QOr doesn't handle two empty ResultSets correctly

[Honza_Kral]

>>> from django.db.models import Q
>>> from django.contrib.auth.models import User
>>> User.objects.filter( Q( id__in=[] ) | Q( id__in=[] ) )

Should be empty queryset, but is not

whe generated WHERE clause is empty, the problem seems to be around
django/db/models/query.py:725

[raphael]

still an issue in this case :

User.objects.filter( Q( idin=[]) | Q( idin=[1,]) | Q(idin=[] ) )

[Honza_Kral]

my take on the patch

[Honza_Kral]

fixes the same problem for queryset-refactor branch

[raphael]

Remove performance penalty issue

[mtredinnick]

Any patches against query creation for trunk aren't likely to survive at the moment, since queryset-refactor is very advanced. I have a feeling this is going to be fixed naturally in that branch (there are some known problems with disjunctive joins at the moment, but they'll be fixed). I'll close this once I've verified it's fixed there and the branch is merged.

[Honza_Kral]

Replying to mtredinnick:

Any patches against query creation for trunk aren't likely to survive at the moment, since queryset-refactor is very advanced. I have a feeling this is going to be fixed naturally in that branch (there are some known problems with disjunctive joins at the moment, but they'll be fixed). I'll close this once I've verified it's fixed there and the branch is merged.

One of the patches fixes the problem in qs-rf branch, where the problem exists as well. ( 6074-queryset-refactor.diff )

[mtredinnick]

Unfortunately, that doesn't fix it properly in the queryset-refactor branch. There is some more general work required to make disjunctions work correctly there (which is why three tests are currently failing in that branch: to remind me to fix that properly). Thanks for trying, anyway. It's just a bit trickier than it looks because this is a special case of something more general.

[Honza_Kral]

Replying to mtredinnick:

Unfortunately, that doesn't fix it properly in the queryset-refactor branch. There is some more general work required to make disjunctions work correctly there (which is why three tests are currently failing in that branch: to remind me to fix that properly). Thanks for trying, anyway. It's just a bit trickier than it looks because this is a special case of something more general.

I didn't presume to fix the whole QOr behavior in qs-rf branch - the patch I supplied merely fixed this little issue (the tests attached failed without it and passes with it).

[mtredinnick]

(In [6868]) queryset-refactor: Fixed disjunctions of empty result sets. Refs #6074.

[mtredinnick]

(In [7477]) Merged the queryset-refactor branch into trunk.

This is a big internal change, but mostly backwards compatible with existing
code. Also adds a couple of new features.

Fixed #245, #1050, #1656, #1801, #2076, #2091, #2150, #2253, #2306, #2400, #2430, #2482, #2496, #2676, #2737, #2874, #2902, #2939, #3037, #3141, #3288, #3440, #3592, #3739, #4088, #4260, #4289, #4306, #4358, #4464, #4510, #4858, #5012, #5020, #5261, #5295, #5321, #5324, #5325, #5555, #5707, #5796, #5817, #5987, #6018, #6074, #6088, #6154, #6177, #6180, #6203, #6658