Discussion

Discussion

This method is functionally equivalent to the following example code:

 NSEntityDescription *entity = [[managedObjectStore.managedObjectModel entitiesByName] objectForKey:entityName];
 return [RKEntityMapping mappingForEntity:entity];

Discussion

Default: YES

Discussion

This is a convenience method for flexibly adding a connection to the receiver. The relationship can be specified with by providing an NSRelationshipDescription object or an NSString specifying the name of the relationship to be connected. The connection specifier can be provided as an NSString object, an NSArray of NSString object, or an NSDictionary with NSString objects for the keys and values. The string objects specify the name of attributes within the entity and destination entity of the specified relationship. The RKConnectionDescription class models a connection as a dictionary in which the keys are NSString objects corresponding to the names of attributes in the source entity of the relationship being connected and the values are NSString objects corresponding to the names of attributes in the destination entity.

When the connectionSpecifier is an NSString value, it is interpretted as the name of an attribute in the specified relationship’s entity. The corresponding attribute in the destination entity is determined by invoking the source to destination key transformation block set via [RKObjectMapping setSourceToDestinationKeyTransformationBlock:]. If no transformation block is configured, then the destination attribute is assumed to have the same name as the source attribute. For example, consider a model in which there entities named ‘User’ and ‘Project’. The ‘User’ entity has a to-many relationship to the ‘Project’ entity named ‘projects’. Both the ‘User’ and the ‘Project’ entities contain an attribute named ‘userID’, which represents the value for the primary key of the ‘User’ in the API the application is communicating with. When the user’s projects are loaded from the ‘/projects’ endpoint, the user ID is sent down in the JSON representation of the Project objects as a numeric value. In order to establish a connection for the ‘projects’ relationship between the ‘User’ and ‘Project’ entities, we could add the connection like so:

// JSON looks like {"project": { "name": "Project Name", "userID": 1234, "projectID": 1 } }
RKEntityMapping *mapping = [RKEntityMapping mappingForEntityForName:@"Project" inManagedObjectStore:managedObjectStore];
[mapping addAttributeMappings:@[ @"name", @"userID", @"projectID" ]];
// Find a 'User' whose value for the 'userID' object is equal to the value stored on the 'userID' attribute of the 'Project' and assign it to the relationship
// In other words, "Find the User whose userID == 1234 and assign that object to the 'user' relationship"
[mapping addConnectionForRelationship:@"user" connectedBy:@"userID"];

When the connection is attempted to be established by an instance of RKRelationshipConnectionOperation, the value for the ‘userID’ attribute will be read from the Project (in this case, @1234) and the managed object context will be searched for a managed object for the ‘User’ entity with a corresponding value for its ‘userID’ attribute.

When the connectionSpecifier is an NSArray object, it is interpretted as containing the names of attributes in the specified relationship’s entity. Just as with a stirng value, the corresponding destination attributes are determined by invoking the source to destination key transformation block or they are assumed to have matching names.

When the connectionSpecifier is an NSDictionary object, the keys are interpretted as containing the names of attributes on the source entity the values are interpretted as the names of attributes on the destination entity. For example:

// Find the User whose userID is equal to the value stored on the 'createdByUserID' attribute
[mapping addConnectionForRelationship:@"createdByUser" connectedBy:@{ @"createdByUserID": @"userID" }];