On Sat, Mar 3, 2012 at 10:03 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Claus Stadler <cstadler@informatik.uni-leipzig.de> writes:
>> Query optimizer glitch: "...WHERE TRUE" condition in union results in
>> bad query plan ...
>
> Yeah, this is because a nonempty WHERE clause defeats simplifying the
> UNION ALL into a simple "append relation" (cf is_safe_append_member()).
> The planner will eventually figure out that WHERE TRUE is a no-op,
> but that doesn't happen till later (and there are good reasons to do
> things in that order).
>
> Sooner or later I'd like to relax the restriction that appendrel members
> can't have extra WHERE clauses, but don't hold your breath waiting...
Does this comment need updating?
* Note: the data structure assumes that append-rel members are single
* baserels. This is OK for inheritance, but it prevents us from pulling
* up a UNION ALL member subquery if it contains a join. While that could
* be fixed with a more complex data structure, at present there's not much
* point because no improvement in the plan could result.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company