Horton

Re: bug in stored generated column over domain with constraints.

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

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: jian he <jian.universality@gmail.com>
Cc: PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2025-04-15T16:13:49Z
Lists: pgsql-hackers 
jian he <jian.universality@gmail.com> writes:
> Thanks for simplifying the tests, overall all looks good.

OK, pushed and back-patched, but only to v14.  I tried to make
a variant that'd work in v13, but it caused ExecCheckPlanOutput
to throw "Query provides a value for a generated column" errors
for UPDATEs.  That's surely related to the major refactoring of
UPDATE targetlists that we did in v14.  Given that we've had
zero field complaints about this problem so far, I think the
risk of breaking something in v13 exceeds the value of fixing
the bug, so I left v13 alone.

			regards, tom lane

Commits

  1. Fix failure for generated column with a not-null domain constraint.

  2. Handle default NULL insertion a little better.