Re: PGXS: REGRESS_OPTS=--load-language=plpgsql

Tom Lane wrote:
> Bruce Momjian <bruce(at)momjian(dot)us> writes:
> > Tom Lane wrote:
> >> Attached is a draft patch (no doc changes) that implements CREATE OR
> >> REPLACE LANGUAGE
>
> > How is pg_migrator affected by this? It always loads the the dump as
> > the super-user. How will the pg_dump use CREATE OR REPLACE LANGUAGE?
>
> pg_dump would issue "CREATE OR REPLACE LANGUAGE plpgsql" which would
> succeed just fine, since it'd be issued by a superuser.
>
> I think the potential downsides of that are significantly smaller than
> having a special case that excludes plpgsql altogether --- for one
> example, it would still succeed in a custom installation that had been
> changed so that plpgsql wasn't installed by default.

Are we doing this just for plpgsql in pg_dump?

> BTW, another problem I just noticed with the current kluge is that it
> fails to transfer any nondefault permissions that might have been
> attached to plpgsql.

Well, I assumed the permissions would still come, just not the CREATE
LANGUAGE, but now that I think about it you might be right.

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com
PG East: http://www.enterprisedb.com/community/nav-pg-east-2010.do
+ If your life is a hard drive, Christ can be your backup. +