From a7d50f5f4a9f69e33075fbe05d4ae42158d1b9c9 Mon Sep 17 00:00:00 2001
From: Nils Philippsen <nils@redhat.com>
Date: Wed, 26 Oct 2011 17:31:36 +0200
Subject: [PATCH 1/2] python: implement and use utf8 stream writer for stdout,
 stderr

The C-side (glib) really wants stuff to be encoded in UTF-8.
(cherry picked from commit b5a5011f31d6062cd00ee6b02ddf356d691e67e6)
---
 lib/python/packagekit/backend.py |   45 ++++++++++++++++++++++++++++++++++++++
 1 files changed, 45 insertions(+), 0 deletions(-)

diff --git a/lib/python/packagekit/backend.py b/lib/python/packagekit/backend.py
index ae7bbad..4c7fd2c 100644
--- a/lib/python/packagekit/backend.py
+++ b/lib/python/packagekit/backend.py
@@ -37,8 +37,49 @@ def _to_unicode(txt, encoding='utf-8'):
             txt = unicode(txt, encoding, errors='replace')
     return txt
 
+def _to_utf8(txt, errors='replace'):
+    '''convert practically anything to a utf-8-encoded byte string'''
+
+    # convert to unicode object
+    if isinstance(txt, str):
+        txt = txt.decode('utf-8', errors=errors)
+    if not isinstance(txt, basestring):
+        # try to convert non-string objects like exceptions
+        try:
+            # if txt.__unicode__() exists, or txt.__str__() returns ASCII
+            txt = unicode(txt)
+        except UnicodeDecodeError:
+            # if txt.__str__() exists
+            txt = str(txt).decode('utf-8', errors=errors)
+        except:
+            # no __str__(), __unicode__() methods, use representation
+            txt = unicode(repr(txt))
+
+    # return encoded as UTF-8
+    return txt.encode('utf-8', errors=errors)
+
 # Classes
 
+class _UTF8Writer(codecs.StreamWriter):
+
+    encoding = 'utf-8'
+
+    def __init__(self, stream, errors='replace'):
+        codecs.StreamWriter.__init__(self, stream, errors)
+
+    def encode(self, inp, errors='strict'):
+        try:
+            l = len(inp)
+        except TypeError:
+            try:
+                l = len(unicode(inp))
+            except:
+                try:
+                    l = len(str(inp))
+                except:
+                    l = 1
+        return (_to_utf8(inp, errors=errors), l)
+
 class PkError(Exception):
     def __init__(self, code, details):
         self.code = code
@@ -49,6 +90,10 @@ class PkError(Exception):
 class PackageKitBaseBackend:
 
     def __init__(self, cmds):
+        # Make sys.stdout/stderr cope with UTF-8
+        sys.stdout = _UTF8Writer(sys.stdout)
+        sys.stderr = _UTF8Writer(sys.stderr)
+
         # Setup a custom exception handler
         installExceptionHandler(self)
         self.cmds = cmds
-- 
1.7.6.4


From 01226fbdfc6cd9f03183048c7531ed324d4b5412 Mon Sep 17 00:00:00 2001
From: Nils Philippsen <nils@redhat.com>
Date: Wed, 26 Oct 2011 17:33:18 +0200
Subject: [PATCH 2/2] yum: don't let yum.misc.setup_locale() override stdout
 codec

Use our own stdout/stderr wrappers, as what codecs.getwriter() supplies
only works for unicode, but not for already encoded data.
(cherry picked from commit 31b9a6942a65f8eeeb67c6ea376429e8c5eac43f)
---
 backends/yum/yumBackend.py |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/backends/yum/yumBackend.py b/backends/yum/yumBackend.py
index 479ee2d..dfed1a2 100755
--- a/backends/yum/yumBackend.py
+++ b/backends/yum/yumBackend.py
@@ -3482,7 +3482,7 @@ class PackageKitYumBase(yum.YumBase):
             raise PkError(ERROR_FAILED_CONFIG_PARSING, _to_unicode(e))
 
         # setup to use LANG for descriptions
-        yum.misc.setup_locale(override_time=True)
+        yum.misc.setup_locale(override_time=True, override_codecs=False)
 
         self.missingGPGKey = None
         self.dsCallback = DepSolveCallback(backend)
-- 
1.7.6.4