Horton

Re: generating function default settings from pg_proc.dat

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
Cc: Andrew Dunstan <andrew@dunslane.net>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2026-02-17T18:41:49Z
Lists: pgsql-hackers

Attachments

Here's the promised v2, which addresses all your review comments.

With respect to the list of supported types in bki.sgml: I wonder if
we should just drop that, because it evidently hasn't been maintained
well.  It wasn't at all in sync with the actual contents of TypInfo[].
I made it be so, but ...

Poking further at that, I found that there were a lot of TypInfo[]
entries that were not actually used and seem to have just been
cargo-culted in.  So this patch removes all the ones that aren't
demonstrably necessary to get through initdb.  Maybe that's too
aggressive, but in view of the potential for maintenance errors
(cf 7cdb633c8) I don't think we should be carrying unused entries
there.

			regards, tom lane

Commits

  1. Simplify creation of built-in functions with non-default ACLs.

  2. Be more wary of false matches in initdb's replace_token().

  3. Simplify creation of built-in functions with default arguments.