gh-150771: Use output charset in set_text_content for shift_jis/euc-jp

Encode the payload with the charset output mapping (iso-2022-jp) when
set_content is called with shift_jis or euc-jp, instead of patching
serialization in body_encode and set_payload. Reverts those changes.

Author: bhuvi27

Lib/email/charset.py

@@ -16,7 +16,6 @@
 import email.quoprimime
 
 from email import errors
-from email import utils
 from email.encoders import encode_7or8bit
 
 
@@ -439,12 +438,5 @@ def body_encode(self, string):
             return email.quoprimime.body_encode(string)
         else:
             if isinstance(string, str):
-                if utils._has_surrogates(string):
-                    string = string.encode('ascii', 'surrogateescape')
-                    if self.input_charset != self.output_charset:
-                        string = (string.decode(self.input_codec)
-                                  .encode(self.output_codec))
-                    string = string.decode('ascii', 'surrogateescape')
-                else:
-                    string = string.encode(self.output_charset).decode('ascii')
+                string = string.encode(self.output_charset).decode('ascii')
             return string

Lib/email/contentmanager.py

@@ -174,7 +174,8 @@ def set_text_content(msg, string, subtype="plain", charset='utf-8', cte=None,
                      params=None, headers=None):
     _prepare_set(msg, 'text', subtype, headers)
 
-    charset = email.charset.Charset(charset).input_charset
+    cs = email.charset.Charset(charset)
+    charset = cs.output_charset
     cte, payload = _encode_text(string, charset, cte, msg.policy)
     msg.set_payload(payload)
     msg.set_param('charset', charset, replace=True)

Lib/email/message.py

@@ -352,9 +352,7 @@ def set_payload(self, payload, charset=None):
                 return
             if not isinstance(charset, Charset):
                 charset = Charset(charset)
-            if not utils._has_surrogates(payload):
-                payload = payload.encode(charset.output_charset,
-                                          'surrogateescape')
+            payload = payload.encode(charset.output_charset, 'surrogateescape')
         if hasattr(payload, 'decode'):
             self._payload = payload.decode('ascii', 'surrogateescape')
         else:

Lib/test/test_email/test_contentmanager.py

@@ -366,11 +366,9 @@ def test_set_text_charset_shift_jis(self):
         m = self._make_message()
         content = "\u65e5\u672c\u8a9e\n"
         raw_data_manager.set_content(m, content, charset='shift_jis')
-        self.assertEqual(m['Content-Type'], 'text/plain; charset="shift_jis"')
-        self.assertEqual(m['Content-Transfer-Encoding'], '8bit')
-        self.assertEqual(m.get_payload(decode=True), content.encode('shift_jis'))
+        self.assertEqual(m['Content-Type'], 'text/plain; charset="iso-2022-jp"')
+        self.assertEqual(m.get_payload(decode=True), content.encode('iso-2022-jp'))
         self.assertEqual(m.get_content(), content)
-        # Serialization converts the payload to iso-2022-jp for output.
         self.assertEqual(str(m), textwrap.dedent("""\
             Content-Type: text/plain; charset="iso-2022-jp"
             Content-Transfer-Encoding: 7bit
@@ -382,11 +380,9 @@ def test_set_text_charset_euc_jp(self):
         m = self._make_message()
         content = "\u65e5\u672c\u8a9e\n"
         raw_data_manager.set_content(m, content, charset='euc-jp')
-        self.assertEqual(m['Content-Type'], 'text/plain; charset="euc-jp"')
-        self.assertEqual(m['Content-Transfer-Encoding'], '8bit')
-        self.assertEqual(m.get_payload(decode=True), content.encode('euc-jp'))
+        self.assertEqual(m['Content-Type'], 'text/plain; charset="iso-2022-jp"')
+        self.assertEqual(m.get_payload(decode=True), content.encode('iso-2022-jp'))
         self.assertEqual(m.get_content(), content)
-        # Serialization converts the payload to iso-2022-jp for output.
         self.assertEqual(str(m), textwrap.dedent("""\
             Content-Type: text/plain; charset="iso-2022-jp"
             Content-Transfer-Encoding: 7bit

Misc/NEWS.d/next/Library/2026-06-09-12-00-00.gh-issue-150771.K7mNx2.rst

@@ -1,3 +1,4 @@
-Fix serialization of :mod:`email` messages using ``shift_jis`` or ``euc-jp``
-charsets.  Converting surrogate-escaped payloads to the required
-``iso-2022-jp`` output charset no longer raises :exc:`UnicodeEncodeError`.
+Fix :mod:`email` messages created with ``shift_jis`` or ``euc-jp`` charsets.
+:func:`email.contentmanager.set_text_content` now stores the payload using
+the output charset (``iso-2022-jp``) so :func:`str` on the message no longer
+raises :exc:`UnicodeEncodeError`.