| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Sami Imseih <samimseih(at)gmail(dot)com> |
| Cc: | "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Controlling the usage of a user-defined cast |
| Date: | 2024-12-12 16:51:37 |
| Message-ID: | 659109.1734022297@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Sami Imseih <samimseih(at)gmail(dot)com> writes:
> When a user-defined CAST is created, it has the ability to break behavior of built-in
> casts that could be performed implicitly, i.e. without a cast defined in pg_cast.
The requirement for ownership of at least one type means that the
example you give could only be done by a superuser. I think there
could be an argument for requiring ownership of *both* types, but
perhaps that would break some useful cases.
> There is currently no way to prevent the usage of a user-defined cast. Should there be one?
I don't think so, and I don't see any reasonable way to do it.
You will get nowhere proposing a GUC that changes query semantics ---
we learned that that was a bad idea decades ago.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | David G. Johnston | 2024-12-12 16:54:21 | Re: Document NULL |
| Previous Message | Tom Lane | 2024-12-12 16:45:01 | Re: Allow subfield references without parentheses |