Horton

Re: [BUG FIX]Connection fails with whitespace after keepalives parameter value

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Fujii Masao <masao.fujii@oss.nttdata.com>, "Yuto Sasaki (Fujitsu)" <sasaki.yuto-00@fujitsu.com>, "pgsql-hackers@lists.postgresql.org" <pgsql-hackers@lists.postgresql.org>
Date: 2024-10-03T23:39:39Z
Lists: pgsql-hackers 
On Thu, Oct 03, 2024 at 11:57:16AM -0400, Tom Lane wrote:
> I don't have a strong opinion one way or the other about whether
> we should make libpq permissive about extra spaces (as per
> Michael's patch).  I guess you could argue that all of these
> fixes are consistent with the principle of "be conservative
> with what you send and liberal with what you accept".  But at
> most I'd fix these remaining things in HEAD.

Removing this extra whitespace from the ECPG strings sounds good here.

FWIW, my argument about doing this in libpq is not really related to
ECPG: it feels inconsistent to apply one rule for the parameters and a
different one for the values in URIs.  So I'd be OK to see how this
goes on as a HEAD-only change.
--
Michael

Commits

  1. libpq: Discard leading and trailing spaces for parameters and values in URIs

  2. ecpg: avoid adding whitespace around '&' in connection URLs.

  3. Parse libpq's "keepalives" option more like other integer options.