_is_unicode(): Use UnicodeType instead of the unicode builtin for

Python 2.1 compatibility.

Lib/email/Charset.py

@@ -7,8 +7,10 @@ except NameError:
     def _is_unicode(x):
         return 1==0
 else:
+    # Use UnicodeType instead of built-in unicode for Py2.1 compatibility
+    from types import UnicodeType
     def _is_unicode(x):
-        return isinstance(x, unicode)
+        return isinstance(x, UnicodeType)
     
 from email.Encoders import encode_7or8bit
 import email.base64MIME