#30771: Filtering on query result overrides GROUP BY of internal query
-------------------------------------+-------------------------------------
     Reporter:  Aur Saraf            |                    Owner:  James
                                     |  Timmins
         Type:  Bug                  |                   Status:  assigned
    Component:  Database layer       |                  Version:  master
  (models, ORM)                      |
     Severity:  Normal               |               Resolution:
     Keywords:  annotate,            |             Triage Stage:  Accepted
  annotation, filter, groupby,       |
  aggregate, aggregation             |
    Has patch:  0                    |      Needs documentation:  0
  Needs tests:  0                    |  Patch needs improvement:  0
Easy pickings:  0                    |                    UI/UX:  0
-------------------------------------+-------------------------------------

Comment (by James Timmins):

 Thanks so much for the help Simon! This is a great jumping-off point.

 There's something that I'm unclear about, which perhaps you can shed some
 light on. While I was able to replicate the bug with 2.2, when I try to
 create a test on Master to validate the bug, the group-by behavior seems
 to have changed.

 Here's the test that I created:

 {{{
     def test_exact_selected_field_rhs_subquery(self):
         author_1 = Author.objects.create(name='one')
         author_2 = Author.objects.create(name='two')
         max_ids =
 
Author.objects.filter(alias__isnull=True).values('alias').annotate(m=Max('id')).values('m')
         authors = Author.objects.filter(id=max_ids[:1])

         self.assertFalse(str(max_ids.query)) # This was just to force the
 test-runner to output the query.

         self.assertEqual(authors[0], author_2)
 }}}

 And here's the resulting query:

 {{{
 SELECT MAX("lookup_author"."id") AS "m" FROM "lookup_author" WHERE
 "lookup_author"."alias" IS NULL GROUP BY "lookup_author"."alias",
 "lookup_author"."name"
 }}}

 It no longer appears to be grouping by the 'alias' field listed in the
 initial `.values()` preceeding the `.annotate()`. I looked at the docs and
 release notes to see if there was a behavior change, but didn't see
 anything listed.

 Do you know if I'm just misunderstanding what's happening here? Or does
 this seem like a possible regression?

-- 
Ticket URL: <https://code.djangoproject.com/ticket/30771#comment:5>
Django <https://code.djangoproject.com/>
The Web framework for perfectionists with deadlines.

-- 
You received this message because you are subscribed to the Google Groups 
"Django updates" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to django-updates+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/django-updates/068.b6ebec92d059fbf92e3798b2115a251a%40djangoproject.com.

Reply via email to