Fixed #27333 -- Prevented BASE64 encoding in message.as_string() on Python 3

Thanks Tim Graham for the review.
This commit is contained in:
Claude Paroz 2016-10-11 20:53:26 +02:00
parent bd7237d7ec
commit 458e2fbfcc
2 changed files with 13 additions and 20 deletions

View File

@ -211,31 +211,20 @@ class SafeMIMEText(MIMEMixin, MIMEText):
def __init__(self, _text, _subtype='plain', _charset=None): def __init__(self, _text, _subtype='plain', _charset=None):
self.encoding = _charset self.encoding = _charset
if _charset == 'utf-8': MIMEText.__init__(self, _text, _subtype=_subtype, _charset=_charset)
# Unfortunately, Python doesn't yet pass a Charset instance as
# MIMEText init parameter to set_payload().
# http://bugs.python.org/issue27445
# We do it manually and trigger re-encoding of the payload.
if six.PY3 and isinstance(_text, bytes):
# Sniffing encoding would fail with bytes content in MIMEText.__init__.
_text = _text.decode('utf-8')
MIMEText.__init__(self, _text, _subtype, None)
del self['Content-Transfer-Encoding']
has_long_lines = any(len(l) > RFC5322_EMAIL_LINE_LENGTH_LIMIT for l in _text.splitlines())
# Quoted-Printable encoding has the side effect of shortening long
# lines, if any (#22561).
self.set_payload(_text, utf8_charset_qp if has_long_lines else utf8_charset)
self.replace_header('Content-Type', 'text/%s; charset="%s"' % (_subtype, _charset))
elif _charset is None:
# the default value of '_charset' is 'us-ascii' on Python 2
MIMEText.__init__(self, _text, _subtype)
else:
MIMEText.__init__(self, _text, _subtype, _charset)
def __setitem__(self, name, val): def __setitem__(self, name, val):
name, val = forbid_multi_line_headers(name, val, self.encoding) name, val = forbid_multi_line_headers(name, val, self.encoding)
MIMEText.__setitem__(self, name, val) MIMEText.__setitem__(self, name, val)
def set_payload(self, payload, charset=None):
if charset == 'utf-8':
has_long_lines = any(len(l) > RFC5322_EMAIL_LINE_LENGTH_LIMIT for l in payload.splitlines())
# Quoted-Printable encoding has the side effect of shortening long
# lines, if any (#22561).
charset = utf8_charset_qp if has_long_lines else utf8_charset
MIMEText.set_payload(self, payload, charset=charset)
class SafeMIMEMultipart(MIMEMixin, MIMEMultipart): class SafeMIMEMultipart(MIMEMixin, MIMEMultipart):

View File

@ -572,6 +572,8 @@ class MailTests(HeadersCheckMixin, SimpleTestCase):
) )
s = msg.message().as_bytes() s = msg.message().as_bytes()
self.assertIn(b'Content-Transfer-Encoding: 8bit', s) self.assertIn(b'Content-Transfer-Encoding: 8bit', s)
s = msg.message().as_string()
self.assertIn(str('Content-Transfer-Encoding: 8bit'), s)
msg = EmailMessage( msg = EmailMessage(
'Subject', 'Body with non latin characters: А Б В Г Д Е Ж Ѕ З И І К Л М Н О П.', 'bounce@example.com', 'Subject', 'Body with non latin characters: А Б В Г Д Е Ж Ѕ З И І К Л М Н О П.', 'bounce@example.com',
@ -579,6 +581,8 @@ class MailTests(HeadersCheckMixin, SimpleTestCase):
) )
s = msg.message().as_bytes() s = msg.message().as_bytes()
self.assertIn(b'Content-Transfer-Encoding: 8bit', s) self.assertIn(b'Content-Transfer-Encoding: 8bit', s)
s = msg.message().as_string()
self.assertIn(str('Content-Transfer-Encoding: 8bit'), s)
def test_dont_base64_encode_message_rfc822(self): def test_dont_base64_encode_message_rfc822(self):
# Ticket #18967 # Ticket #18967