Jackson CSV 映射器不适用于自定义架构

Question

我有这个代码：

CsvSchema sema = CsvSchema.builder()
    .addColumn("name")
    .addColumn("year", CsvSchema.ColumnType.NUMBER)
    .build().withHeader();

ObjectReader reader = new CsvMapper().readerFor(JsonNode.class).with(sema);

JsonNode o = reader.readValue(new FileInputStream(new File("/path/to/test.csv")));
System.out.println(o);

而test.csv 是：

test, year
1,    1

该代码应将CSV 解析为下一种格式的JSON：

{"name":"1","year":1}

但我的输出是：

{"name":"1","year":"1"}

问题是：Jackson解析年份为String，但我在CSV Schema中配置的那一年是Number。有人知道是什么问题吗？

Jackson 版本是2.9.8，我也在2.7.1上试过

Answer 1

经过几个小时的工作，我为您找到了解决方案。

我使用 FlexJson 来配置你的 json 的序列化。

    <!-- https://mvnrepository.com/artifact/net.sf.flexjson/flexjson -->
<dependency>
    <groupId>net.sf.flexjson</groupId>
    <artifactId>flexjson</artifactId>
    <version>2.0</version>
</dependency>

它不是很漂亮，但它确实有效。

我希望这对您有所帮助，我相信您可以改进此代码

public String generateJsonFromCSV(File csvFile, File schemaJson) throws IOException, ClassNotFoundException, NoSuchMethodException, IllegalAccessException, InvocationTargetException, InstantiationException {

    //Get CsvSchema as Map
    Map<String, CsvSchema.ColumnType> map = getSchemaMapFromJson(schemaJson);
    //Create CsvSchema from CsvSchemaMap
    CsvSchema.Builder schemaBuilder = new CsvSchema.Builder();
    map.forEach(schemaBuilder::addColumn);
    CsvSchema schema = schemaBuilder.build();
    //read CSV
    CsvMapper csvMapper = new CsvMapper();
    MappingIterator<Map<?, ?>> mappingIterator = csvMapper.readerFor(Map.class).with(schema).readValues(csvFile);
    //Get configured JsonSerializer from CsvSchemaMap
    JSONSerializer jsonSerializer = getJsonSerializer(map);
    List<Map<?, ?>> lines = mappingIterator.readAll();
    //remove first line
    lines.remove(0);
    //serialize
    return jsonSerializer.deepSerialize(lines);
}

/**
 *
 * @param schemaMap mapping field to ColumnType
 * @return a configured JSONSerializer
 */
private JSONSerializer getJsonSerializer(Map<String, CsvSchema.ColumnType> schemaMap){
    Map<CsvSchema.ColumnType, Transformer> transformerMap = new EnumMap<>(CsvSchema.ColumnType.class);
    transformerMap.put(CsvSchema.ColumnType.STRING, new StringTransformer());
    transformerMap.put(CsvSchema.ColumnType.NUMBER, new NumberTransformer());
    JSONSerializer jsonSerializer = new JSONSerializer();
    for (Map.Entry<String, CsvSchema.ColumnType> columnTypeEntry : schemaMap.entrySet()) {
        jsonSerializer.transform(transformerMap.get(columnTypeEntry.getValue()),columnTypeEntry.getKey());
    }
    return jsonSerializer;
}

/**
 /**
 *
 * @param file JSON CsvSchema
 * @return fieldname ColumnType mapping
 * @throws ClassNotFoundException
 */
private Map<String, CsvSchema.ColumnType> getSchemaMapFromJson(File file) throws ClassNotFoundException {
    Map<String, String> schema = new JSONDeserializer<Map<String,String>>().deserialize(getResourceFileAsString(file.getName()));
    Map<String, CsvSchema.ColumnType> result = new HashMap<>(schema.size());
    for (Map.Entry<String, String> columnSchema : schema.entrySet()) {
        result.put(columnSchema.getKey(), CsvSchema.ColumnType.valueOf(columnSchema.getValue().toUpperCase()));
    }
    return result;
}

输出将是

[{"name":"foobar","year":1986},{"name":"testtest","year":777}]

Answer 2

要强制CsvMapper 使用给定类型，最好的方法是使用POJO。在这种情况下，CsvMapper 知道类型并在可能的情况下自动转换它。让我们创建一个示例数据：

name,year
1,1
2,2
3,2

（在您的示例中，第一列名称是test，但我认为这只是一个错误，应该是name。）

下面的应用展示了如何使用POJO解析CSV和写入JSON：

import com.fasterxml.jackson.annotation.JsonCreator;
import com.fasterxml.jackson.annotation.JsonProperty;
import com.fasterxml.jackson.databind.MappingIterator;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.fasterxml.jackson.databind.ObjectReader;
import com.fasterxml.jackson.databind.SerializationFeature;
import com.fasterxml.jackson.dataformat.csv.CsvMapper;
import com.fasterxml.jackson.dataformat.csv.CsvSchema;

import java.io.File;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;

public class CsvApp {

    public static void main(String[] args) throws IOException {
        File csvFile = new File("./resource/test.csv").getAbsoluteFile();

        CsvMapper csvMapper = new CsvMapper();
        CsvSchema bootstrapSchema = CsvSchema.emptySchema().withHeader();
        ObjectReader reader = csvMapper.readerFor(User.class).with(bootstrapSchema);
        MappingIterator<User> iterator = reader.readValues(csvFile);

        List<User> users = new ArrayList<>();
        iterator.forEachRemaining(users::add);

        System.out.println("Users read from CSV file:");
        users.forEach(System.out::println);

        System.out.println();
        System.out.println("Users in JSON format:");
        ObjectMapper jsonMapper = new ObjectMapper();
        jsonMapper.enable(SerializationFeature.INDENT_OUTPUT);

        System.out.println(jsonMapper.writeValueAsString(users));
    }
}

class User {

    private final String name;
    private final int year;

    @JsonCreator
    public User(@JsonProperty("name") String name, @JsonProperty("year") int year) {
        this.name = name;
        this.year = year;
    }

    public String getName() {
        return name;
    }

    public int getYear() {
        return year;
    }

    @Override
    public String toString() {
        return "User{" +
                "name='" + name + '\'' +
                ", year=" + year +
                '}';
    }
}

上面的代码打印：

Users read from CSV file:
User{name='1', year=1}
User{name='2', year=2}
User{name='3', year=2}

Users in JSON format:
[ {
  "name" : "1",
  "year" : 1
}, {
  "name" : "2",
  "year" : 2
}, {
  "name" : "3",
  "year" : 2
} ]