Kevin Grittner <kgrittn@ymail.com> wrote:
> Michael Paquier <michael.paquier@gmail.com> wrote:
>
>> I am not sure that adding a boolean flag introducing a concept
>> related to matview inside checkRuleResultList is the best
>> approach to solve that. checkRuleResultList is something related
>> only to rules, and has nothing related to matviews in it yet.
>
> Well, I was tempted to keep that concept in DefineQueryRewrite()
> where the call is made, by calling the new bool something like
> requireColumnNameMatch and not having checkRuleResultList()
> continue to use isSelect for that purpose at all.
> DefineQueryRewrite() already references RELKIND_RELATION once,
> RELKIND_MATVIEW twice, and RELKIND_VIEW three times, so it would
> hardly be introducing a new concept there.
Upon reflection, that seemed to be cleaner. Pushed fix that way.
Not much of a change from the previously posted patch, but attached
here in case anyone wants to argue against this approach.
Thanks for the report!
--
Kevin Grittner
EDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
