Re: BUG #7873: pg_restore --clean tries to drop tables that don't exist

Josh Kupershmidt <schmiddy@gmail.com>

From: Josh Kupershmidt <schmiddy@gmail.com>
To: Pavel Stehule <pavel.stehule@gmail.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Bruce Momjian <bruce@momjian.us>, Dave Rolsky <autarch@urth.org>, pgsql-bugs@postgresql.org
Date: 2013-02-21T02:28:42Z
Lists: pgsql-bugs, pgsql-hackers
On Tue, Feb 19, 2013 at 6:00 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:
> 2013/2/16 Pavel Stehule <pavel.stehule@gmail.com>:
>> 2013/2/16 Tom Lane <tgl@sss.pgh.pa.us>:
>>> I think it has come up before.  I wouldn't object to a pg_dump option to
>>> add IF EXISTS to all the drop commands (though changing the default
>>> behavior would be more controversial).  Don't intend to spend my own
>>> time on it though ...

FYI, it was proposed here:
http://www.postgresql.org/message-id/507AD08C.5020603@dalibo.com

> here is patch, that we use about one year - originally for 9.1 - I did
> port to 9.3

dropdb and dropuser both support a similar option named --if-exists. I
suggest --if-exists instead of --conditional-drops for consistency.
I've only glanced at the patch, but if it makes no sense to use
--conditional-drops (or --if-exists, whatever it ends up being called)
without --clean, then attempting to do so should raise an error.

Josh