gh-136702: Deprecate passing non-ascii *encoding* (str) to encodings.normalize_encoding (#140030)

Closes #136702

Author: StanFromIreland

Doc/deprecations/pending-removal-in-3.17.rst

@@ -23,6 +23,12 @@ Pending removal in Python 3.17
     (Contributed by Shantanu Jain in :gh:`91896`.)
 
 
+* :mod:`encodings`:
+
+  - Passing non-ascii *encoding* names to :func:`encodings.normalize_encoding`
+    is deprecated and scheduled for removal in Python 3.17.
+    (Contributed by Stan Ulbrych in :gh:`136702`)
+
 * :mod:`typing`:
 
   - Before Python 3.14, old-style unions were implemented using the private class

Lib/email/_header_value_parser.py

@@ -796,6 +796,10 @@ def params(self):
                         value = urllib.parse.unquote(value, encoding='latin-1')
                     else:
                         try:
+                            # Explicitly look up the codec for warning generation, see gh-140030
+                            # Can be removed in 3.17
+                            import codecs
+                            codecs.lookup(charset)
                             value = value.decode(charset, 'surrogateescape')
                         except (LookupError, UnicodeEncodeError):
                             # XXX: there should really be a custom defect for

Lib/email/utils.py

@@ -460,6 +460,10 @@ def collapse_rfc2231_value(value, errors='replace',
         charset = fallback_charset
     rawbytes = bytes(text, 'raw-unicode-escape')
     try:
+        # Explicitly look up the codec for warning generation, see gh-140030
+        # Can be removed in 3.17
+        import codecs
+        codecs.lookup(charset)
         return str(rawbytes, charset, errors)
     except LookupError:
         # charset is not a known codec.

Lib/encodings/__init__.py

@@ -26,7 +26,7 @@
 
 (c) Copyright CNRI, All Rights Reserved. NO WARRANTY.
 
-"""#"
+"""
 
 import codecs
 import sys
@@ -56,6 +56,12 @@ def normalize_encoding(encoding):
     if isinstance(encoding, bytes):
         encoding = str(encoding, "ascii")
 
+    if not encoding.isascii():
+        import warnings
+        warnings.warn(
+            "Support for non-ascii encoding names will be removed in 3.17",
+            DeprecationWarning, stacklevel=2)
+
     return _normalize_encoding(encoding)
 
 def search_function(encoding):

Lib/test/test_codecs.py

@@ -3886,22 +3886,26 @@ def search_function(encoding):
         self.assertEqual(codecs.lookup('TEST.AAA 8'), ('test.aaa-8', 2, 3, 4))
         self.assertEqual(codecs.lookup('TEST.AAA---8'), ('test.aaa---8', 2, 3, 4))
         self.assertEqual(codecs.lookup('TEST.AAA   8'), ('test.aaa---8', 2, 3, 4))
-        self.assertEqual(codecs.lookup('TEST.AAA\xe9\u20ac-8'), ('test.aaa\xe9\u20ac-8', 2, 3, 4))
         self.assertEqual(codecs.lookup('TEST.AAA.8'), ('test.aaa.8', 2, 3, 4))
         self.assertEqual(codecs.lookup('TEST.AAA...8'), ('test.aaa...8', 2, 3, 4))
+        with self.assertWarns(DeprecationWarning):
+            self.assertEqual(codecs.lookup('TEST.AAA\xe9\u20ac-8'), ('test.aaa\xe9\u20ac-8', 2, 3, 4))
 
     def test_encodings_normalize_encoding(self):
-        # encodings.normalize_encoding() ignores non-ASCII characters.
         normalize = encodings.normalize_encoding
         self.assertEqual(normalize('utf_8'), 'utf_8')
-        self.assertEqual(normalize('utf\xE9\u20AC\U0010ffff-8'), 'utf_8')
         self.assertEqual(normalize('utf   8'), 'utf_8')
         # encodings.normalize_encoding() doesn't convert
         # characters to lower case.
         self.assertEqual(normalize('UTF 8'), 'UTF_8')
         self.assertEqual(normalize('utf.8'), 'utf.8')
         self.assertEqual(normalize('utf...8'), 'utf...8')
 
+        # Non-ASCII *encoding* is deprecated.
+        with self.assertWarnsRegex(DeprecationWarning,
+                "Support for non-ascii encoding names will be removed in 3.17"):
+            self.assertEqual(normalize('utf\xE9\u20AC\U0010ffff-8'), 'utf_8')
+
 
 if __name__ == "__main__":
     unittest.main()

Lib/test/test_email/test_email.py

@@ -5738,7 +5738,8 @@ def test_rfc2231_bad_character_in_encoding(self):
 
 """
         msg = email.message_from_string(m)
-        self.assertEqual(msg.get_filename(), 'myfile.txt')
+        with self.assertWarns(DeprecationWarning):
+            self.assertEqual(msg.get_filename(), 'myfile.txt')
 
     def test_rfc2231_single_tick_in_filename_extended(self):
         eq = self.assertEqual

Lib/test/test_email/test_headerregistry.py

@@ -247,7 +247,15 @@ def content_type_as_value(self,
         decoded =  args[2] if l>2 and args[2] is not DITTO else source
         header = 'Content-Type:' + ' ' if source else ''
         folded = args[3] if l>3 else header + decoded + '\n'
-        h = self.make_header('Content-Type', source)
+        # Both rfc2231 test cases with utf-8%E2%80%9D raise warnings,
+        # clear encoding cache to ensure test isolation.
+        if 'utf-8%E2%80%9D' in source and 'ascii' not in source:
+            import encodings
+            encodings._cache.clear()
+            with self.assertWarns(DeprecationWarning):
+                h = self.make_header('Content-Type', source)
+        else:
+            h = self.make_header('Content-Type', source)
         self.assertEqual(h.content_type, content_type)
         self.assertEqual(h.maintype, maintype)
         self.assertEqual(h.subtype, subtype)

Misc/NEWS.d/next/Library/2025-10-13-11-25-41.gh-issue-136702.uvLGK1.rst

@@ -0,0 +1,3 @@
+:mod:`encodings`: Deprecate passing a non-ascii *encoding* name to
+:func:`encodings.normalize_encoding` and schedule removal of support for
+Python 3.17.