Fix for ascii decode in DataSourceAzure (LP: #1422993).

4 files changed, 15 insertions, 4 deletions

diff --git a/cloudinit/sources/DataSourceAzure.py b/cloudinit/sources/DataSourceAzure.py
index 444070bb..6e030217 100644
--- a/cloudinit/sources/DataSourceAzure.py
+++ b/cloudinit/sources/DataSourceAzure.py
@@ -137,7 +137,7 @@ class DataSourceAzureNet(sources.DataSource):
 
         if found != ddir:
             cached_ovfenv = util.load_file(
-                os.path.join(ddir, 'ovf-env.xml'), quiet=True)
+                os.path.join(ddir, 'ovf-env.xml'), quiet=True, decode=False)
             if cached_ovfenv != files['ovf-env.xml']:
                 # source was not walinux-agent's datadir, so we have to clean
                 # up so 'wait_for_files' doesn't return early due to stale data
@@ -593,7 +593,7 @@ def load_azure_ds_dir(source_dir):
     if not os.path.isfile(ovf_file):
         raise NonAzureDataSource("No ovf-env file found")
 
-    with open(ovf_file, "r") as fp:
+    with open(ovf_file, "rb") as fp:
         contents = fp.read()
 
     md, ud, cfg = read_azure_ovf(contents)
diff --git a/tests/unittests/helpers.py b/tests/unittests/helpers.py
index ce77af93..7516bd02 100644
--- a/tests/unittests/helpers.py
+++ b/tests/unittests/helpers.py
@@ -287,10 +287,11 @@ def populate_dir(path, files):
     if not os.path.exists(path):
         os.makedirs(path)
     for (name, content) in files.items():
-        with open(os.path.join(path, name), "w") as fp:
-            fp.write(content)
+        with open(os.path.join(path, name), "wb") as fp:
+            fp.write(content.encode('utf-8'))
             fp.close()
 
+
 try:
     skipIf = unittest.skipIf
 except AttributeError:
diff --git a/tests/unittests/test_datasource/test_azure.py b/tests/unittests/test_datasource/test_azure.py
index 965bce4b..38d70fcd 100644
--- a/tests/unittests/test_datasource/test_azure.py
+++ b/tests/unittests/test_datasource/test_azure.py
@@ -360,6 +360,12 @@ class TestAzureDataSource(TestCase):
         self.assertTrue(os.path.exists(ovf_env_path))
         self.assertEqual(xml, load_file(ovf_env_path))
 
+    def test_ovf_can_include_unicode(self):
+        xml = construct_valid_ovf_env(data={})
+        xml = u'\ufeff{0}'.format(xml)
+        dsrc = self._get_ds({'ovfcontent': xml})
+        dsrc.get_data()
+
     def test_existing_ovf_same(self):
         # waagent/SharedConfig left alone if found ovf-env.xml same as cached
         odata = {'UserData': b64e("SOMEUSERDATA")}
diff --git a/tox.ini b/tox.ini
index 686a1e7a..20ba6546 100644
--- a/tox.ini
+++ b/tox.ini
@@ -11,6 +11,8 @@ deps =
      nose
      pep8==1.5.7
      pyflakes
+setenv =
+    LC_ALL = C
 
 [testenv:py26]
 commands = nosetests {posargs:tests}
@@ -21,3 +23,5 @@ deps =
      nose
      pep8==1.5.7
      pyflakes
+setenv =
+    LC_ALL = C