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

[Martin Packman]

Require UnicodeFilenameFeature for new tests in bb.test_conflicts

Fixes another an oversight from <lp:~gz/bzr/conflicts_non_ascii_ui_686161> by requiring unicode filenames to be supported to run the new tests in bb.test_conflicts so they'll be skipped on systems without a locale set.

Confusingly, the Feature I needed wasn't in bzrlib.tests.features and appears to be duplicated as both UnicodeFilenameFeature and UnicodeFilename.

[Jelmer Vernooij (jelmer)]

On Sun, 2011-05-22 at 12:01 +0000, Martin [gz] wrote:
> Confusingly, the Feature I needed wasn't in bzrlib.tests.features and appears to be duplicated as both UnicodeFilenameFeature and UnicodeFilename.
Can you file a bug about that?

 review approve

> differences between files attachment (review-diff.txt)
> === modified file 'bzrlib/tests/blackbox/test_conflicts.py'
> --- bzrlib/tests/blackbox/test_conflicts.py 2011-05-21 16:29:38 +0000
> +++ bzrlib/tests/blackbox/test_conflicts.py 2011-05-22 12:01:44 +0000
> @@ -19,7 +19,7 @@
> tests,
> workingtree,
> )
> -from bzrlib.tests import script
> +from bzrlib.tests import script, UnicodeFilenameFeature

We're usually writing these imports on several lines now, to make
merging less problematic. I guess it's not really a big deal in this
case.

Cheers,

Jelmer

review: Approve

[Martin Packman (gz)]

> Can you file a bug about that?

Filed bug 786537 and will probably tackle it later today.

> We're usually writing these imports on several lines now, to make
> merging less problematic. I guess it's not really a big deal in this
> case.

Yep, and had I not looked in the wrong place first could have not touched the import and just used tests.UnicodeFilenameFeature instead. I'll need to change this line anyway to import the features module shortly.

[Martin Packman (gz)]

sent to pqm by email

[Andrew Bennetts (spiv)]

Jelmer Vernooij wrote:
[…]
> > -from bzrlib.tests import script
> > +from bzrlib.tests import script, UnicodeFilenameFeature
>
> We're usually writing these imports on several lines now, to make
> merging less problematic. I guess it's not really a big deal in this
> case.

It'd be neat if someone wrote a per-file merger that could handle
merging import lines of a .py file…

-Andrew

=== modified file 'bzrlib/tests/blackbox/test_conflicts.py'
--- bzrlib/tests/blackbox/test_conflicts.py	2011-05-21 16:29:38 +0000
+++ bzrlib/tests/blackbox/test_conflicts.py	2011-05-22 12:01:44 +0000
@@ -19,7 +19,7 @@
     tests,
     workingtree,
     )
-from bzrlib.tests import script
+from bzrlib.tests import script, UnicodeFilenameFeature
 
 
 def make_tree_with_conflicts(test, this_path='this', other_path='other',
@@ -86,6 +86,7 @@
 class TestUnicodePaths(tests.TestCaseWithTransport):
     """Unicode characters in conflicts should be displayed properly"""
 
+    _test_needs_features = [UnicodeFilenameFeature]
     encoding = "UTF-8"
 
     def _as_output(self, text):