More robust conversion to unicode

Fixes #72.

Author: jessevdk

cldoc/utf8.py

@@ -5,7 +5,10 @@
     def makeutf8(s):
         if not isinstance(s, unicode):
             if hasattr(s, '__unicode__'):
-                return unicode(s)
+                if isinstance(s, str) or isinstance(s, buffer):
+                    return unicode(s, 'utf-8')
+                else:
+                    return unicode(s)
 
             return str(s).decode('utf-8')
 
@@ -25,8 +28,8 @@ def makeutf8(s):
 string = basecls
 
 class utf8(string):
-    def __init__(self, s):
-        super(utf8, self).__init__(makeutf8(s))
+    def __new__(cls, s):
+        return super(utf8, cls).__new__(cls, makeutf8(s))
 
     def __str__(self):
         if not isinstance(self, str):

tests/input/utf8.hh

@@ -0,0 +1,4 @@
+/* Copyright © */
+class A
+{
+};

tests/output/utf8-A.xml

@@ -0,0 +1,4 @@
+<?xml version='1.0' encoding='utf-8'?>
+<class id="A" name="A" xmlns="http://jessevdk.github.com/cldoc/1.0">
+  <brief>Copyright ©</brief>
+</class>

tests/output/utf8-index.xml

@@ -0,0 +1,6 @@
+<?xml version='1.0' encoding='utf-8'?>
+<index xmlns="http://jessevdk.github.com/cldoc/1.0">
+  <class name="A" ref="A#A">
+    <brief>Copyright ©</brief>
+  </class>
+</index>