Dean Rasheed <dean.a.rasheed@gmail.com> writes:
> On 2 July 2013 09:30, Dean Rasheed <dean.a.rasheed@gmail.com> wrote:
>> I think the rewritten query should only use inheritance if inheritance
>> was requested in the original query, *and* if inheritance was enabled
>> in the view's query, per attached patch against HEAD.
> On second thoughts, I think this should only apply to INSERT.
> UPDATE and DELETE should continue work the same as SELECT, respecting
> the inheritance flag from the view query regardless of the inheritance
> flag in the outer query.
[ thinks for a bit... ] Yeah, I agree.
> Attached is an updated patch for HEAD, with regression tests. This
> should also be applied to the 9.3beta branch.
Applied with minor adjustments --- mainly, I took out the inFromCl
twiddling, which AFAICS is neither necessary (nothing downstream of this
looks at inFromCl) nor clearly correct.
regards, tom lane