[darcs-devel] darcs patch: add test for replace that messes with un... (and 3 more)

[Eric Y. Kow]

> > Yeah, I figured it might be safe from a darcs-internals point of view.
> > I'll push it sometime this week or maybe this weekend.  That said, would
> > it work as a solution to just remove the
> >    isJust (apply_to_slurpy (tokreplace f toks old new) work)
> > since the pristine cache check is all that we really need?
> 
> Right, that would give the behavior you've described, but I prefer our
> current behavior, which I think is rather more intuitive.

Ah, and I've finally convinced myself that the current behaviour is ok,
and for that matter desirable.

The trick for me was to realise that it only applies cleanly to working
if all instances of the target token are removed, so doing a replace is
perfectly 'ok' in that there will be no surprises (again, in strictly UI
terms).  Somehow I was carrying around the idea that we would be
affecting pristine tokens the user hadn't thought about, but that isn't
going to happen because they'd all have been hunk-removed by previous
changes in working.

(I was already planning to push to patch in as-is anyway, but hey being
convinced is always a nice bonus)

-- 
Eric Kow                     http://www.loria.fr/~kow
PGP Key ID: 08AC04F9         Merci de corriger mon français.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 186 bytes
Desc: not available
Url : http://lists.osuosl.org/pipermail/darcs-devel/attachments/20070204/8a10a563/attachment.pgp