Re: BUG #5560: pg_dump generates invalid DDL

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "Shigeru Hanada" <hanada(at)metrosystems(dot)co(dot)jp>
Cc: pgsql-bugs(at)postgresql(dot)org, Peter Eisentraut <peter_e(at)gmx(dot)net>
Subject: Re: BUG #5560: pg_dump generates invalid DDL
Date: 2010-07-13 22:31:22
Message-ID: 18612.1279060282@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

"Shigeru Hanada" <hanada(at)metrosystems(dot)co(dot)jp> writes:
> When target database contains user mapping for public, pg_dump generates
> invalid DDL with empty string like this.

> CREATE USER MAPPING FOR "" SERVER srv;

I can reproduce this behavior.

Another serious problem with the pg_dump code for user mappings is that
it fails outright if the user is not superuser --- meaning that that
patch has completely broken non-superuser use of pg_dump. That's
entirely unacceptable IMO.

The simplest fix seems to be to have pg_dump look at
the pg_user_mappings view rather than directly at pg_user_mapping.
That both provides the special case for PUBLIC and gets around the
permissions problem. It does mean that user mappings will be silently
dumped without the options data if you don't have privileges for the
specific mapping, but ISTM that beats failing entirely. (Alternatively,
we could probably persuade pg_dump to not dump mappings the caller
hasn't got permissions for ... is that better?)

Peter, any comments?

regards, tom lane

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Jaime Casanova 2010-07-14 00:39:40 Re: BUG #5557: Problema com Bytea
Previous Message Tom Lane 2010-07-13 22:16:03 Re: BUG #5561: Tsearch across schemas