Merge lp:~vila/bzr/1086209-lc-all-c into lp:bzr

[Vincent Ladeuil]

Fix LC_ALL=C test failures related to utf8 stderr encoding

This should fix the LC_ALL=C test failures.

It's a bit weird that we didn't try to use the self.stderr encoding so I
went with a conservative fix: if something bad happens, we'll fall back to
the previous implementation.

[John A Meinel (jameinel)]

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

On 2012-12-04 18:29, Vincent Ladeuil wrote:
> Vincent Ladeuil has proposed merging lp:~vila/bzr/1086209-lc-all-c
> into lp:bzr.
>
> Requested reviews: bzr-core (bzr-core) Related bugs: Bug #1086209
> in Bazaar: "test failures in bzr.dev when run with LC_ALL=C"
> https://bugs.launchpad.net/bzr/+bug/1086209
>
> For more details, see:
> https://code.launchpad.net/~vila/bzr/1086209-lc-all-c/+merge/137862
>
> This should fix the LC_ALL=C test failures.
>
> It's a bit weird that we didn't try to use the self.stderr encoding
> so I went with a conservative fix: if something bad happens, we'll
> fall back to the previous implementation.
>

+* Fix test failures occurring only with LC_ALL=C, when prompting a user,
+ ``stdee`` encoding should be preferred over terminal encoding if
+ available. (Vincent Ladeuil, #1086209)

tyop

Otherwise:
 review: approve

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (Cygwin)
Comment: Using GnuPG with undefined - http://www.enigmail.net/

iEYEARECAAYFAlC+HCMACgkQJdeBCYSNAAP4dACfd6iuMu3YbSn+ICQMQKKZqs2L
JZUAn0iYiaxAE1q0Z0r4W1dAVkHwODx+
=lqmC
-----END PGP SIGNATURE-----

review: Approve

[Vincent Ladeuil (vila)]

sent to pqm by email

=== modified file 'bzrlib/tests/test_ui.py'
--- bzrlib/tests/test_ui.py	2012-09-17 09:06:56 +0000
+++ bzrlib/tests/test_ui.py	2012-12-04 16:40:37 +0000
@@ -103,7 +103,6 @@
         ui.stdout = tests.StringIOWrapper()
         ui.stderr = tests.StringIOWrapper()
         ui.stderr.encoding = ui.stdout.encoding = ui.stdin.encoding = 'utf8'
-        pb = ui.nested_progress_bar()
         password = ui.get_password(u'Hello \u1234 %(user)s', user=u'some\u1234')
         self.assertEqual(u'baz\u1234', password)
         self.assertEqual(u'Hello \u1234 some\u1234: ',
 
=== modified file 'bzrlib/ui/text.py'
--- bzrlib/ui/text.py	2012-10-10 18:45:21 +0000
+++ bzrlib/ui/text.py	2012-12-04 16:40:37 +0000
@@ -336,7 +336,13 @@
         if kwargs:
             # See <https://launchpad.net/bugs/365891>
             prompt = prompt % kwargs
-        prompt = prompt.encode(osutils.get_terminal_encoding(), 'replace')
+        try:
+            prompt = prompt.encode(self.stderr.encoding)
+        except (UnicodeError, AttributeError):
+            # If stderr has no encoding attribute or can't properly encode,
+            # fallback to terminal encoding for robustness (better display
+            # something to the user than aborting with a traceback).
+            prompt = prompt.encode(osutils.get_terminal_encoding(), 'replace')
         self.clear_term()
         self.stdout.flush()
         self.stderr.write(prompt)
 
=== modified file 'doc/en/release-notes/bzr-2.6.txt'
--- doc/en/release-notes/bzr-2.6.txt	2012-10-23 10:22:25 +0000
+++ doc/en/release-notes/bzr-2.6.txt	2012-12-04 16:40:37 +0000
@@ -104,9 +104,9 @@
 Testing
 *******
 
-.. Fixes and changes that are only relevant to bzr's test framework and 
-   suite.  This can include new facilities for writing tests, fixes to 
-   spurious test failures and changes to the way things should be tested.
+* Fix test failures occurring only with LC_ALL=C, when prompting a user,
+  ``stderr`` encoding should be preferred over terminal encoding if
+  available. (Vincent Ladeuil, #1086209)
 
 
 bzr 2.6b2