| From: | Henson Choi <assam258(at)gmail(dot)com> |
|---|---|
| To: | Sho Ito <i(dot)sho(dot)0628(at)gmail(dot)com> |
| Cc: | Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com>, pgsql-hackers(at)lists(dot)postgresql(dot)org |
| Subject: | Re: [PATCH v1] Preserve implicit composite casts in ROW deparsing |
| Date: | 2026-11-21 12:11:44 |
| Message-ID: | CAAAe_zBze+MYfU9rYr8HszJ_JYcxQTHM4oq=EwNorkiKzu+X=g@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi Sho,
> > However, I think the comment might be more useful if it focused on
> > explaining why RECORDOID can be excluded even when showimplicit is
> > true.
>
> Thanks for reviewing. Attached v3 revises the comment to explain that
> a bare `ROW()` already has type `RECORD`, so omitting an implicit
> cast to `RECORD` does not affect overload resolution on reparse.
I agree with Horiguchi-san: leaving out RECORDOID is the part of the
condition that is not obvious, so that is what the comment should
explain. The v3 wording reads well to me.
> There are no code or test changes from v2.
I tested v3 and found no problems. The added VARIADIC and column
default test cases look good to me.
Best regards,
Henson
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Zhijie Hou | 2026-11-21 12:18:45 | Re: Stop masking visibility information during WAL consistency checking |
| Previous Message | Álvaro Herrera | 2026-11-21 11:31:25 | Re: pgsql: Don't trust NOT NULL on virtual generated columns of partitioned |