Re: Patch review for logging hooks (CF 2012-01)

On Sun, Mar 4, 2012 at 10:45 AM, Andrew Dunstan <andrew(at)dunslane(dot)net> wrote:
> I'm just looking at this patch, and I agree, it should be testable. I'm
> wondering if it wouldn't be a good idea to have a module or set of modules
> for demonstrating and testing bits of the API that we expose. src/test/api
> or something similar? I'm not sure how we'd automate a test for this case,
> though. I guess we could use something like pg_logforward and have a UDP
> receiver catch the messages and write them to a file. Something like that
> should be possible to rig up in Perl. But all that seems a lot of work at
> this stage of the game. So the question is do we want to commit this patch
> without it?

The latest version of this patch looks sound to me. We haven't
insisted on having even a sample application for every hook before,
let alone a regression test, so I don't think this patch needs one
either. Now, it might be fairly said that we ought to have regression
tests for a lot more things than we do right now, but that's basically
a limitation of our regression-testing environment which the author of
this patch shouldn't be obliged to fix.

So my vote is to go ahead and commit it.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company