Re: Split index and table statistics into different types of stats

Hi,

On 2022-11-18 12:18:38 +0100, Drouvot, Bertrand wrote:
> On 11/16/22 9:12 PM, Andres Freund wrote:
> > This still leaves a fair bit of boilerplate. ISTM that the function body
> > really should just be a single line.
> >
> > Might even be worth defining the whole function via a macro. Perhaps something like
> >
> > PGSTAT_DEFINE_REL_FIELD_ACCESSOR(PGSTAT_KIND_INDEX, pg_stat_get_index, numscans);
>
> Thanks for the feedback!
>
> Right, what about something like the following?
>
> "
> #define PGSTAT_FETCH_STAT_ENTRY(pgstat_entry_kind, pgstat_fetch_stat_function, relid, stat_name) \
> do { \
> pgstat_entry_kind *entry = pgstat_fetch_stat_function(relid); \
> PG_RETURN_INT64(entry == NULL ? 0 : (int64) (entry->stat_name)); \
> } while (0)
>
> Datum
> pg_stat_get_index_numscans(PG_FUNCTION_ARGS)
> {
> PGSTAT_FETCH_STAT_ENTRY(PgStat_StatIndEntry, pgstat_fetch_stat_indentry, PG_GETARG_OID(0), numscans);
> }
> "

That's better, but still seems like quite a bit of repetition, given the
number of accessors. I think I like my idea of a macro defining the whole
function a bit better.

I'd define a "base" macro and then a version that's specific to tables and
indexes each, so that the pieces related to that don't have to be repeated as
often.

> > This should probably be done in a preparatory commit.
>
> Proposal submitted in [1].

Now merged.

Greetings,

Andres Freund