On 2016-02-08 09:41:10 +0000, Dean Rasheed wrote:
> On 7 February 2016 at 23:21, Andrew Gierth <andrew@tao11.riddles.org.uk> wrote:
> > I think for now the following patch is best. This preserves the ability
> > to push degenerate HAVING clauses through an explicit GROUP BY () as
> > well as an implicit one, but prevents pushdown in all other cases.
> >
> > Includes comment and regression test.
> >
>
> Yes, I think that's probably the best solution for now. Longer term I
> think it might be possible to do better, along the lines discussed
> above, but I haven't had time to explore any of those options.
>
> Note that with this change to the "if" condition, the next "else if"
> condition can be simplified back to just:
>
> else if (parse->groupClause)
> {
> /* move it to WHERE */
> ...
FWIW, I find the current phrasing clearer.
> Andres, are you going to take this one?
Working on it as we speak. I thought you might appreciate doing your
first commit under a littlebit less pressure.
Andres