encoding.to_message ignores multiline headers when there's non-ascii chars

[EmilStenstrom]

>>> msg1 = email.message_from_string("To: Person1 <[email protected]>, Person2\n\t<[email protected]>")
>>> msg2 = email.message_from_string("To: =?utf-8?q?Emil_Stenstr=C3=B6m?= <[email protected]>, Person2\n\t<[email protected]>")
>>> print encoding.to_message(encoding.from_message(msg1))
From nobody Tue Dec  3 12:42:05 2013
MIME-Version: 1.0
To: Person1 <person1@domain.com>, Person2
    <person2@domain.com>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
>>> print encoding.to_message(encoding.from_message(msg2))
From nobody Tue Dec  3 12:42:09 2013
MIME-Version: 1.0
To: "=?utf-8?q?Emil_Stenstr=C3=B6m?=" <person1@domain.com>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

Note that the from_message step seems to work fine:

>>> print encoding.from_message(msg2)["to"]
Emil Stenström <person1@domain.com>, Person2
    <person2@domain.com>

[EmilStenstrom]

The problem is on properly_encode_header. email.parseaddr just parses and returns the first address if there are more than one. Fix is to split the address on "," and then use email.getaddresses instead. Pull request on the way.

[EmilStenstrom]

After this fix, the test case I gave above works fine:

>>> msg2 = email.message_from_string("To: =?utf-8?q?Emil_Stenstr=C3=B6m?= <[email protected]>, Person2\n\t<[email protected]>")
>>> print encoding.to_message(encoding.from_message(msg2))
From nobody Tue Dec  3 13:52:03 2013
MIME-Version: 1.0
To: "=?utf-8?q?Emil_Stenstr=C3=B6m?=" <person1@domain.com>,
 "=?utf-8?q?Person2?=" <person2@domain.com>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

[EmilStenstrom]

@zedshaw Any chance to get this merged so I can stop using my own fork?

[tino]

Probably not. Look around, you'll see Zed abandoned this project. Please use https://github.com/moggers87/salmon/ as it is the best maintained fork