Re: tests for client programs - Mailing list pgsql-hackers

From Peter Eisentraut
Subject Re: tests for client programs
Date
Msg-id 1391915261.920.1.camel@vanquo.pezone.net
Whole thread Raw
In response to Re: tests for client programs  ("Erik Rijkers" <er@xs4all.nl>)
Responses Re: tests for client programs  (Peter Eisentraut <peter_e@gmx.net>)
List pgsql-hackers
On Wed, 2014-01-15 at 20:56 +0100, Erik Rijkers wrote:
> 2 tests stumbled:
>
> 1. One test  ( pg_ctl/t/001_start_stop.pl )  failed because I had PGDATA set.  I unset all PG+ vars after that.  No a
big
> problem but nonetheless it might be better if the test suite removes /controls the variables before running.
>
> 2. The pg_isready test failed command_fails()  ('fails with no server running') because it defaults to the
compiled-in
> server-port (and that server was running).  I added the test-designated port (65432, as defined in TestLib.pm).  This
> simple change is in the attached patch.
>
>
> With these two changes the whole test suite passed.

Fixed those two things by unsetting environment variables and picking a
different port.

New patch attached.

Attachment

pgsql-hackers by date:

Previous
From: Peter Eisentraut
Date:
Subject: Re: clang's -Wmissing-variable-declarations shows some shoddy programming
Next
From: Peter Eisentraut
Date:
Subject: Re: Review: tests for client programs