Re: Planning counters in pg_stat_statements (using pgss_store)

On 2020/05/22 15:10, Andy Fan wrote:
>
>
> On Thu, May 21, 2020 at 3:49 PM Julien Rouhaud <rjuju123@gmail.com <mailto:rjuju123@gmail.com>> wrote:
>
>     Le jeu. 21 mai 2020 à 09:17, Michael Paquier <michael@paquier.xyz <mailto:michael@paquier.xyz>> a écrit :
>
>         On Thu, May 21, 2020 at 08:49:53AM +0200, Julien Rouhaud wrote:
>          > On Tue, May 19, 2020 at 4:29 AM Andy Fan <zhihui.fan1213@gmail.com <mailto:zhihui.fan1213@gmail.com>> wrote:
>          >> Thanks for the excellent extension. I want to add 5 more fields to satisfy the
>          >> following requirements.
>          >>
>          >> int   subplan; /* No. of subplan in this query */
>          >> int   subquery; /* No. of subquery */
>          >> int   joincnt; /* How many relations are joined */
>          >> bool hasagg; /* if we have agg function in this query */
>          >> bool hasgroup; /* has group clause */
>          >
>          > Most of those fields can be computed using the raw sql satements.  Why
>          > not adding functions like query_has_agg(querytext) to get the
>          > information from pgss stored query text instead?
>
>         Yeah I personally find concepts related only to the query string
>         itself not something that needs to be tied to pg_stat_statements.
>         ...
>
>
>     Indeed cte will bring additional concerns about the fields semantics. That's another good reason to go with external functions so you can add extra parameters for that if needed.
>
>
> There are something more we can't get from query string easily. like:
> 1. view involved.   2. subquery are pulled up so there is not subquery
> indeed.  3. sublink are pull-up or become as an InitPlan rather than subPlan.
> 4. joins are removed by remove_useless_joins.

If we can store the plan for each statement, e.g., like pg_store_plans
extension [1] does, rather than such partial information, which would
be enough for your cases?