Question

I have a problem with Core Data in my iOS app.

First off, the code for my ManagedObject:

@interface MyBook : NSManagedObject
  @property (retain, nonatomic) NSString *book_name;
@end


@implementation MyBook    
  @synthesize book_name;
@end

And the code sample:

NSFetchRequest *fr = [NSFetchRequest fetchRequestWithEntityName:@"MyBook"];
fr.predicate = [NSPredicate predicateWithFormat:@"book_name == %@", @"Something"];

NSArray *result = [s.managedObjectContext executeFetchRequest:fr error:&error];

MyBook *book = nil;
if (result.count) {
    book = [result objectAtIndex:0];
    NSLog(@"Updating old book: %@", book);
} else {
    book = [NSEntityDescription insertNewObjectForEntityForName:@"Book"
               inManagedObjectContext:s.managedObjectContext];
    NSLog(@"Creating new book");
    shelf.book_name = // some unique book name is set here
}

NSLog(@"result: %@", book.book_name);

The code basically…

• Tries to look up a book by name and grab that.
• If the book does not exist yet, create one and set the book name.
  • Otherwise, grab the existing book.

The weird thing I’m experiencing is that when the book already exists, the property book_name is null. If the object was created, book_name will have the string I set before.

I guess access to the property is not causing a faulting on the fetched book instance. That’s probably why [book valueForKey:@"book_name"] works and after that the property has been populated as well.

What am I missing here? I’m expecting Core Data to notice that I’m accessing the property and internally grabbing the data for me transparently.

Thank you

Answer 1

Big WHOOPS!

Just after posting this original question, I realized that I was using @synthesize instead of @dynamic.

However, I’ll leave this question here, if anyone else experiences such an effect in the future. 🙂