Index: Lib/html/parser.py
===================================================================
--- Lib/html/parser.py	(revision 86342)
+++ Lib/html/parser.py	(working copy)
@@ -362,6 +362,8 @@
 
     # Internal -- helper to remove special character quoting
     entitydefs = None
+    # many web pages accidentally use MS code page 1252 characters instead of iso-8859-1 or unicode characters
+    cp1252_to_unicode = {92:2019, 145:8216, 146:8217, 147:8220, 148:8221, 149:8226, 150:8211, 151:8212, 152:732, 153:8482}    
     def unescape(self, s):
         if '&' not in s:
             return s
@@ -373,6 +375,8 @@
                     c = int(s[1:], 16)
                 else:
                     c = int(s)
+                if c in HTMLParser.cp1252_to_unicode:
+                    c = HTMLParser.cp1252_to_unicode[c]                    
                 return chr(c)
             else:
                 # Cannot use name2codepoint directly, because HTMLParser
Index: Lib/test/test_htmlparser.py
===================================================================
--- Lib/test/test_htmlparser.py	(revision 86342)
+++ Lib/test/test_htmlparser.py	(working copy)
@@ -320,10 +320,17 @@
                 ("starttag", "html", [("foo", "\u20AC&aa&unsupported;")])
                 ])
 
+    def test_unescaping(self):
+        parser = html.parser.HTMLParser()
+        escaped = "<p>There&#146;s the C&ocirc;te</p>"
+        correct_unescaped = "<p>There"+chr(8217) + "s the C" + chr(0x00f4) \
+                            + "te</p>"
+        actual_unescaped = parser.unescape(escaped)
+        if actual_unescaped != correct_unescaped:
+            self.fail ("failed to unescape properly")
 
 def test_main():
     support.run_unittest(HTMLParserTestCase)
 
-
 if __name__ == "__main__":
     test_main()