Core Data memory usage while importing large dataset

Question

I\'m now stuck for about two weeks with a nasty Core Data problem. I read lots of blogpost, articles and SO questions/answers but I\'m still not able to solve my problem.

<

Answer 1

I have an import helper that does something very similar.

Have a look at the code below and see if it helps you

__block NSUInteger i = 0;
NSArray *jsonArray = ...
for (NSDictionary *dataStucture in jsonArray)
{
    [managedObjectContext performBlock:^{
        @autoreleasepool {
            i++;
            A *a = (A*)[self newManagedObjectOfType:@"A" inManagedObjectContext:managedObjectContext];
            [self parseData:[dataStucture objectForKey:@"a"]
                 intoObject:a
     inManagedObjectContext:managedObjectContext];

            [managedObjectContext refreshObject:a
                                   mergeChanges:YES];
            if (i > 20) // Arbitrary number here
            {
                NSError *error = nil;
                [managedObjectContext save:&error];
                [managedObjectContext reset];
            }

            [managedObjectContext refreshObject:a
                                   mergeChanges:YES];

        }
        dispatch_async(dispatch_get_main_queue(), ^{
            [self saveMainThreadManagedObjectContext];

            NSLog(@"DONE");
            // parsing is done, now you see that there are still
            // A's, B's and C's left in memory.
            // Every managedObjectContext is saved and no references are kept
            // to any A, B and C so they should be released. This is not true,
            // so a managedObjectContext is keeping a strong reference to these
            // objects.
        });
    }];
}