Hi there,
In "Core Data Programming Guide" Apple recommend a way to batch fire fault a
collection of objects, It should be implemented as below:
NSArray * arrayOfFaults = [NSArray arrayWithObjects:object1,object2,nil];
NSPredicate *predicate = [NSPredicate predicateWithFormat:@"self IN %@",
arrayOfFaults];
NSFetchRequest *fetchRequest = [[NSFetchRequest alloc]
initWithEntityName:@"Foo"];
[fetchRequest setPredicate:predicate];
[fetchRequest setReturnsObjectsAsFaults:NO];
[managedObjectContext executeFetchRequest:fetchRequest error:NULL];
However using "Core Data" Instrument on iOS 6 simulator, I found this fetch
almost as same time as other fetch. In other words, it goes to disk and did not
use the cache. It's unreasonable . Since it's an " IN " clause , Core Data
could use the objectID to find object in cache. There are no need to go to the
disk if all of them could be found in cache.
I am pretty sure these objects in arrayOfFaults are in cache. If I replace
above code with this one, not cache miss are reported:
for (NSManagedObject *object in arrayOfFaults){
[object willAccessValueForKey:nil];
}
So my question is : are there any way to batch fault objects while take
advantage of the cache ?
Bob Cromwell
_______________________________________________
Cocoa-dev mailing list ([email protected])
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to [email protected]
