Newline embedded in email RFC2047 encoding raises exception when parsed #114906
Description
Bug report
Bug description:
I came across some messages where the sender had embedded a newline in the From: header's display string. This was (ostensibly) a legitimate sender, possibly hoping to get more real estate for their message in the recipient's inbox listing or something, or just making a configuration mistake.
Unfortunately, this crashes Python's email parser with policy=default. (The legacy parser works fine, simply because it doesn't attempt to unpeel RFC2047 encoding by default.)
>>> from email import message_from_bytes
>>> from email.policy import default
>>> message = message_from_bytes(b'From: =?UTF-8?Q?=0AMy_self?= <me@example.org>\r\n\r\nHello\r\n', policy=default)
>>> message.items()
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/Users/fsc-eriker/git/cpython/email/message.py", line 491, in items
return [(k, self.policy.header_fetch_parse(k, v))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/fsc-eriker/git/cpython/email/message.py", line 491, in <listcomp>
return [(k, self.policy.header_fetch_parse(k, v))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/fsc-eriker/git/cpython/email/policy.py", line 163, in header_fetch_parse
return self.header_factory(name, value)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/fsc-eriker/git/cpython/email/headerregistry.py", line 604, in __call__
return self[name](name, value)
^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/fsc-eriker/git/cpython/email/headerregistry.py", line 192, in __new__
cls.parse(value, kwds)
File "/Users/fsc-eriker/git/cpython/email/headerregistry.py", line 346, in parse
[Address(mb.display_name or '',
File "/Users/fsc-eriker/git/cpython/email/headerregistry.py", line 346, in <listcomp>
[Address(mb.display_name or '',
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/fsc-eriker/git/cpython/email/headerregistry.py", line 33, in __init__
raise ValueError("invalid arguments; address parts cannot contain CR or LF")
ValueError: invalid arguments; address parts cannot contain CR or LF
>>> While the error message is correct, it should probably not raise an exception; perhaps instead register a defect?
I have tested this on 3.11 out of the box and with the current sources from the cpython Github repo; but I would expect it to manifest on all versions of the modern email module and all platforms.
CPython versions tested on:
3.11
Operating systems tested on:
macOS