Rev 4238: (Jelmer) Make function for escaping invalid XML characters public. in file:///home/pqm/archives/thelove/bzr/%2Btrunk/

[Canonical.com Patch Queue Manager]

At file:///home/pqm/archives/thelove/bzr/%2Btrunk/

------------------------------------------------------------
revno: 4238
revision-id: pqm at pqm.ubuntu.com-20090402135806-w2vi86mmi1s39b4h
parent: pqm at pqm.ubuntu.com-20090402093220-mqwls6dsncdfegbq
parent: jelmer at samba.org-20090331203433-ncsd0sar9rc7wkqd
committer: Canonical.com Patch Queue Manager <pqm at pqm.ubuntu.com>
branch nick: +trunk
timestamp: Thu 2009-04-02 14:58:06 +0100
message:
  (Jelmer) Make function for escaping invalid XML characters public.
modified:
  bzrlib/commit.py               commit.py-20050511101309-79ec1a0168e0e825
  bzrlib/xml_serializer.py       xml.py-20050309040759-57d51586fdec365d
    ------------------------------------------------------------
    revno: 4222.1.1
    revision-id: jelmer at samba.org-20090331203433-ncsd0sar9rc7wkqd
    parent: pqm at pqm.ubuntu.com-20090331122525-77wqe7gqpr2gcl7u
    committer: Jelmer Vernooij <jelmer at samba.org>
    branch nick: escape-commit-message
    timestamp: Tue 2009-03-31 22:34:33 +0200
    message:
      Make function for escaping invalid XML characters public.
    modified:
      bzrlib/commit.py               commit.py-20050511101309-79ec1a0168e0e825
      bzrlib/xml_serializer.py       xml.py-20050309040759-57d51586fdec365d
=== modified file 'bzrlib/commit.py'
--- a/bzrlib/commit.py	2009-03-27 04:36:58 +0000
+++ b/bzrlib/commit.py	2009-03-31 20:34:33 +0000
@@ -62,6 +62,7 @@
     revision,
     trace,
     tree,
+    xml_serializer,
     )
 from bzrlib.branch import Branch
 import bzrlib.config
@@ -607,14 +608,7 @@
         # serialiser not by commit. Then we can also add an unescaper
         # in the deserializer and start roundtripping revision messages
         # precisely. See repository_implementations/test_repository.py
-
-        # Python strings can include characters that can't be
-        # represented in well-formed XML; escape characters that
-        # aren't listed in the XML specification
-        # (http://www.w3.org/TR/REC-xml/#NT-Char).
-        self.message, escape_count = re.subn(
-            u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
-            lambda match: match.group(0).encode('unicode_escape'),
+        self.message, escape_count = xml_serializer.escape_invalid_chars(
             self.message)
         if escape_count:
             self.reporter.escaped(escape_count, self.message)

=== modified file 'bzrlib/xml_serializer.py'
--- a/bzrlib/xml_serializer.py	2009-03-23 14:59:43 +0000
+++ b/bzrlib/xml_serializer.py	2009-03-31 20:34:33 +0000
@@ -185,3 +185,19 @@
 format_registry.register_lazy('6', 'bzrlib.xml6', 'serializer_v6')
 format_registry.register_lazy('7', 'bzrlib.xml7', 'serializer_v7')
 format_registry.register_lazy('8', 'bzrlib.xml8', 'serializer_v8')
+
+
+def escape_invalid_chars(message):
+    """Escape the XML-invalid characters in a commit message.
+
+    :param message: Commit message to escape
+    :param count: Number of characters that were escaped
+    """
+    # Python strings can include characters that can't be
+    # represented in well-formed XML; escape characters that
+    # aren't listed in the XML specification
+    # (http://www.w3.org/TR/REC-xml/#NT-Char).
+    return re.subn(u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
+            lambda match: match.group(0).encode('unicode_escape'),
+            message)
+