user_data: fix error when user-data is not utf-8 decodable

when user-data was not decodable, cloud-init would raise exception.
This also changes the signature of user_data.convert_string.
The 'headers' argument was never used, and woudl have been broken
if it was, as it was expected to be a dictionary but then was
passed in with *headers.

LP: #1532072

3 files changed, 34 insertions, 12 deletions

diff --git a/ChangeLog b/ChangeLog
index 8edf3249..fa5e7df4 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -125,6 +125,7 @@
  - Remove trailing dot from GCE metadata URL (LP: #1581200) [Phil Roche]
  - support network rendering to sysconfig (for centos and RHEL)
  - write_files: if no permissions are given, just use default without warn.
+ - user_data: fix error when user-data is not utf-8 decodable (LP: #1532072)
 
 0.7.6:
  - open 0.7.6
diff --git a/cloudinit/user_data.py b/cloudinit/user_data.py
index f0631906..393bf0bb 100644
--- a/cloudinit/user_data.py
+++ b/cloudinit/user_data.py
@@ -334,19 +334,23 @@ def is_skippable(part):
 
 
 # Coverts a raw string into a mime message
-def convert_string(raw_data, headers=None):
+def convert_string(raw_data, content_type=NOT_MULTIPART_TYPE):
     if not raw_data:
         raw_data = ''
-    if not headers:
-        headers = {}
-    data = util.decode_binary(util.decomp_gzip(raw_data))
-    if "mime-version:" in data[0:4096].lower():
-        msg = util.message_from_string(data)
-        for (key, val) in headers.items():
-            _replace_header(msg, key, val)
-    else:
-        mtype = headers.get(CONTENT_TYPE, NOT_MULTIPART_TYPE)
-        maintype, subtype = mtype.split("/", 1)
-        msg = MIMEBase(maintype, subtype, *headers)
+
+    def create_binmsg(data, content_type):
+        maintype, subtype = content_type.split("/", 1)
+        msg = MIMEBase(maintype, subtype)
         msg.set_payload(data)
+        return msg
+
+    try:
+        data = util.decode_binary(util.decomp_gzip(raw_data))
+        if "mime-version:" in data[0:4096].lower():
+            msg = util.message_from_string(data)
+        else:
+            msg = create_binmsg(data, content_type)
+    except UnicodeDecodeError:
+        msg = create_binmsg(raw_data, content_type)
+
     return msg
diff --git a/tests/unittests/test_data.py b/tests/unittests/test_data.py
index 1923e2af..13db8a4c 100644
--- a/tests/unittests/test_data.py
+++ b/tests/unittests/test_data.py
@@ -557,3 +557,20 @@ class TestUDProcess(helpers.ResourceUsingTestCase):
         ud_proc = ud.UserDataProcessor(self.getCloudPaths())
         message = ud_proc.process(msg)
         self.assertTrue(count_messages(message) == 1)
+
+
+class TestConvertString(helpers.TestCase):
+    def test_handles_binary_non_utf8_decodable(self):
+        blob = b'\x32\x99'
+        msg = ud.convert_string(blob)
+        self.assertEqual(blob, msg.get_payload(decode=True))
+
+    def test_handles_binary_utf8_decodable(self):
+        blob = b'\x32\x32'
+        msg = ud.convert_string(blob)
+        self.assertEqual(blob, msg.get_payload(decode=True))
+
+    def test_handle_headers(self):
+        text = "hi mom"
+        msg = ud.convert_string(text)
+        self.assertEqual(text, msg.get_payload(decode=False))