Merge lp:~gz/bzr/trivial_mac_insensitive_glob into lp:bzr

[Martin Packman]

Skip win32utils case insensitive globbing tests on OS X

Refine the skip condition of a couple of win32utils tests so OSX isn't surprised when it runs them and they don't work. In practice, these tests require not only an case-insensitive filesystem, but beyond that working support in os.path itself.

I've not made this into a new feature because something that comes up in one file only isn't refined enough to make into an api, especially when the current check is depending on implementation details of several standard library modules. Nevertheless, it's more accurate than the current.

[John A Meinel (jameinel)]

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

On 12/7/2010 11:45 AM, Martin [gz] wrote:
> Martin [gz] has proposed merging lp:~gz/bzr/trivial_mac_insensitive_glob into lp:bzr.
>
> Requested reviews:
> bzr-core (bzr-core)
> Related bugs:
> #686595 Windows-specific case insensitive globbing tests fail on Mac OSX
> https://bugs.launchpad.net/bugs/686595
>
>
> Refine the skip condition of a couple of win32utils tests so OSX isn't surprised when it runs them and they don't work. In practice, these tests require not only an case-insensitive filesystem, but beyond that working support in os.path itself.
>
> I've not made this into a new feature because something that comes up in one file only isn't refined enough to make into an api, especially when the current check is depending on implementation details of several standard library modules. Nevertheless, it's more accurate than the current.

 review: approve

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkz+0JMACgkQJdeBCYSNAAMLhgCgtDFCT4Fr1DHWBmYUJWhQnsJ9
EhoAoNIECpstAzkPdiByd2nNlo5L/fB9
=byqi
-----END PGP SIGNATURE-----

review: Approve

[Vincent Ladeuil (vila)]

Thanks ! You beat me too it :)

This makes sense when reading the cover letter but is a bit terse otherwise. In the long term, we want to be able to test any sort of file system by using in-memory trees, so it's a good thing to keep precise tests (instead of running the whole module on windows only) but we also need to document the details.

So I'll copy some of the cover letter into comments before merging.

review: Approve

[Vincent Ladeuil (vila)]

Meh, the skip condition documents the point well enough, what crossed my mind ?

[Vincent Ladeuil (vila)]

sent to pqm by email

=== modified file 'bzrlib/tests/test_win32utils.py'
--- bzrlib/tests/test_win32utils.py	2010-06-22 23:58:34 +0000
+++ bzrlib/tests/test_win32utils.py	2010-12-07 17:45:00 +0000
@@ -120,7 +120,8 @@
             ])
 
     def test_case_insensitive_globbing(self):
-        self.requireFeature(tests.CaseInsCasePresFilenameFeature)
+        if os.path.normcase("AbC") == "AbC":
+            self.skip("Test requires case insensitive globbing function")
         self.build_ascii_tree()
         self._run_testset([
             [[u'A'], [u'A']],
@@ -326,7 +327,8 @@
             single_quotes_allowed=True)
 
     def test_case_insensitive_globs(self):
-        self.requireFeature(tests.CaseInsCasePresFilenameFeature)
+        if os.path.normcase("AbC") == "AbC":
+            self.skip("Test requires case insensitive globbing function")
         self.build_tree(['a/', 'a/b.c', 'a/c.c', 'a/c.h'])
         self.assertCommandLine([u'A/b.c'], 'A/B*')