[issue2327] hg bisect should have -f option

Greg Ward greg-hg at gerg.ca
Fri Aug 13 08:18:44 CDT 2010


On Thu, Aug 12, 2010 at 7:33 PM, Russ Cox <bugs at mercurial.selenic.com> wrote:
> Almost every time I want to bisect, it's because I've found
> that changing something in the tree exposes the bug.
> Now, it's possible to run bisect, make the change that
> exposes the bug, test, revert the change, run bisect, etc.,
> but usually update's merger can take care of that for me.
> Except that bisect doesn't have a -f flag.  This patch adds
> it with the obvious meaning.

Pasting patches into the bug tracker doesn't work.  Committers aren't
going to bother looking for patches there or dealing with the
inevitable whitespace issues.  Please see

  http://mercurial.selenic.com/wiki/ContributingChanges
  http://mercurial.selenic.com/wiki/SuccessfulPatch

for the best way to get a patch accepted.

> diff -u commands0.py commands.py
> --- commands0.py        2010-08-12 16:27:58.000000000 -0700
> +++ commands.py 2010-08-12 16:30:11.000000000 -0700
[...]
> @@ -4001,7 +4003,8 @@
>           ('s', 'skip', False, _('skip testing changeset')),
>           ('c', 'command', '',
>            _('use command to check changeset state'), _('CMD')),
> -          ('U', 'noupdate', False, _('do not update to target'))],
> +          ('U', 'noupdate', False, _('do not update to target')),
> +          ('f', 'force', False, _('skip check for outstanding uncommitted
> changes'))],

I would make the help message

  do not check for outstanding uncommitted changes

or, on second thought:

  do not check for uncommitted changes

Yeah, that feels right.

Other than that, I like it.  Resend to the devel list and see what
people think!  Thanks for your contribution --

Greg


More information about the Mercurial-devel mailing list