Horton

Re: MERGE ... RETURNING

Jeff Davis <pgsql@j-davis.com>

View thread
From: Jeff Davis <pgsql@j-davis.com>
To: Dean Rasheed <dean.a.rasheed@gmail.com>
Cc: Isaac Morland <isaac.morland@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2023-07-13T19:14:52Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Add RETURNING support to MERGE.

  2. doc: Improve a couple of places in the MERGE docs.

  3. doc: improve description of privileges for MERGE and update glossary.

  4. Fix RLS policy usage in MERGE.

  5. Fix leak of LLVM "fatal-on-oom" section counter.

  6. Implement outer-level aggregates to conform to the SQL spec, with

On Thu, 2023-07-13 at 18:01 +0100, Dean Rasheed wrote:
> For some use cases, I can imagine allowing OLD/NEW.colname would mean
> you wouldn't need pg_merge_action() (if the column was NOT NULL), so
> I
> think the features should work well together.

For use cases where a user could do it either way, which would you
expect to be the "typical" way (assuming we supported the new/old)?

  MERGE ... RETURNING pg_merge_action(), id, val;

or

  MERGE ... RETURNING id, OLD.val, NEW.val;

?

I am still bothered that pg_merge_action() is so context-sensitive.
"SELECT pg_merge_action()" by itself doesn't make any sense, but it's
allowed in the v8 patch. We could make that a runtime error, which
would be better, but it feels like it's structurally wrong. This is not
an objection, but it's just making me think harder about alternatives.

Maybe instead of a function it could be a special table reference like:

  MERGE ... RETURNING MERGE.action, MERGE.action_number, id, val?

Regards,
	Jeff Davis