[PATCH] pgpassfile connection option - Mailing list pgsql-hackers

From Julian Markwort
Subject [PATCH] pgpassfile connection option
Date
Msg-id a4b4f4f1-7b58-a0e8-5268-5f7db8e8ccaa@uni-muenster.de
Whole thread Raw
Responses Re: [PATCH] pgpassfile connection option  (Andrew Dunstan <andrew@dunslane.net>)
Re: [PATCH] pgpassfile connection option  (Fabien COELHO <coelho@cri.ensmp.fr>)
List pgsql-hackers
Hello psql-hackers!

We thought it would be advantageous to be able to specify a 'custom'
pgpassfile within the connection string along the lines of the existing
parameters sslkey and sslcert.

Which is exactly what this very compact patch does.
The patch is minimally invasive - when no pgpassfile attribute is
provided in the connection string, the regular pgpassfile is used.
The security-measures (which are limited to checking the permissions for
0600) are kept, however we could loosen that restriciton to allow group
access as well along the lines of the ssl key file , if this is
preferred. (in case multiple users belonging to the same group would
like to connect using the same file).

The patch applies cleanly to master and compiles and runs as expected
(as there are no critical alterations).
I've not written any documentation as of now, but I'll follow up closely
if there is any interest for this patch.

notes:
  - using ~ to denote the user's home directory in the path does not
work, however $HOME works (as this is translated by bash beforehand).
  - the notation in the custom pgpassfile should follow the notation of
the 'default' pgpass files:
     hostname:port:database:username:password
  - this has only been tested on linux so far, however due to the nature
of the changes I suspect that there is nothing that could go wrong in
other environments, although I could test that as well, if deemed necessary.


I'm looking forward to any feedback,
Julian

--

Julian Markwort
Westphalian Wilhelms-University in Münster
julian.markwort@uni-muenster.de


Attachment

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: Executor's internal ParamExecData Params versus EvalPlanQual
Next
From: Andrew Dunstan
Date:
Subject: Re: [PATCH] pgpassfile connection option