def write_mime_message_payloads(
output_dir: str,
transaction_id: str,
msg: MIMEMessage,
):
"""For each payload in the MIME message, write out the ones that are
attached as files.
Args:
msg (MIMEMessage): The parsed MIME message.
transaction_id (str): The transaction ID of the email.
output_dir (str): The directory to write the payloads in.
"""
for part in msg.walk():
if is_html(part) and not is_file(part):
write_html(output_dir, transaction_id, part)
# for some reason, outlook doesn't like EML files with HTML in them,
# so we blank out the HTML payload here to display it correctly
part.set_payload("")
elif is_file(part) and not part.is_multipart():
write_attachment(output_dir, transaction_id, part)
def anonymise_mime_message(msg: MIMEMessage) -> MIMEMessage:
"""Replace all of the sensitive fields in the MIME message with messages
indicating that this message has been anonymised.
The fields that get replaced are "Subject", "Thread-Topic", as well as the
body of the email. Attachments are preserved.
"""
msg.replace_header("Subject", "The subject has been removed for anonymity")
# the thread topic isn't always present
if msg.get("Thread-Topic") is not None:
msg.replace_header("Thread-Topic",
"The topic has been removed for anonymity")
for part in msg.walk():
# the only parts that contain the body are text/html and text/plain
# TODO(sam): I think sometimes these can be encoded as base64?
# Might need special handling...
if part.get_content_type() == "text/html" \
or part.get_content_type() == "text/plain":
if not is_file(part):
part.set_payload("The body has been removed for anonymity")
return msg
