On 8/28/20 8:44 AM, Chris Green wrote: > Stefan Ram <r...@zedat.fu-berlin.de> wrote: >> Chris Green <c...@isbd.net> writes: >>> Therein lies the problem, the incoming byte stream *isn't* ASCII, it's >>> an E-Mail message which may, for example, have UTF-8 or other encoded >>> characters in it. Hopefully it will have an encoding given in the >>> header but that's only if the sender is 'well behaved', one needs to >>> be able to handle almost anything and it must be done without 'manual' >>> interaction. >> I would make a difference between "scoring" and "transport": >> >> To transfer a message into an mbox it can be transferred as it is. >> Just the bytes from the POP3 server. Let mutt deal with them. >> > That's what I do at present in Python 2, the problem is that Python 3 > complains when I use the standard library to put the message into the > mbox. > > I want to transport the message into my mbox and Python 3 won't do it > without knowing how it's encoded whereas Python 2 just stuffed it in > there 'as is'. > > I want Python 3's mailbox class to juyst put what I tell it (even if > mis-formatted or mis-encoded) into the mbox. > It looks like the mailbox class has gotten 'pickier' in Python 3, and won't accept a message as a byte string, just as either a email message or a real string. My guess would be that 'simplest' path would be to convert your message into a parsed Message class, and add that.
-- Richard Damon -- https://mail.python.org/mailman/listinfo/python-list
