Re: Regression in statement locations

On Mon, May 19, 2025 at 02:31:24PM +0000, David Steele wrote:
> But in PG18 we now get:
>
> NOTICE: AUDIT: SESSION,35,2,DDL,CREATE TABLE,TABLE,public.do_table,"CREATE
> TABLE do_table (""weird name"" INT)",<none>
> NOTICE: AUDIT: SESSION,35,3,DDL,DROP TABLE,TABLE,public.do_table,"CREATE
> TABLE do_table (""weird name"" INT)
>
> ; DROP table do_table",<none>
>
> The create table statement is fine but the drop table statement has create
> table glued to it. I tried removing the extra quotes, whitespace, etc. from
> our test and the result is the same.

Yes, I've noticed this one while looking at the output generated by
pgaudit through the new logic when I've posted this message:
https://www.postgresql.org/message-id/ZzqXU-2uTWzC38jO@paquier.xyz

If I remember what was on my mind back then (don't have this \r\n case
specifically written down on my notes from last November as far as I
can see now), I was thinking that this new output is expected as this
is the full query string for the nested statement passed down to the
EXECUTE clause done in this PL function. Anthonin, what's your take?

I am adding Jian in CC, who also participated on the thread.
--
Michael