bpo-43323: Fix UnicodeEncodeError in the email module (GH-32137)

It was raised if the charset itself contains characters not encodable
in UTF-8 (in particular \udcxx characters representing non-decodable
bytes in the source).
(cherry picked from commit e91dee8)

Co-authored-by: Serhiy Storchaka <[email protected]>

Author: miss-islington

Lib/email/_encoded_words.py

@@ -179,15 +179,15 @@ def decode(ew):
     # Turn the CTE decoded bytes into unicode.
     try:
         string = bstring.decode(charset)
-    except UnicodeError:
+    except UnicodeDecodeError:
         defects.append(errors.UndecodableBytesDefect("Encoded word "
-            "contains bytes not decodable using {} charset".format(charset)))
+            f"contains bytes not decodable using {charset!r} charset"))
         string = bstring.decode(charset, 'surrogateescape')
-    except LookupError:
+    except (LookupError, UnicodeEncodeError):
         string = bstring.decode('ascii', 'surrogateescape')
         if charset.lower() != 'unknown-8bit':
-            defects.append(errors.CharsetError("Unknown charset {} "
-                "in encoded word; decoded as unknown bytes".format(charset)))
+            defects.append(errors.CharsetError(f"Unknown charset {charset!r} "
+                f"in encoded word; decoded as unknown bytes"))
     return string, charset, lang, defects
 
 

Lib/email/_header_value_parser.py

@@ -781,7 +781,7 @@ def params(self):
                     else:
                         try:
                             value = value.decode(charset, 'surrogateescape')
-                        except LookupError:
+                        except (LookupError, UnicodeEncodeError):
                             # XXX: there should really be a custom defect for
                             # unknown character set to make it easy to find,
                             # because otherwise unknown charset is a silent

Lib/test/test_email/test__encoded_words.py

@@ -130,6 +130,13 @@ def test_unknown_charset(self):
                    # XXX Should this be a new Defect instead?
                    defects = [errors.CharsetError])
 
+    def test_invalid_character_in_charset(self):
+        self._test('=?utf-8\udce2\udc80\udc9d?q?foo=ACbar?=',
+                   b'foo\xacbar'.decode('ascii', 'surrogateescape'),
+                   charset = 'utf-8\udce2\udc80\udc9d',
+                   # XXX Should this be a new Defect instead?
+                   defects = [errors.CharsetError])
+
     def test_q_nonascii(self):
         self._test('=?utf-8?q?=C3=89ric?=',
                    'Éric',

Lib/test/test_email/test_email.py

@@ -5323,6 +5323,15 @@ def test_rfc2231_unknown_encoding(self):
 Content-Transfer-Encoding: 8bit
 Content-Disposition: inline; filename*=X-UNKNOWN''myfile.txt
 
+"""
+        msg = email.message_from_string(m)
+        self.assertEqual(msg.get_filename(), 'myfile.txt')
+
+    def test_rfc2231_bad_character_in_encoding(self):
+        m = """\
+Content-Transfer-Encoding: 8bit
+Content-Disposition: inline; filename*=utf-8\udce2\udc80\udc9d''myfile.txt
+
 """
         msg = email.message_from_string(m)
         self.assertEqual(msg.get_filename(), 'myfile.txt')

Lib/test/test_email/test_headerregistry.py

@@ -698,6 +698,18 @@ def content_type_as_value(self,
             " charset*=unknown-8bit''utf-8%E2%80%9D\n",
             ),
 
+        'rfc2231_nonascii_in_charset_of_charset_parameter_value': (
+            "text/plain; charset*=utf-8”''utf-8%E2%80%9D",
+            'text/plain',
+            'text',
+            'plain',
+            {'charset': 'utf-8”'},
+            [],
+            'text/plain; charset="utf-8”"',
+            "Content-Type: text/plain;"
+            " charset*=utf-8''utf-8%E2%80%9D\n",
+            ),
+
         'rfc2231_encoded_then_unencoded_segments': (
             ('application/x-foo;'
                 '\tname*0*="us-ascii\'en-us\'My";'

Misc/NEWS.d/next/Library/2022-03-27-12-40-16.bpo-43323.9mFPuI.rst

@@ -0,0 +1,2 @@
+Fix errors in the :mod:`email` module if the charset itself contains
+undecodable/unencodable characters.