| From: | "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com> |
|---|---|
| To: | "Vitale, Anthony, Sony Music" <anthony(dot)vitale(at)sonymusic(dot)com> |
| Cc: | PABLO ANDRES IBARRA DUPRAT <Pablo(dot)Ibarra(at)itau(dot)cl>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "pgsql-admin(at)lists(dot)postgresql(dot)org" <pgsql-admin(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Scripting a ALTER PROCEDURE or FUNCTION to Change OWNER |
| Date: | 2024-06-18 23:45:55 |
| Message-ID: | CAKFQuwbtpXWGjyi_6qmi3a1-j7kLKvkPpha2_QqjTSK71K9JEQ@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-admin |
On Tue, Jun 18, 2024 at 3:02 PM Vitale, Anthony, Sony Music <
anthony(dot)vitale(at)sonymusic(dot)com> wrote:
> I would this this is what you are looking for
>
For loops and SQL injection risks, not an ideal way to write SQL programs.
Though it does allow you to avoid using psql but in which case you need to
do away with the \set metacommands. If you are going to use psql I
strongly suggest running the select query, inspecting the results, then
changing said query to use \gexec. A lot fewer moving parts that dealing
with plpgsql.
Good call on making it dynamic on prokind though. And the owner matching.
Tom's got the right idea of just casting the OID for the main naming scheme
- it does the SQL injection mitigation.
David J.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Achilleas Mantzios | 2024-06-19 04:14:54 | Re: Statement_timeout in procedure block |
| Previous Message | lennam | 2024-06-18 22:38:03 | RE: Materialized views & dead tuples |