{{{
Book.objects
.annotate(
labels=ArrayAgg(
"reviews__labels",
filter=Q(reviews__labels__isnull=False),
distinct=True,
)
)
.annotate(
fav_count=Subquery(
FavoriteBook.objects.filter(book=OuterRef("pk"))
.annotate(count=Count("id"))
.values("count")[:1],
output_field=IntegerField(),
)
)
.values('id')
}}}
This crashes with
{{{
django.db.utils.ProgrammingError: column "fav_count" does not exist
LINE 1: ..._labels"."review_id") GROUP BY "myapp_book"."id", "fav_count...
^
}}}
(Disregard that it makes no sense to do annotations if we're only using
"id" values - it's for the purpose of this example)
Asking for `.values('fav_count')` works though. I've compiled a small
example project that contains all the models and a test case that
reproduces this issue.
--
Ticket URL: <https://code.djangoproject.com/ticket/31152>
Django <https://code.djangoproject.com/>
The Web framework for perfectionists with deadlines.
* Attachment "proj.zip" added.
Project
* status: new => closed
* resolution: => duplicate
* component: Uncategorized => Database layer (models, ORM)
Comment:
Duplicate of #31136.
--
Ticket URL: <https://code.djangoproject.com/ticket/31152#comment:1>