Re: [PATCH] Add pretty formatting to pg_get_triggerdef

Philip Alger <paalger0@gmail.com>

From: Philip Alger <paalger0@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Chao Li <li.evan.chao@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2025-11-06T20:19:29Z
Lists: pgsql-hackers
Hi Tom,



> This patch will cause psql's \d to fail hard against any pre-v19
> server.  That's not acceptable.  Note the comment at the top of
> describe.c:
>
>  * Support for the various \d ("describe") commands.  Note that the current
>  * expectation is that all functions in this file will succeed when working
>  * with servers of versions 9.2 and up.  It's okay to omit irrelevant
>  * information for an old server, but not to fail outright.  (But failing
>  * against a pre-9.2 server is allowed.)
>
>
I didn't see that at the top of the file, and didn't think the new function
would break it all.


> You could get around that with a conditional on the server version,
> as is done elsewhere in describe.c.  But I kind of wonder why we
> need this new function at all.  Isn't "pg_get_triggerdef(oid, false)"
> close enough?


No. That works to print out the CREATE TRIGGER statement with the
schema.relation name. I am trying to make `pretty` print similar to what is
in pg_get_viewdef and pg_get_functiondef which formats the output.


> An alternative suggestion is to keep "pg_get_triggerdef(oid, true)"
> doing exactly what it does today and have the newly-introduced
> function be the one that has new behavior.  This might be a smarter
> plan anyway to avoid side-swiping other clients that are expecting
> the current behavior.
>

Thinking aloud, I would need to create a new function like
pg_get_triggerdef_worker and let it handle the formatting.

-- 
Best,
Phil Alger
EDB: https://www.enterprisedb.com