Opened 17 years ago

Last modified 17 years ago

#1435 closed defect

Invalid aggregate query — at Initial Version

Reported by: x@… Owned by: Radek Bartoň
Priority: normal Component: DiscussionPlugin
Severity: normal Keywords: sql
Cc: Trac Release: 0.10

Description

Saw this in my logs:

Apr 6 16:29:53 ops5 postgres[17289]: [2-1] ERROR: column "topic.subject" must appear in the GROUP BY clause or be used in an aggregate function Apr 6 16:29:53 ops5 postgres[17289]: [2-2] STATEMENT: SELECT m.id, m.forum, m.topic, m.time, m.author, m.body, t.subject FROM message m LEFT JOIN (SELECT subject, id FROM Apr 6 16:29:53 ops5 postgres[17289]: [2-3] topic GROUP BY id) t ON t.id = m.topic WHERE body LIKE '%subversion%'

Appears to be coming from:

sql = "SELECT m.id, m.forum, m.topic, m.time, m.author, m.body," \ " t.subject FROM message m LEFT JOIN (SELECT subject, id FROM" \ " topic GROUP BY id) t ON t.id = m.topic WHERE body LIKE '%%%s%%'" \

% (query)

That subquery is not valid usage of a group by; ie, which subject should it be yielding for each id? Using min() or max() might be a reasonable stop gap. However, AFA Postgres is concerned, I think DISTINCT ON () would be the usual way to go. Of course, I'm not sure of DISTINCT ON's portability(dont use mysql or sqlite).

Change History (0)

Note: See TracTickets for help on using tickets.