[merge] win32- fix test_bundle

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Aaron Bentley wrote:
> John Arbash Meinel wrote:
>>> Hopefully this is my last win32 related bugfix for a while. And it
>>> should be, since the test suite now passes :)
>>>
>>> Anyway, I updated the test suite to use a starting space, rather than a
>>> trailing space. Which is supported on win32.
> 
> I thought that it made a lot of sense to be testing for trailing spaces.
>  That's the kind of area where bundles are most vulnerable.
> 
> I'd prefer to move this and the tab test into a different test, and skip
> it on Win32.

Sure. As it stands test_bundle is pretty ugly. I had written it quite a
while ago, and was using more of a shotgun approach to testing. "If
testing X once is good, testing it twice is better".

I didn't really want to take the time to refactor it, but I can
certainly add a new test for this kind of whitespace stuff.

I guess part of it also tests rollup bundles, etc.
...

> @@ -411,6 +418,10 @@
>              rh = self.b1.revision_history()
>              tree.branch.set_revision_history(rh[:rh.index(rev_id)+1])
>              tree.update()
> +            delta = compare_trees(tree,
> +                                self.b1.repository.revision_tree(rev_id))
> +            self.assertFalse(delta.has_changed(),
> +                             'Working tree has modifications')
>          return tree
> 
> I'd like to switch the order of the trees.  When I examined the delta
> here, it said 'file with trailing space added', which was confusing
> because that path was actually missing from the working tree.
> 
> Aaron

I completely agree. It was probably just that way to make it shorter on
the line. But the diff would be confusing.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEpzlbJdeBCYSNAAMRArPOAKCNaVFpXl5MiVAOIFHo4X1/VhdECgCdHvWl
w3KOprsAIgO28JgmRpqnonQ=
=xDWC
-----END PGP SIGNATURE-----