Re: Behaviour of objectsWithFetchSpecification()
Re: Behaviour of objectsWithFetchSpecification()
- Subject: Re: Behaviour of objectsWithFetchSpecification()
- From: Ken Anderson <email@hidden>
- Date: Thu, 22 Jan 2009 08:52:48 -0500
I hear you - it would be nice if there were a merged view of what's in-
memory vs what's in the database... but there isn't :)
On Jan 22, 2009, at 6:47 AM, Paul Hoadley wrote:
Hi Ken,
On 22/01/2009, at 10:01 PM, Ken Anderson wrote:
Deleting an EO is really just like setting a flag until you save,
so this behavior is really no surprise.
Yeah, it certainly matches exactly with the description in the API
docs, so I was only surprised because I hadn't read them closely
enough.
Actually, though, my surprise wasn't so much that a deleted object
was being returned, but that an object that had been modified in the
EC so that, say, if I hadn't deleted it, but instead called
saveChanges(), it wouldn't have matched the fetchSpec. That is, the
fetchSpec asked for objects where bar() == theBar. I had already
called setBar(null), which is why I wasn't expecting to see that
object in the results. That I had also called deleteObject() was
incidental. Anyway, it _still_ matches the description in the API
docs (the promise is just that the matching EO in the external store
won't overwrite the (now not matching) object in the EC), so the
surprise is entirely self-inflicted.
It doesn't exactly match your needs, but hopefully it will help!
Thanks for that code.
--
Paul.
w http://logicsquad.net/
h http://paul.hoadley.name/
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden