[ZODB-Dev] Re: false write conflicts

[John Belmonte]

Jeremy Hylton wrote:
> On Tue, 2004-03-02 at 11:15, John Belmonte wrote:
> 
>>This wrapper for Persistent.__setattr__ looks like it will suit my needs:
>>
>>   class MyPersistent(Persistent):
>>     def __setattr__(self, key, val):
>>       if not (self.__dict__.has_key(key) and self.__dict__[key] is val):
>>         Persistent.__setattr__(self, key, val)
> 
> 
> My preference would be to avoid a setattr hook at all costs, but that's
> just my preference.
> 
> We've revised the way Persistent objects should implement getattr and
> setattr hooks in ZODB 3.3.  (I'm not sure how it worked in earlier
> versions of ZODB.)  There is no documentation yet, but the tests in
> ZODB.tests.test_overriding_attrs explain it pretty clearly.
> 
> You need to let Persistent run before any user code, because the object
> could be a ghost.  If it's a ghost, then __dict__ will be an empty
> dictionary and the setattr won't do what you expect.  First, you should
> call Persistent._p_setattr(self, key, val).  If it returns True if
> Persistent handled the attribute; for example, if you assigned to
> _p_oid, the Persistent will handle the attribute assignment.

Thanks for the heads-up.  I'm wondering if the ghosted object issue can 
be safely ignored in my case.  The worst my hook can do is call 
Persistent.__setattr__ when it doesn't need to, flagging the object as 
dirty.  However, in the case of what I'm trying to address (in-place 
modification of an attribute object), a __getattr__ will have already 
occurred, so the containing object will normally not be ghosted.

-John


-- 
http:// if  ile.org/