从 json 插入核心数据模型

Question

我有一项服务可以获取使用以下 json 响应的员工：

[{ "initials": "PR", "name": "Paul", "accounts": ["SPA", "ITA"] },  
{ "initials": "SR", "name": "William", "accounts": ["ITA"] } ]

和另一个服务来获取导致这样的 json 的帐户：

[{ "code": "SPA", "name": "Spain", "employees": ["PR"]  },
 { "code": "ITA", "name": "Italy", "employees": ["PR", "SR"]  } ]

我需要解析接收到的信息并将其插入到核心数据数据库中，该数据库具有两个核心数据模型，员工和帐户，声明了多对多的关系。它们声明如下：

@interface Employee : NSManagedObject

    @property (nonatomic, retain) NSString *initials;
    @property (nonatomic, retain) NSString *name;
    @property (nonatomic, retain) NSSet *accounts;

@end

@interface Account : NSManagedObject

    @property (nonatomic, retain) NSString *code;
    @property (nonatomic, retain) NSString *name;
    @property (nonatomic, retain) NSSet *employees;

@end

我现在做的是先插入员工，然后创建每个帐户并查询要插入每个帐户的员工。显然，这样做的性能非常糟糕，因为我需要为每个帐户做一个请求。

- (void)loadEmployeesJSON { ... }
- (void)loadAccountsJSON {
    NSManagedObjectContext *context = self.managedObjectContext;

    NSError *err = nil;
    NSString *dataPath = [[NSBundle mainBundle] pathForResource:@"Accounts" ofType:@"json"];
    NSArray *accounts = [NSJSONSerialization JSONObjectWithData:[NSData dataWithContentsOfFile:dataPath] options:kNilOptions error:&err];

    [accounts enumerateObjectsUsingBlock:^(NSDictionary *obj, NSUInteger idx, BOOL *stop) {
        Account *account = [NSEntityDescription insertNewObjectForEntityForName:@"Account" inManagedObjectContext:context];
        account.code = [obj objectForKey:@"code"];
        account.name = [obj objectForKey:@"name"];
        account.employees = [self accountsForEmployee:[obj objectForKey:@"employees"]];
        NSError *error;
        if (![context save:&error]) {
            NSLog(@"Whoops, couldn't save: %@", [error localizedDescription]);
        }
    }];
}

- (NSSet*)accountsForEmployee:(NSArray*)initialsArray {
    NSManagedObjectContext *context = self.managedObjectContext;
    NSError *error = nil;

    NSFetchRequest *fetchRequest = [[NSFetchRequest alloc] init];
    NSEntityDescription *entity = [NSEntityDescription entityForName:@"Employee" inManagedObjectContext:context];
    [fetchRequest setPredicate:[NSPredicate predicateWithFormat:@"initials IN %@", initialsArray]];
    [fetchRequest setEntity:entity];
    NSArray *fetchedObjects = [context executeFetchRequest:fetchRequest error:&error];
    return [NSSet setWithArray:fetchedObjects];
}

如何在不要求每个帐户的员工的情况下链接它们？ 我认为我可以在内存中映射所有员工并要求他们。但是，如果我的员工数据库太大（+100.000）而无法存储在内存中怎么办？

Answer 1

无论大小有多大，您都可以将所有员工都保存在内存中。为此，在您创建员工实体并保存它（最好是批量保存）后，通过以下方式将它们重置为故障：

NSManagedObjectContext *moc = ...;
NSManagedObject *employee = ...;
[moc refreshObject:employee mergeChanges:NO];

通过在保存后调用-refreshObject:mergeChanges:，它将将该对象的属性和关系从内存中删除，并将其内存占用减少到接近于零。从那里您可以将其存储在NSDictionary 中以供以后进行关系映射。

这将避免所有的提取并显着提高您的导入性能。