From 782d89986fb5a234c65eddbd0765bf788d340706 Mon Sep 17 00:00:00 2001
From: Chris McDonough <chrism@plope.com>
Date: Fri, 2 Apr 2010 17:26:30 +0000
Subject: [PATCH] - Fix bug in serialization of non-Unicode values in the
 ``String`` class.

---
 CHANGES.txt          |  5 +++++
 colander/__init__.py | 15 +++++++++++++--
 colander/tests.py    |  8 ++++++++
 3 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/CHANGES.txt b/CHANGES.txt
index 41fd0b3..77e5d02 100644
--- a/CHANGES.txt
+++ b/CHANGES.txt
@@ -1,6 +1,11 @@
 Changes
 =======
 
+Next release
+------------
+
+- Fix bug in serialization of non-Unicode values in the ``String`` class.
+
 0.5.1 (2010-04-02)
 ------------------
 
diff --git a/colander/__init__.py b/colander/__init__.py
index 64c81b7..94277a9 100644
--- a/colander/__init__.py
+++ b/colander/__init__.py
@@ -525,6 +525,11 @@ class String(Type):
     which should be applied to object serialization.  It defaults to
     ``utf-8`` if not provided.
 
+    If a string (as opposed to a unicode object) is provided as a
+    value to either the serialize or deserialize method of this type,
+    it must be encoded with the type's encoding; an
+    :exc:`colander.Invalid` error will result if not.
+
     The subnodes of the :class:`colander.SchemaNode` that wraps
     this type are ignored.
     """
@@ -545,10 +550,16 @@ class String(Type):
 
     def serialize(self, node, value):
         try:
-            return unicode(value).encode(self.encoding or default_encoding)
+            encoding = self.encoding or default_encoding
+            if isinstance(value, unicode):
+                result = value.encode(encoding)
+            else:
+                # do validation here
+                result = unicode(value, encoding).encode(encoding)
+            return result
         except Exception, e:
             raise Invalid(node,
-                          '%r is cannot be serialized to str: %s' % (value, e))
+                          '%r cannot be serialized to str: %s' % (value, e))
 
 Str = String
 
diff --git a/colander/tests.py b/colander/tests.py
index 3a4c0d4..bbdf631 100644
--- a/colander/tests.py
+++ b/colander/tests.py
@@ -703,6 +703,14 @@ class TestString(unittest.TestCase):
         result = typ.serialize(node, uni)
         self.assertEqual(result, utf16)
 
+    def test_serialize_string_with_high_unresolveable_high_order_chars(self):
+        not_utf8 = '\xff\xfe\xf8\x00'
+        node = DummySchemaNode(None)
+        typ = self._makeOne('utf-8')
+        e = invalid_exc(typ.serialize, node, not_utf8)
+        self.failUnless('cannot be serialized to str' in e.msg)
+        
+
 class TestInteger(unittest.TestCase):
     def _makeOne(self):
         from colander import Integer