From: | Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> |
---|---|
To: | Joel Jacobson <joel(at)gluefinance(dot)com> |
Cc: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: UPDATE pg_catalog.pg_proc.prosrc OK? |
Date: | 2010-12-28 13:28:27 |
Message-ID: | 4D19E5FB.9030500@enterprisedb.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On 28.12.2010 15:19, Joel Jacobson wrote:
> 2010/12/28 Robert Haas<robertmhaas(at)gmail(dot)com>
>
>> I'm not immediately sure whether it's safe, but it seems like playing
>> with fire, and I don't see any particular advantage to doing it this
>> way over using CREATE OR REPLACE FUNCTION.
>
> While I understand some of the SQL commands affecting pg_catalog have side
> effects, such as CREATE DATABASE, others seems to lack side effects.
>
> To use CREATE OR REPLACE FUNCTION, I would have to assemble SQL from the
> data in pg_proc.* for each function, which is quite a complex task (e.g.,
> pg_dump does this, and it's far from trivial, due to differences between
> different versions etc).
Since 8.4, there is a function called pg_get_functiondef() which does
all the hard work. Or you could use pg_dump.
--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com
From | Date | Subject | |
---|---|---|---|
Next Message | Magnus Hagander | 2010-12-28 13:31:56 | pg_primary_conninfo |
Previous Message | Joel Jacobson | 2010-12-28 13:19:02 | Re: UPDATE pg_catalog.pg_proc.prosrc OK? |