On Tue, Feb 24, 2009 at 06:45:13PM +0100, sean finney wrote: > On Tue, Feb 24, 2009 at 11:15:14AM +0100, Guido Günther wrote: > > > the fix is to use regexpObj.search() instead of regexpObj.match(), as the > > > former will match anywhere in the line, wherease the latter only matches > > > starting at the beginning of the line. pretty trivial, but patch attached > > > anyway :) > > This is actually intentional. I don't want to match it on random Closes: > > in the commit body. You can have the full commit body with --full which > > gives you the closes in the commit message then. > > Am I missing a usage case? > > okay... i was trying to include (Closes: #foo) from a more prose-like > part of a commit message without having to include it in the fairly > space-limited portion of the short commit msg. > > there didn't seem to be any documentation in the fine manual for how this > is supposed to work apart from the description of --meta/--meta-closes. > the description of these options doesn't make any mention of such an > assumption, so i assumed it was an unintended bug. > > so i have no problem with it being said that the Closes: stuff should be > on its own line, maybe it's even nicer that way. perhaps this bug can thus > be re-categorized as "document better how the meta/closes feature works" :) A patch for this would be very welcome too! -- Guido
-- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
