[PATCH] test-manifestv2.t: test that we can convert to manifestv2 with clone

Drew Gottlieb drgott at google.com
Tue May 5 20:02:22 CDT 2015


I agree. Matt, feel free to drop those two lines.

On Tue, May 5, 2015 at 4:27 PM Adrian Buehlmann <adrian at cadifra.com> wrote:

> On 2015-05-06 00:59, Drew Gottlieb wrote:
> > # HG changeset patch
> > # User Drew Gottlieb <drgott at google.com>
> > # Date 1430865181 25200
> > #      Tue May 05 15:33:01 2015 -0700
> > # Node ID 3df093b5de908329a655061d0d70662e9ec07c4c
> > # Parent  c5d4f9cc8da7bb2068457e96e4f74ff694514ced
> > test-manifestv2.t: test that we can convert to manifestv2 with clone
> >
> > This just adds a test to ensure that we can convert a manifestv1 repo to
> a
> > manifestv2 repo (for future commits only) by means of a local clone.
> >
> > diff --git a/tests/test-manifestv2.t b/tests/test-manifestv2.t
> > --- a/tests/test-manifestv2.t
> > +++ b/tests/test-manifestv2.t
> > @@ -1,4 +1,71 @@
> > -Check that entry is added to .hg/requires
> > +Create repo with old manifest
> > +
> > +  $ hg init existing
> > +  $ cd existing
> > +  $ echo footext > foo
> > +  $ hg add foo
> > +  $ hg commit -m initial
> > +
> > +We're using v1, so no manifestv2 entry is in requires yet.
> > +
> > +  $ grep manifestv2 .hg/requires
> > +  [1]
> > +
> > +Let's clone this with manifestv2 enabled to switch to the new format for
> > +future commits.
> > +
> > +  $ cd ..
> > +  $ hg clone --pull existing new --config experimental.manifestv2=1
> > +  requesting all changes
> > +  adding changesets
> > +  adding manifests
> > +  adding file changes
> > +  added 1 changesets with 1 changes to 1 files
> > +  updating to branch default
> > +  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
> > +  $ cd new
> > +
> > +Check that entry was added to .hg/requires.
> > +
> > +  $ grep manifestv2 .hg/requires
> > +  manifestv2
> > +
> > +Make a new commit.
> > +
> > +  $ echo newfootext > foo
> > +  $ hg commit -m new
> > +
> > +Check that the manifest actually switched to v2.
> > +
> > +  $ hg debugdata -m 0
> > +  foo\x0021e958b1dca695a60ee2e9cf151753204ee0f9e9 (esc)
> > +
> > +  $ hg debugdata -m 1
> > +  \x00 (esc)
> > +  \x00foo\x00 (esc)
> > +  I\xab\x7f\xb8(\x83\xcas\x15\x9d\xc2\xd3\xd3:5\x08\xbad5_ (esc)
> > +
> > +Check that manifestv2 is used if the requirement is present, even if
> it's
> > +disabled in the config.
> > +
> > +  $ echo newerfootext > foo
> > +  $ hg --config experimental.manifestv2=False commit -m newer
> > +
>
> ..
>
> > +  $ grep manifestv2 .hg/requires
> > +  manifestv2
>
> Nitpick: I think you can leave away this one (goal: keep tests as
> minimal as possible). It's very unlikely that anyone ever manages to get
> a patch in which would remove an entry from .hg/requires. That file is
> written at repo creation.
>
> > +  $ hg debugdata -m 2
> > +  \x00 (esc)
> > +  \x00foo\x00 (esc)
> > +  \xa6\xb1\xfb\xef]\x91\xa1\x19`\xf3.#\x90S\xf8\x06 \xe2\x19\x00 (esc)
> > +
> > +Check that we can still read v1 manifests.
> > +
> > +  $ hg files -r 0
> > +  foo
> > +
> > +  $ cd ..
> > +
> > +Check that entry is added to .hg/requires on repo creation
> >
> >    $ hg --config experimental.manifestv2=True init repo
> >    $ cd repo
>
> Other than that: Nice testcase additions!
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://selenic.com/pipermail/mercurial-devel/attachments/20150506/dfd587bd/attachment.html>


More information about the Mercurial-devel mailing list