non-recursive walk of email message from mailbox message

Question

I'm trying to work with email messages in Python 3.7 and struggling with what looks like compatibility issues. The docs mention email.message.Message having an iter_parts method that should allow me to do a non-recursive walk of message parts.

This doesn't exist on messages returned from mailbox messages and it's taken me a while to get it behaving. For example, I can generate a dummy message with:

from email.message import EmailMessage

msg = EmailMessage()
msg['Subject'] = 'msg 1'
msg.add_alternative("Plain text body", subtype='plain')
msg.add_alternative("<html><body><p>HTML body</p></body></html>", subtype='html')
msg.add_attachment(b"Nothing to see here!", maintype='data', subtype='raw')

and then dump out the parts with:

def iter_parts(msg):
  ret = msg.get_content_type()
  if msg.is_multipart():
    parts = ', '.join(iter_parts(m) for m in msg.iter_parts())
    ret = f'{ret} [{parts}]'
  return ret

iter_parts(msg)

which gives me: multipart/mixed [multipart/alternative [text/plain, text/plain], data/raw]

but if I save this to a mbox file and reload it:

import mailbox
mbox = mailbox.mbox('/tmp/test.eml')
mbox.add(msg)
iter_parts(mbox[0])

it tells me AttributeError: 'mboxMessage' object has no attribute 'iter_parts'

Initially I thought it might be related to https://stackoverflow.com/a/45804980/1358308 but setting factory=None doesn't seem to do much in Python 3.7.

Am posting my solution, but would like to know if there are better options!

score 1 · Answer 1 · answered Sep 24 '18 at 17:36

After much poking and reading of source I found that I can instead do:

from email import policy
from email.parser import BytesParser

mbox = mailbox.mbox('/tmp/test.eml', factory=BytesParser(policy=policy.default).parse)

and then I get objects with an iter_parts method.

non-recursive walk of email message from mailbox message

1 Answers1