[PATCH] Fixed issue 884: Build products not ignored until .hgignore touched

[Renato Cunha]

On Tue, Jul 06, 2010 at 09:11:18AM +0900, Nicolas Dumazet wrote:
> nitpick: You are probably expected to prefix the commit message by
> "inotify: " here.

Thanks! I'll learn. Eventually...

> >
> > This is an attempt to fix issue 884. In summary, the bug was that "hg status"
> > displayed ignored files as untracked, instead of not displaying them. A new
> > test has been added to the main inotify test to prevent any regressions.
> >
> > diff --git a/hgext/inotify/server.py b/hgext/inotify/server.py
> > --- a/hgext/inotify/server.py
> > +++ b/hgext/inotify/server.py
> > @@ -213,7 +213,7 @@
> >             if time != int(st_mtime):
> >                 return 'l'
> >             return 'n'
> > -        if type_ == '?' and self.dirstate._ignore(fn):
> > +        if type_ == '?' and self.dirstate._dirignore(fn):
> >             return 'i'
> >         return type_
> 
> At first, I was very confused by this fix.
> I mean. As far as I can tell, in this context we are SURE that fn is a
> file, and not a directory.
> It hence looks strange to check _dirignore on this file.
> 
> But you are right, in the sense that ignore patterns should match each
> component of the path.

Good to know that you agree with it. It took me quite some time to find the
source of the problem, and I wasn't very confident about my change.

> This fix looks good, kudos to you :)

Thanks again. ;)

-- 
Renato Cunha <http://renatocunha.com>
Blog: http://valedotrovao.com
"Whatever happens, happens"