Different `diff` output for null repository

[Stephen Rasku]

Hi Mischa,

Thanks for your answer.  I've tested it and I see that's how it works.

This clause from `hg help diff` should probably be corrected, then:

    If only one revision is specified then that revision is compared to the
working directory

because it's not accurate if you are at the null revision.  Personally, I
would prefer that the behaviour was changed to match the documentation
rather than the documentation being changed to match the behaviour.
However, I know the project takes great pains to not modify the command
line behaviour so I understand why they wouldn't want to do that.

In any case, thanks again for the answer.  It gives me a workaround for
what I'm trying to do.

...Stephen

On Mon, Nov 14, 2016 at 11:59 AM, Becker, Mischa J <mischa.becker at kroger.com
> wrote:

> > From: Mercurial [mailto:mercurial-bounces at mercurial-scm.org] On Behalf
> Of Stephen Rasku
> > Sent: Sunday, November 13, 2016 4:31 PM
> > To: mercurial at mercurial-scm.org
> > Subject: Different `diff` output for null repository
> >
> > I have the following test file.
> >  $ hg init diff
> >  $ cd diff
> >  $ echo foo > foo
> >  $ hg add foo
> >  $ hg ci -m "foo"
> >  $ echo bar >> foo
> >  $ hg ci -m "foo+bar"
> >  $ hg up -r 1
> >  0 files updated, 0 files merged, 0 files removed, 0 files unresolved
> >  $ hg cat -r tip foo > foo
> >  $ cat foo
> >  foo
> >  bar
> >  $ hg diff -r tip
> >  $ hg up null
> >  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
> >  $ hg cat -r tip foo > foo
> >  $ cat foo
> >  foo
> >  bar
> >  $ hg diff -r tip
> >
> > I would expect the two diff  commands to have the same output (i.e.
> nothing).  However,
> > I get the following output:
> >
> > $ cram diff.t
> >!
> > --- diff.t
> > +++ diff.t.err
> > @@ -19,3 +19,9 @@
> >   foo
> >   bar
> >   $ hg diff -r tip
> > +  diff -r 6695c077f203 foo
> > +  --- a/foo\tMon Nov 14 00:29:46 2016 +0000 (esc)
> > +  +++ /dev/null\tThu Jan 01 00:00:00 1970 +0000 (esc)
> > +  @@ -1,2 +0,0 @@
> > +  -foo
> > +  -bar
> >
> > # Ran 1 tests, 0 skipped, 1 failed.
> > Am I wrong in my expectations?  Is this documented anywhere?
> >...Stephen
>
> If you were doing the above diff test on revisions 0 and 1 then it would
> work as expected.  The null revision however is before revision 0.  At that
> point foo is not a tracked file and will be ignored by hg diff until you hg
> add it again.  Running hg stat --all will show you the difference.
>
> Same thing would happen if foo wasn't added to your repo until revision 4
> and you ran the above test on revision 3 instead of null.
>
> Mischa
>
> ________________________________
>
> This e-mail message, including any attachments, is for the sole use of the
> intended recipient(s) and may contain information that is confidential and
> protected by law from unauthorized disclosure. Any unauthorized review,
> use, disclosure or distribution is prohibited. If you are not the intended
> recipient, please contact the sender by reply e-mail and destroy all copies
> of the original message.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mercurial-scm.org/pipermail/mercurial/attachments/20161114/a0ffe96c/attachment-0002.html>