On Wed, 21 Dec 2011, Martin Packman wrote:
> Thanks Raphael! That cut down approach does work, though it also lacks
> some of the benefits of the larger change.
> It's a judgement call whether warning about invalid versions during
> merge is worthwhile
For this I have added a check in the Dpkg::Changelog parsing code:
commit 66a76d5abbc8dce24296108d109a96e7affad7cc
Author: Raphaël Hertzog <hert...@debian.org>
Date: Wed Dec 21 09:59:39 2011 +0100
Dpkg::Changelog::Entry::Debian: check_header() now validates the version
> the rearrangement to only construct version objects once per entry in
> the common case seems worth doing regardless to me.
I found that the speed gain was not worth the amount of supplementary
code.
Cheers,
--
Raphaël Hertzog ◈ Debian Developer
Pre-order a copy of the Debian Administrator's Handbook and help
liberate it: http://debian-handbook.info/liberation/
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
