Re: [PATCH] Add PQconninfoParseParams and PQconninfodefaultsMerge to libpq - Mailing list pgsql-hackers

From Amit Kapila
Subject Re: [PATCH] Add PQconninfoParseParams and PQconninfodefaultsMerge to libpq
Date
Msg-id 005001ce140a$fcb4aa10$f61dfe30$@kapila@huawei.com
Whole thread Raw
In response to Re: [PATCH] Add PQconninfoParseParams and PQconninfodefaultsMerge to libpq  (Heikki Linnakangas <hlinnakangas@vmware.com>)
List pgsql-hackers
On Tuesday, February 26, 2013 1:36 PM Heikki Linnakangas wrote:
> On 26.02.2013 09:06, Amit Kapila wrote:
> > On Monday, February 25, 2013 11:26 PM Heikki Linnakangas wrote:
> >> On 21.02.2013 16:09, Amit Kapila wrote:
> >>> On Thursday, February 21, 2013 12:46 AM Heikki Linnakangas wrote:
> >> I've committed those patches, with some further changes. If you have
> >> the
> >> time, please take another look at the committed patches. Thanks!
> >
> > 1. For pg_dumpall, incase user gives dbname in connection string, for
> ex.
> > "user=amit dbname=db_regress port=5434"
> >     it will be ignored and postgres (default database name) will be
> used.
> 
> Yeah, that is mentioned in the docs. 

Sorry, I have missed that part, this just came to me when I was looking at
the code.

>The other option would be to use
> the database name from the connection string for the initial
> connection,
> ie. the same as specifying a database name with the -l option.

This is what come to my mind initially, as if it is allowed with -l option
it is better if the same should be allowed through connection string as
well.
We are ignoring for pg_basebackup as well, but in pg_basebackup there is no
way for user to
specify database name.


With Regards,
Amit Kapila.




pgsql-hackers by date:

Previous
From: Marios Vodas
Date:
Subject: GIST_LEAF() causes error in GiST methods union and penalty
Next
From: Andres Freund
Date:
Subject: Re: [RFC] Extend namespace of valid guc names