Index: Lib/encodings/utf_8_sig.py
===================================================================
RCS file: Lib/encodings/utf_8_sig.py
diff -N Lib/encodings/utf_8_sig.py
--- /dev/null	1 Jan 1970 00:00:00 -0000
+++ Lib/encodings/utf_8_sig.py	5 Apr 2005 20:25:28 -0000
@@ -0,0 +1,57 @@
+""" Python 'utf-8-sig' Codec
+This work similar to UTF-8 with the following changes:
+
+* On encoding/writing a UTF-8 encoded BOM will be prepended/written as the
+  first three bytes.
+
+* On decoding/reading if the first three bytes are an UTF-8 encoded BOM, these
+  bytes will be skipped.
+"""
+import codecs
+
+### Codec APIs
+
+def encode(input, errors='strict'):
+    return (codecs.BOM_UTF8 + codecs.utf_8_encode(input, errors)[0], len(input))
+
+def decode(input, errors='strict'):
+    prefix = 0
+    if input.startswith(codecs.BOM_UTF8):
+        input = input[3:]
+        prefix = 3
+    (output, consumed) = codecs.utf_8_decode(input, errors, True)
+    return (output, consumed+prefix)
+
+class StreamWriter(codecs.StreamWriter):
+    def reset(self):
+        codecs.StreamWriter.reset(self)
+        try:
+            del self.encode
+        except AttributeError:
+            pass
+
+    def encode(self, input, errors='strict'):
+        self.encode = codecs.utf_8_encode
+        return encode(input, errors)
+
+class StreamReader(codecs.StreamReader):
+    def reset(self):
+        codecs.StreamReader.reset(self)
+        try:
+            del self.decode
+        except AttributeError:
+            pass
+
+    def decode(self, input, errors='strict'):
+        if len(input) < 3 and codecs.BOM_UTF8.startswith(input):
+            # not enough data to decide if this is a BOM
+            # => try again on the next call
+            return (u"", 0)
+        self.decode = codecs.utf_8_decode
+        return decode(input, errors)
+
+### encodings module API
+
+def getregentry():
+
+    return (encode,decode,StreamReader,StreamWriter)
Index: Lib/test/test_codecs.py
===================================================================
RCS file: /cvsroot/python/python/dist/src/Lib/test/test_codecs.py,v
retrieving revision 1.22
diff -u -r1.22 test_codecs.py
--- Lib/test/test_codecs.py	4 Apr 2005 21:38:47 -0000	1.22
+++ Lib/test/test_codecs.py	5 Apr 2005 20:25:30 -0000
@@ -315,6 +315,32 @@
             ]
         )
 
+class UTF8SigTest(ReadTest):
+    encoding = "utf-8-sig"
+
+    def test_partial(self):
+        self.check_partial(
+            u"u\feff\x00\xff\u07ff\u0800\uffff",
+            [
+                u"",
+                u"",
+                u"",
+                u"",
+                u"",
+                u"\ufeff",
+                u"\ufeff\x00",
+                u"\ufeff\x00\xff",
+                u"\ufeff\x00\xff",
+                u"\ufeff\x00\xff\u07ff",
+                u"\ufeff\x00\xff\u07ff",
+                u"\ufeff\x00\xff\u07ff",
+                u"\ufeff\x00\xff\u07ff\u0800",
+                u"\ufeff\x00\xff\u07ff\u0800",
+                u"\ufeff\x00\xff\u07ff\u0800",
+                u"\ufeff\x00\xff\u07ff\u0800\uffff",
+            ]
+        )
+
 class EscapeDecodeTest(unittest.TestCase):
     def test_empty_escape_decode(self):
         self.assertEquals(codecs.escape_decode(""), ("", 0))
@@ -762,6 +788,7 @@
     "utf_16_le",
     "utf_7",
     "utf_8",
+    "utf_8_sig",
 ]
 
 if hasattr(codecs, "mbcs_encode"):