#19063: fix set_payload handling of non-ASCII string input.
This version of the fix raises an error instead of accepting the invalid
input (ie: if a non-ASCII string is used but no charset is specified).
diff --git a/Lib/email/message.py b/Lib/email/message.py
index ce673b0..9d295fc 100644
--- a/Lib/email/message.py
+++ b/Lib/email/message.py
@@ -301,9 +301,23 @@
Optional charset sets the message's default character set. See
set_charset() for details.
"""
- if isinstance(payload, bytes):
- payload = payload.decode('ascii', 'surrogateescape')
- self._payload = payload
+ if hasattr(payload, 'encode'):
+ if charset is None:
+ try:
+ payload.encode('ascii', 'surrogateescape')
+ except UnicodeError:
+ raise TypeError("charset argument must be specified"
+ " when non-ASCII characters are used in the"
+ " payload") from None
+ self._payload = payload
+ return
+ if not isinstance(charset, Charset):
+ charset = Charset(charset)
+ payload = payload.encode(charset.output_charset)
+ if hasattr(payload, 'decode'):
+ self._payload = payload.decode('ascii', 'surrogateescape')
+ else:
+ self._payload = payload
if charset is not None:
self.set_charset(charset)
@@ -342,7 +356,7 @@
try:
cte(self)
except TypeError:
- self._payload = charset.body_encode(self._payload)
+ self._payload = charset.body_encode(self.get_payload(decode=True))
self.add_header('Content-Transfer-Encoding', cte)
def get_charset(self):