[Sqlalchemy-tickets] [sqlalchemy] #2845: Query.group_by() ignores arguments

[sqlalchemy <mi...@zz...>]

#2845: Query.group_by() ignores arguments
--------------------+-----------------------------------------
 Reporter:  dwt     |           Owner:  zzzeek
     Type:  defect  |          Status:  new
 Priority:  medium  |       Milestone:
Component:  orm     |        Severity:  no triage selected yet
 Keywords:          |  Progress State:  awaiting triage
--------------------+-----------------------------------------
 Hi there,

 I just spent an hour to debug why group_by was ignored for a query I was
 transforming, so I gather it's worth it to spend some more time to report
 a bug. We are using version 0.7.x so there may have been changes since,
 but I wasn't yet able to invest the time to create a testcase that
 reproduces this.

 Here's what happens. Using the ORM I transform a query like this:

 {{{
         query =
 statistics.track_items_for_topics_in_date_range_query([topic],
 early_cutoff, late_cutoff)
         from sqlalchemy.sql import func
         query = query.with_entities(TrackItem.question_id,
 TrackItem.user_id,
 func.max(TrackItem.creation_time).label('max_creation_time'))
         query = query.group_by(TrackItem.question, TrackItem.user)
 }}}

 The problem/workaround is to not use the objects in the group_by() but
 instead the _id suffixes to tell sqlalchemy that it should group by the
 ids instead.

 Now what happens if you don't know/see this is that the group_by() call is
 just ignored and the query is the exact same query after the call as it
 was before hand.

 Which is, uhm, a bit hard to debug. :)

--
Ticket URL: <http://www.sqlalchemy.org/trac/ticket/2845>
sqlalchemy <http://www.sqlalchemy.org/>
The Database Toolkit for Python