Re: MERGE ... RETURNING

From: walther(at)technowledgy(dot)de
To: Jeff Davis <pgsql(at)j-davis(dot)com>, Dean Rasheed <dean(dot)a(dot)rasheed(at)gmail(dot)com>
Cc: Vik Fearing <vik(at)postgresfriends(dot)org>, Gurjeet Singh <gurjeet(at)singh(dot)im>, Isaac Morland <isaac(dot)morland(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, jian he <jian(dot)universality(at)gmail(dot)com>
Subject: Re: MERGE ... RETURNING
Date: 2024-03-08 08:41:53
Message-ID: 506fc86e-1f6b-415f-86a2-0350a7e7819a@technowledgy.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Jeff Davis:
> To summarize, most of the problem has been in retrieving the action
> (INSERT/UPDATE/DELETE) taken or the WHEN-clause number applied to a
> particular matched row. The reason this is important is because the row
> returned is the old row for a DELETE action, and the new row for an
> INSERT or UPDATE action. Without a way to distinguish the particular
> action, the RETURNING clause returns a mixture of old and new rows,
> which would be hard to use sensibly.

It seems to me that all of this is only a problem, because there is only
one RETURNING clause.

Dean Rasheed wrote in the very first post to this thread:
> I considered allowing a separate RETURNING list at the end of each
> action, but rapidly dismissed that idea. Firstly, it introduces
> shift/reduce conflicts to the grammar. These can be resolved by making
> the "AS" before column aliases non-optional, but that's pretty ugly,
> and there may be a better way. More serious drawbacks are that this
> syntax is much more cumbersome for the end user, having to repeat the
> RETURNING clause several times, and the implementation is likely to be
> pretty complex, so I didn't pursue it.

I can't judge the grammar and complexity issues, but as a potential user
it seems to me to be less complex to have multiple RETURNING clauses,
where I could inject my own constants about the specific actions, than
to have to deal with any of the suggested functions / clauses. More
repetitive, yes - but not more complex.

More importantly, I could add RETURNING to only some of the actions and
not always all at the same time - which seems pretty useful to me.

Best,

Wolfgang

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Sutou Kouhei 2024-03-08 09:17:47 Re: meson: Specify -Wformat as a common warning flag for extensions
Previous Message vignesh C 2024-03-08 08:41:12 Re: speed up a logical replica setup