hg identify first draft

[Matt Mackall]

On Wed, Jun 15, 2005 at 09:35:46AM +0200, Thomas Arendsen Hein wrote:
> * Thomas Arendsen Hein <thomas at intevation.de> [20050615 09:12]:
> > * Matt Mackall <mpm at selenic.com> [20050615 08:54]:
> > > On Wed, Jun 15, 2005 at 07:45:38AM +0200, Thomas Arendsen Hein wrote:
> > > > Attached is my version based on yours, pullable from
> > > > http://hg.intevation.org/mercurial-identify/
> > > > 
> > > > Matt: This is not yet in mercurial-tah, please state if I should
> > > > undo (or revert) my hgweb fix.
> > > 
> > > Still mulling over that one. Pop it for now, please.
> > 
> > Ok, I did undo it, nothing else changed.
> > 
> > Now, what is the correct way to merge all these trees in a clean way?
> > I always end up with commands.py containing not all changes from all
> > trees.
> 
> Ah! hg.py versus hgmerge:
> 
> In hg.py:
>         a = self.wjoin(fn)
>         b = temp("other", other)
>         c = temp("base", base)
>         r = os.system("%s %s %s %s" % (cmd, a, b, c))
> 
> In hgmerge:
> LOCAL=$1
> BASE=$2
> OTHER=$3

Ugh. Good catch.

> I knew that it was working when I still was using the old tkmerge
> script. I suggest swapping b and c in hg.py, *but* please take a
> close look at hgmerge if this breaks anything.

I'll do you one better: I added a merge conflict test case.

> http://hg.intevation.org/mercurial-tah/ now includes hg identify, too.

Pulled.

-- 
Mathematics is the supreme nostalgia of our time.