filemerge: A feature request

Thomas Arendsen Hein thomas at intevation.de
Wed Feb 6 11:46:36 CST 2008


* Matt Mackall <mpm at selenic.com> [20080206 18:33]:
> 
> On Wed, 2008-02-06 at 17:34 +0100, Thomas Arendsen Hein wrote:
> > * Matt Mackall <mpm at selenic.com> [20080206 17:25]:
> > > On Wed, 2008-02-06 at 00:51 -0600, Steve Borho wrote:
> > > > You can specify a binary merging tool in the merge-tools specification.
> > > > 
> > > > [merge-tools]
> > > > mytool.args = $local $base $other
> > > > mytool.binary = true
> > > >
> > > > However, there's no explicit way to say:  binary -> internal:local
> > > 
> > > Well... you can do:
> > > 
> > > [merge-tools]
> > > internal:other.binary = True
> > > internal:other.priority = -100
> > > 
> > > ..which says "if nothing else can merge this binary. take the other
> > > version".
> > 
> > No, "internal:other..." is interpreted as "internal = other...".
> 
> Forgot about that. Perhaps we should change the ":" now before we're
> stuck with it.
> 
> On the other hand, I was planning on having any Python merge extensions
> being python:<func> like we use in hooks. Argh.
> 
> Perhaps we should instead replace/fix ConfigParser. Do we document
> "name: val" anywhere?

I don't think so and as this yields more surprises than good, it
would be good to get rid of that. Bonus points for having ordered
entries in sections to make e.g. encode/decode happy :)

Thomas

-- 
thomas at intevation.de - http://intevation.de/~thomas/ - OpenPGP key: 0x5816791A
Intevation GmbH, Osnabrueck - Register: Amtsgericht Osnabrueck, HR B 18998
Geschaeftsfuehrer: Frank Koormann, Bernhard Reiter, Dr. Jan-Oliver Wagner


More information about the Mercurial-devel mailing list