUnit with Apache HttpComponents 4.5+, you will need to use HtmlUnit 2.18 or higher.

We can easily create an HtmlUnit WebClient that integrates with MockMvc using the MockMvcWebClientBuilder as follows.

@Autowired
WebApplicationContext context;

WebClient webClient;

@Before
public void setup() {
	webClient = MockMvcWebClientBuilder
		.webAppContextSetup(context)
		.build();
}
Spring框架参考文档 Spring Framework Reference Documentation

This is a simple example of using MockMvcWebClientBuilder. For advanced usage see the section called “Advanced MockMvcWebClientBuilder”

This will ensure that any URL referencing localhost as the server will be directed to our MockMvc instance without the need for a real HTTP connection. Any other URL will be requested using a network connection as normal. This allows us to easily test the use of CDNs.

MockMvc and HtmlUnit Usage

Now we can use HtmlUnit as we normally would, but without the need to deploy our application to a Servlet container. For example, we can request the view to create a message with the following.

HtmlPage createMsgFormPage = webClient.getPage("http://localhost/messages/form");
Spring框架参考文档 Spring Framework Reference Documentation

The default context path is "". Alternatively, we can specify the context path as illustrated in the section called “Advanced MockMvcWebClientBuilder”.

Once we have a reference to the HtmlPage, we can then fill out the form and submit it to create a message.

HtmlForm form = createMsgFormPage.getHtmlElementById("messageForm");
HtmlTextInput summaryInput = createMsgFormPage.getHtmlElementById("summary");
summaryInput.setValueAttribute("Spring Rocks");
HtmlTextArea textInput = createMsgFormPage.getHtmlElementById("text");
textInput.setText("In case you didn't know, Spring Rocks!");
HtmlSubmitInput submit = form.getOneHtmlElementByAttribute("input", "type", "submit");
HtmlPage newMessagePage = submit.click();

Finally, we can verify that a new message was created successfully. The following assertions use the AssertJ library.

assertThat(newMessagePage.getUrl().toString()).endsWith("/messages/123");
String id = newMessagePage.getHtmlElementById("id").getTextContent();
assertThat(id).isEqualTo("123");
String summary = newMessagePage.getHtmlElementById("summary").getTextContent();
assertThat(summary).isEqualTo("Spring Rocks");
String text = newMessagePage.getHtmlElementById("text").getTextContent();
assertThat(text).isEqualTo("In case you didn't know, Spring Rocks!");

This improves on our MockMvc test in a number of ways. First we no longer have to explicitly verify our form and then create a request that looks like the form. Instead, we request the form, fill it out, and submit it, thereby significantly reducing the overhead.

Another important factor is that HtmlUnit uses the Mozilla Rhino engine to evaluate JavaScript. This means that we can test the behavior of JavaScript within our pages as well!

Refer to the HtmlUnit documentation for additional information about using HtmlUnit.

Advanced MockMvcWebClientBuilder

In the examples so far, we have used MockMvcWebClientBuilder in the simplest way possible, by building a WebClient based on the WebApplicationContextloaded for us by the Spring TestContext Framework. This approach is repeated here.

@Autowired
WebApplicationContext context;

WebClient webClient;

@Before
public void setup() {
	webClient = MockMvcWebClientBuilder
		.webAppContextSetup(context)
		.build();
}

We can also specify additional configuration options.

WebClient webClient;

@Before
public void setup() {
	webClient = MockMvcWebClientBuilder
		// demonstrates applying a MockMvcConfigurer (Spring Security)
		.webAppContextSetup(context, springSecurity())
		// for illustration only - defaults to ""
		.contextPath("")
		// By default MockMvc is used for localhost only;
		// the following will use MockMvc for example.com and example.org as well
		.useMockMvcForHosts("example.com","example.org")
		.build();
}

As an alternative, we can perform the exact same setup by configuring the MockMvc instance separately and supplying it to the MockMvcWebClientBuilder as follows.

MockMvc mockMvc = MockMvcBuilders
		.webAppContextSetup(context)
		.apply(springSecurity())
		.build();

webClient = MockMvcWebClientBuilder
		.mockMvcSetup(mockMvc)
		// for illustration only - defaults to ""
		.contextPath("")
		// By default MockMvc is used for localhost only;
		// the following will use MockMvc for example.com and example.org as well
		.useMockMvcForHosts("example.com","example.org")
		.build();

This is more verbose, but by building the WebClient with a MockMvc instance we have the full power of MockMvc at our fingertips.

Spring框架参考文档 Spring Framework Reference Documentation

For additional information on creating a MockMvc instance refer to the section called “Setup Options”.

MockMvc and WebDriver

In the previous sections, we have seen how to use MockMvc in conjunction with the raw HtmlUnit APIs. In this section, we will leverage additional abstractions within the Selenium WebDriver to make things even easier.

Why WebDriver and MockMvc?

We can already use HtmlUnit and MockMvc, so why would we want to use WebDriver? The Selenium WebDriver provides a very elegant API that allows us to easily organize our code. To better understand, let’s explore an example.

Spring框架参考文档 Spring Framework Reference Documentation

Despite being a part of Selenium, WebDriver does not require a Selenium Server to run your tests.

Suppose we need to ensure that a message is created properly. The tests involve finding the HTML form input elements, filling them out, and making various assertions.

This approach results in numerous, separate tests because we want to test error conditions as well. For example, we want to ensure that we get an error if we fill out only part of the form. If we fill out the entire form, the newly created message should be displayed afterwards.

If one of the fields were named "summary", then we might have something like the following repeated in multiple places within our tests.

HtmlTextInput summaryInput = currentPage.getHtmlElementById("summary");
summaryInput.setValueAttribute(summary);

So what happens if we change the id to "smmry"? Doing so would force us to update all of our tests to incorporate this change! Of course, this violates the DRY Principle; so we should ideally extract this code into its own method as follows.

public HtmlPage createMessage(HtmlPage currentPage, String summary, String text) {
	setSummary(currentPage, summary);
	// ...
}

public void setSummary(HtmlPage currentPage, String summary) {
	HtmlTextInput summaryInput = currentPage.getHtmlElementById("summary");
	summaryInput.setValueAttribute(summary);
}

This ensures that we do not have to update all of our tests if we change the UI.

We might even take this a step further and place this logic within an Object that represents the HtmlPage we are currently on.

public class CreateMessagePage {

	final HtmlPage currentPage;

	final HtmlTextInput summaryInput;

	final HtmlSubmitInput submit;

	public CreateMessagePage(HtmlPage currentPage) {
		this.currentPage = currentPage;
		this.summaryInput = currentPage.getHtmlElementById("summary");
		this.submit = currentPage.getHtmlElementById("submit");
	}

	public <T> T createMessage(String summary, String text) throws Exception {
		setSummary(summary);

		HtmlPage result = submit.click();
		boolean error = CreateMessagePage.at(result);

		return (T) (error ? new CreateMessagePage(result) : new ViewMessagePage(result));
	}

	public void setSummary(String summary) throws Exception {
		summaryInput.setValueAttribute(summary);
	}

	public static boolean at(HtmlPage page) {
		return "Create Message".equals(page.getTitleText());
	}
}

Formerly, this pattern is known as the Page Object Pattern. While we can certainly do this with HtmlUnit, WebDriver provides some tools that we will explore in the following sections to make this pattern much easier to implement.

MockMvc and WebDriver Setup

To use Selenium WebDriver with the Spring MVC Test framework, make sure that your project includes a test dependency onorg.seleniumhq.selenium:selenium-htmlunit-driver.

We can easily create a Selenium WebDriver that integrates with MockMvc using the MockMvcHtmlUnitDriverBuilder as follows.

@Autowired
WebApplicationContext context;

WebDriver driver;

@Before
public void setup() {
	driver = MockMvcHtmlUnitDriverBuilder
		.webAppContextSetup(context)
		.build();
}
Spring框架参考文档 Spring Framework Reference Documentation

This is a simple example of using MockMvcHtmlUnitDriverBuilder. For more advanced usage, refer to the section called “Advanced MockMvcHtmlUnitDriverBuilder”

This will ensure that any URL referencing localhost as the server will be directed to our MockMvc instance without the need for a real HTTP connection. Any other URL will be requested using a network connection as normal. This allows us to easily test the use of CDNs.

MockMvc and WebDriver Usage

Now we can use WebDriver as we normally would, but without the need to deploy our application to a Servlet container. For example, we can request the view to create a message with the following.

CreateMessagePage page = CreateMessagePage.to(driver);

We can then fill out the form and submit it to create a message.

ViewMessagePage viewMessagePage =
	page.createMessage(ViewMessagePage.class, expectedSummary, expectedText);

This improves on the design of our HtmlUnit test by leveraging the Page Object Pattern. As we mentioned in the section called “Why WebDriver and MockMvc?”, we can use the Page Object Pattern with HtmlUnit, but it is much easier with WebDriver. Let’s take a look at our new CreateMessagePage implementation.

public class CreateMessagePage
		extends AbstractPage { Spring框架参考文档 Spring Framework Reference DocumentationSpring框架参考文档 Spring Framework Reference Documentationprivate WebElement summary;
	private WebElement text;

	Spring框架参考文档 Spring Framework Reference Documentation@FindBy(css = "input[type=submit]")
	private WebElement submit;

	public CreateMessagePage(WebDriver driver) {
		super(driver);
	}

	public <T> T createMessage(Class<T> resultPage, String summary, String details) {
		this.summary.sendKeys(summary);
		this.text.sendKeys(details);
		this.submit.click();
		return PageFactory.initElements(driver, resultPage);
	}

	public static CreateMessagePage to(WebDriver driver) {
		driver.get("http://localhost:9990/mail/messages/form");
		return PageFactory.initElements(driver, CreateMessagePage.class);
	}
}

Spring框架参考文档 Spring Framework Reference Documentation

The first thing you will notice is that CreateMessagePage extends the AbstractPage. We won’t go over the details of AbstractPage, but in summary it contains common functionality for all of our pages. For example, if our application has a navigational bar, global error messages, etc., this logic can be placed in a shared location.

Spring框架参考文档 Spring Framework Reference Documentation

The next thing you will notice is that we have a member variable for each of the parts of the HTML page that we are interested in. These are of type WebElement.WebDriver's PageFactory allows us to remove a lot of code from the HtmlUnit version of CreateMessagePage by automatically resolving each WebElement. ThePageFactory#initElements(WebDriver,Class<T>) method will automatically resolve each WebElement by using the field name and looking it up by the id or nameof the element within the HTML page.

Spring框架参考文档 Spring Framework Reference Documentation

We can use the @FindBy annotation to override the default lookup behavior. Our example demonstrates how to use the @FindBy annotation to look up our submit button using a css selector, input[type=submit].

Finally, we can verify that a new message was created successfully. The following assertions use the FEST assertion library.

assertThat(viewMessagePage.getMessage()).isEqualTo(expectedMessage);
assertThat(viewMessagePage.getSuccess()).isEqualTo("Successfully created a new message");

We can see that our ViewMessagePage allows us to interact with our custom domain model. For example, it exposes a method that returns a Message object.

public Message getMessage() throws ParseException {
	Message message = new Message();
	message.setId(getId());
	message.setCreated(getCreated());
	message.setSummary(getSummary());
	message.setText(getText());
	return message;
}

We can then leverage the rich domain objects in our assertions.

Lastly, don’t forget to close the WebDriver instance when the test is complete.

@After
public void destroy() {
	if (driver != null) {
		driver.close();
	}
}

For additional information on using WebDriver, refer to the Selenium WebDriver documentation.

Advanced MockMvcHtmlUnitDriverBuilder

In the examples so far, we have used MockMvcHtmlUnitDriverBuilder in the simplest way possible, by building a WebDriver based on theWebApplicationContext loaded for us by the Spring TestContext Framework. This approach is repeated here.

@Autowired
WebApplicationContext context;

WebDriver driver;

@Before
public void setup() {
	driver = MockMvcHtmlUnitDriverBuilder
		.webAppContextSetup(context)
		.build();
}

We can also specify additional configuration options.

WebDriver driver;

@Before
public void setup() {
	driver = MockMvcHtmlUnitDriverBuilder
		// demonstrates applying a MockMvcConfigurer (Spring Security)
		.webAppContextSetup(context, springSecurity())
		// for illustration only - defaults to ""
		.contextPath("")
		// By default MockMvc is used for localhost only;
		// the following will use MockMvc for example.com and example.org as well
		.useMockMvcForHosts("example.com","example.org")
		.build();
}

As an alternative, we can perform the exact same setup by configuring the MockMvc instance separately and supplying it to the MockMvcHtmlUnitDriverBuilder as follows.

MockMvc mockMvc = MockMvcBuilders
		.webAppContextSetup(context)
		.apply(springSecurity())
		.build();

driver = MockMvcHtmlUnitDriverBuilder
		.mockMvcSetup(mockMvc)
		// for illustration only - defaults to ""
		.contextPath("")
		// By default MockMvc is used for localhost only;
		// the following will use MockMvc for example.com and example.org as well
		.useMockMvcForHosts("example.com","example.org")
		.build();

This is more verbose, but by building the WebDriver with a MockMvc instance we have the full power of MockMvc at our fingertips.

Spring框架参考文档 Spring Framework Reference Documentation

For additional information on creating a MockMvc instance refer to the section called “Setup Options”.

MockMvc and Geb

In the previous section, we saw how to use MockMvc with WebDriver. In this section, we will use Geb to make our tests even Groovy-er.

Why Geb and MockMvc?

Geb is backed by WebDriver, so it offers many of the same benefits that we get from WebDriver. However, Geb makes things even easier by taking care of some of the boilerplate code for us.

MockMvc and Geb Setup

We can easily initialize a Geb Browser with a Selenium WebDriver that uses MockMvc as follows.

def setup() {
	browser.driver = MockMvcHtmlUnitDriverBuilder
		.webAppContextSetup(context)
		.build()
}
Spring框架参考文档 Spring Framework Reference Documentation

This is a simple example of using MockMvcHtmlUnitDriverBuilder. For more advanced usage, refer to the section called “Advanced MockMvcHtmlUnitDriverBuilder”

This will ensure that any URL referencing localhost as the server will be directed to our MockMvc instance without the need for a real HTTP connection. Any other URL will be requested using a network connection as normal. This allows us to easily test the use of CDNs.

MockMvc and Geb Usage

Now we can use Geb as we normally would, but without the need to deploy our application to a Servlet container. For example, we can request the view to create a message with the following:

to CreateMessagePage

We can then fill out the form and submit it to create a message.

when:
form.summary = expectedSummary
form.text = expectedMessage
submit.click(ViewMessagePage)

Any unrecognized method calls or property accesses/references that are not found will be forwarded to the current page object. This removes a lot of the boilerplate code we needed when using WebDriver directly.

As with direct WebDriver usage, this improves on the design of our HtmlUnit test by leveraging the Page Object Pattern. As mentioned previously, we can use the Page Object Pattern with HtmlUnit and WebDriver, but it is even easier with Geb. Let’s take a look at our new Groovy-based CreateMessagePage implementation.

class CreateMessagePage extends Page {
	static url = 'messages/form'
	static at = { assert title == 'Messages : Create'; true }
	static content =  {
		submit { $('input[type=submit]') }
		form { $('form') }
		errors(required:false) { $('label.error, .alert-error')?.text() }
	}
}

The first thing you will notice is that our CreateMessagePage extends Page. We won’t go over the details of Page, but in summary it contains common functionality for all of our pages. The next thing you will notice is that we define a URL in which this page can be found. This allows us to navigate to the page as follows.

to CreateMessagePage

We also have an at closure that determines if we are at the specified page. It should return true if we are on the correct page. This is why we can assert that we are on the correct page as follows.

then:
at CreateMessagePage
errors.contains('This field is required.')
Spring框架参考文档 Spring Framework Reference Documentation

We use an assertion in the closure, so that we can determine where things went wrong if we were at the wrong page.

Next we create a content closure that specifies all the areas of interest within the page. We can use a jQuery-ish Navigator API to select the content we are interested in.

Finally, we can verify that a new message was created successfully.

then:
at ViewMessagePage
success == 'Successfully created a new message'
id
date
summary == expectedSummary
message == expectedMessage

For further details on how to get the most out of Geb, consult The Book of Geb user’s manual.

15.6.3 Client-Side REST Tests

Client-side tests can be used to test code that internally uses the RestTemplate. The idea is to declare expected requests and to provide "stub" responses so that you can focus on testing the code in isolation, i.e. without running a server. Here is an example:

RestTemplate restTemplate = new RestTemplate();

MockRestServiceServer mockServer = MockRestServiceServer.bindTo(restTemplate).build();
mockServer.expect(requestTo("/greeting")).andRespond(withSuccess());

// Test code that uses the above RestTemplate ...

mockServer.verify();

In the above example, MockRestServiceServer, the central class for client-side REST tests, configures the RestTemplate with a customClientHttpRequestFactory that asserts actual requests against expectations and returns "stub" responses. In this case we expect a request to "/greeting" and want to return a 200 response with "text/plain" content. We could define as additional expected requests and stub responses as needed. When expected requests and stub responses are defined, the RestTemplate can be used in client-side code as usual. At the end of testing mockServer.verify() can be used to verify that all expectations have been satisfied.

By default requests are expected in the order in which expectations were declared. You can set the ignoreExpectOrder option when building the server in which case all expectations are checked (in order) to find a match for a given request. That means requests are allowed to come in any order. Here is an example:

server = MockRestServiceServer.bindTo(restTemplate).ignoreExpectOrder(true).build();

Even with unordered requests by default each request is allowed to execute once only. The expect method provides an overloaded variant that accepts anExpectedCount argument that specifies a count range, e.g. once, manyTimes, max, min, between, and so on. Here is an example:

RestTemplate restTemplate = new RestTemplate();

MockRestServiceServer mockServer = MockRestServiceServer.bindTo(restTemplate).build();
mockServer.expect(times(2), requestTo("/foo")).andRespond(withSuccess());
mockServer.expect(times(3), requestTo("/bar")).andRespond(withSuccess());

// ...

mockServer.verify();

Note that when ignoreExpectOrder is not set (the default), and therefore requests are expected in order of declaration, then that order only applies to the first of any expected request. For example if "/foo" is expected 2 times followed by "/bar" 3 times, then there should be a request to "/foo" before there is a request to "/bar" but aside from that subsequent "/foo" and "/bar" requests can come at any time.

As an alternative to all of the above the client-side test support also provides a ClientHttpRequestFactory implementation that can be configured into aRestTemplate to bind it to a MockMvc instance. That allows processing requests using actual server-side logic but without running a server. Here is an example:

MockMvc mockMvc = MockMvcBuilders.webAppContextSetup(this.wac).build();
this.restTemplate = new RestTemplate(new MockMvcClientHttpRequestFactory(mockMvc));

// Test code that uses the above RestTemplate ...

mockServer.verify();

Static Imports

Just like with server-side tests, the fluent API for client-side tests requires a few static imports. Those are easy to find by searching "MockRest*". Eclipse users should add "MockRestRequestMatchers.*" and "MockRestResponseCreators.*" as "favorite static members" in the Eclipse preferences under Java → Editor → Content Assist → Favorites. That allows using content assist after typing the first character of the static method name. Other IDEs (e.g. IntelliJ) may not require any additional configuration. Just check the support for code completion on static members.

Further Examples of Client-side REST Tests

Spring MVC Test’s own tests include example tests of client-side REST tests.

15.7 PetClinic Example

The PetClinic application, available on GitHub, illustrates several features of the Spring TestContext Framework in a JUnit 4 environment. Most test functionality is included in the AbstractClinicTests, for which a partial listing is shown below:

import static org.junit.Assert.assertEquals;
// import ...

@ContextConfiguration
public abstract class AbstractClinicTests extends AbstractTransactionalJUnit4SpringContextTests {

    @Autowired
    protected Clinic clinic;

    @Test
    public void getVets() {
        Collection<Vet> vets = this.clinic.getVets();
        assertEquals("JDBC query must show the same number of vets",
            super.countRowsInTable("VETS"), vets.size());
        Vet v1 = EntityUtils.getById(vets, Vet.class, 2);
        assertEquals("Leary", v1.getLastName());
        assertEquals(1, v1.getNrOfSpecialties());
        assertEquals("radiology", (v1.getSpecialties().get(0)).getName());
        // ...
    }

    // ...
}

Notes:

  • This test case extends the AbstractTransactionalJUnit4SpringContextTests class, from which it inherits configuration for Dependency Injection (through theDependencyInjectionTestExecutionListener) and transactional behavior (through the TransactionalTestExecutionListener).
  • The clinic instance variable — the application object being tested — is set by Dependency Injection through @Autowired semantics.
  • The getVets() method illustrates how you can use the inherited countRowsInTable() method to easily verify the number of rows in a given table, thus verifying correct behavior of the application code being tested. This allows for stronger tests and lessens dependency on the exact test data. For example, you can add additional rows in the database without breaking tests.
  • Like many integration tests that use a database, most of the tests in AbstractClinicTests depend on a minimum amount of data already in the database before the test cases run. Alternatively, you might choose to populate the database within the test fixture set up of your test cases — again, within the same transaction as the tests.

The PetClinic application supports three data access technologies: JDBC, Hibernate, and JPA. By declaring @ContextConfiguration without any specific resource locations, the AbstractClinicTests class will have its application context loaded from the default location, AbstractClinicTests-context.xml, which declares a common DataSource. Subclasses specify additional context locations that must declare a PlatformTransactionManager and a concrete implementation of Clinic.

For example, the Hibernate implementation of the PetClinic tests contains the following implementation. For this example, HibernateClinicTests does not contain a single line of code: we only need to declare @ContextConfiguration, and the tests are inherited from AbstractClinicTests. Because @ContextConfiguration is declared without any specific resource locations, the Spring TestContext Framework loads an application context from all the beans defined inAbstractClinicTests-context.xml (i.e., the inherited locations) and HibernateClinicTests-context.xml, with HibernateClinicTests-context.xmlpossibly overriding beans defined in AbstractClinicTests-context.xml.

@ContextConfiguration
public class HibernateClinicTests extends AbstractClinicTests { }

In a large-scale application, the Spring configuration is often split across multiple files. Consequently, configuration locations are typically specified in a common base class for all application-specific integration tests. Such a base class may also add useful instance variables — populated by Dependency Injection, naturally — such as aSessionFactory in the case of an application using Hibernate.

As far as possible, you should have exactly the same Spring configuration files in your integration tests as in the deployed environment. One likely point of difference concerns database connection pooling and transaction infrastructure. If you are deploying to a full-blown application server, you will probably use its connection pool (available through JNDI) and JTA implementation. Thus in production you will use a JndiObjectFactoryBean or <jee:jndi-lookup> for the DataSource andJtaTransactionManager. JNDI and JTA will not be available in out-of-container integration tests, so you should use a combination like the Commons DBCPBasicDataSource and DataSourceTransactionManager or HibernateTransactionManager for them. You can factor out this variant behavior into a single XML file, having the choice between application server and a 'local' configuration separated from all other configuration, which will not vary between the test and production environments. In addition, it is advisable to use properties files for connection settings. See the PetClinic application for an example.

16. Further Resources

Consult the following resources for more information about testing:

  • JUnit: "A programmer-oriented testing framework for Java". Used by the Spring Framework in its test suite.
  • TestNG: A testing framework inspired by JUnit with added support for annotations, test groups, data-driven testing, distributed testing, etc.
  • AssertJ: "Fluent assertions for Java" including support for Java 8 lambdas, streams, etc.
  • Mock Objects: Article in Wikipedia.
  • MockObjects.com: Web site dedicated to mock objects, a technique for improving the design of code within test-driven development.
  • Mockito: Java mock library based on the test spy pattern.
  • EasyMock: Java library "that provides Mock Objects for interfaces (and objects through the class extension) by generating them on the fly using Java’s proxy mechanism." Used by the Spring Framework in its test suite.
  • JMock: Library that supports test-driven development of Java code with mock objects.
  • DbUnit: JUnit extension (also usable with Ant and Maven) targeted for database-driven projects that, among other things, puts your database into a known state between test runs.
  • The Grinder: Java load testing framework.

Part V. Data Access

 

This part of the reference documentation is concerned with data access and the interaction between the data access layer and the business or service layer.

Spring’s comprehensive transaction management support is covered in some detail, followed by thorough coverage of the various data access frameworks and technologies that the Spring Framework integrates with.

17. Transaction Management

17.1 Introduction to Spring Framework transaction management

Comprehensive transaction support is among the most compelling reasons to use the Spring Framework. The Spring Framework provides a consistent abstraction for transaction management that delivers the following benefits:

  • Consistent programming model across different transaction APIs such as Java Transaction API (JTA), JDBC, Hibernate, Java Persistence API (JPA), and Java Data Objects (JDO).
  • Support for declarative transaction management.
  • Simpler API for programmatic transaction management than complex transaction APIs such as JTA.
  • Excellent integration with Spring’s data access abstractions.

The following sections describe the Spring Framework’s transaction value-adds and technologies. (The chapter also includes discussions of best practices, application server integration, and solutions to common problems.)

17.2 Advantages of the Spring Framework’s transaction support model

Traditionally, Java EE developers have had two choices for transaction management: global or local transactions, both of which have profound limitations. Global and local transaction management is reviewed in the next two sections, followed by a discussion of how the Spring Framework’s transaction management support addresses the limitations of the global and local transaction models.

17.2.1 Global transactions

Global transactions enable you to work with multiple transactional resources, typically relational databases and message queues. The application server manages global transactions through the JTA, which is a cumbersome API to use (partly due to its exception model). Furthermore, a JTA UserTransaction normally needs to be sourced from JNDI, meaning that you also need to use JNDI in order to use JTA. Obviously the use of global transactions would limit any potential reuse of application code, as JTA is normally only available in an application server environment.

Previously, the preferred way to use global transactions was via EJB CMT (Container Managed Transaction): CMT is a form of declarative transaction management (as distinguished from programmatic transaction management). EJB CMT removes the need for transaction-related JNDI lookups, although of course the use of EJB itself necessitates the use of JNDI. It removes most but not all of the need to write Java code to control transactions. The significant downside is that CMT is tied to JTA and an application server environment. Also, it is only available if one chooses to implement busin

相关文章：

猜你喜欢
相关资源
相似解决方案
热门标签
Java Python linux javascript Mysql C# Docker 算法 前端 SpringBoot Redis Vue spring 设计模式 .net core .net kubernetes c++ 数据库 数据结构 大数据 js 机器学习 微服务 Android Go 程序员 面试 JVM ASP.net core 云原生 人工智能 后端 PHP git CSS golang k8s Nginx Django mybatis 深度学习 多线程 React 架构 devops 爬虫 云计算 Spring Boot LeetCode