Problem with parsing email message with extraneous MIME information

[Steven Allport]

I am working on processing eml email message using the email module (python
2.5), on files exported from an Outlook PST file, to extract the composite
parts of the email. In most instances this works fine, the message is read
in using message_from_file, is_multipart returns True and I can process each
component and extract message attachments.

I am however running into problem with email messages that contain emails
forwarded as attachments. The email has some additional encapulated header
information from each of the forwared emails.When I processes the files
is_multipart returns False the content-type is reported as text/plain
and the payload includes all the message body from 'This message is in MIME
format' though to the end.

for example.

<email header>
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0)
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_000_01C43634.1A06A235
------_=_NextPart_001_01C43634.1A06A235
------_=_NextPart_001_01C43634.1A06A235
------_=_NextPart_001_01C43634.1A06A235--
------_=_NextPart_000_01C43634.1A06A235
<attached message header>
------_=_NextPart_002_01C43634.1A06A235
------_=_NextPart_003_01C43634.1A06A235
------_=_NextPart_003_01C43634.1A06A235
------_=_NextPart_003_01C43634.1A06A235--
------_=_NextPart_002_01C43634.1A06A235
------_=_NextPart_002_01C43634.1A06A235--
------_=_NextPart_000_01C43634.1A06A235
Mime-Version: 1.0
Content-Type: multipart/mixed;
 boundary="------------m.182DA3C.BE6A21A3"
<rest of the message body>

If I remove the section of the email from the 'This is in MIME format'
through to Mime-Version: 1.0 the message is processed correctly. (ie.
is_multipart = True , Content-Type = multipart/mixed etc.)

Could anybody tell me if the above message header breaks the conventions for
email messages or is it just some that is not handled correctly by the email
module.

I would appreciate any feedback from anyone else who has experienced such
problems or could provide hints to a reliable solution.

Thanks,
Steve