Parse POP3 email message

mailparse.py

def mail_body_decode(message):
    for part in email.header.decode_header(message):
        binary, charset = part
        if isinstance(binary, str):
            yield str(binary)
        else:
            dammit = UnicodeDammit(binary, ['utf8', 'gb2312'] + [charset])
            decoded_str = dammit.unicode_markup
            yield decoded_str


def get_message_part(message_part, key):
    existed = list(filter(lambda x : x[0].lower() == key.lower(), message_part._headers))
    if not existed:
        return None
    return existed[0][1]


def parse_attachment(message_part):
    content_disposition = get_message_part(message_part, 'Content-Disposition')
    if content_disposition:
        content_disposition = ''.join(mail_body_decode(content_disposition))
        content_type = get_message_part(message_part, "Content-Type")
        content_type = ''.join(mail_body_decode(content_type))
        dispositions = content_disposition.strip().split(";")
        content_types = content_type.strip().split(";")
        if bool(content_disposition and dispositions[0].lower() == "attachment"):
            file_data = message_part.get_payload(decode=True)
            attachment = BytesIO(file_data)
            attachment.content_type = message_part.get_content_type()
            attachment.size = len(file_data)
            attachment.name = None
            attachment.create_date = None
            attachment.mod_date = None
            attachment.read_date = None

            for param in dispositions[1:] + content_types[1:]:
                name,value = param.split("=", 1)
                name = name.lower().strip()
                value = value.strip()
                if name == 'name':
                    attachment.name = value.strip('"')
                elif name == "filename":
                    attachment.name = value.strip('"')
                elif name == "create-date":
                    attachment.create_date = value  #TODO: datetime
                elif name == "modification-date":
                    attachment.mod_date = value #TODO: datetime
                elif name == "read-date":
                    attachment.read_date = value #TODO: datetime
            return attachment

    return None


def parse_mail(raw_email):
    dammit = UnicodeDammit(raw_email, ["utf8", "gb2312"])
    logger.debug(dammit.tried_encodings)
    raw_email_str = dammit.unicode_markup
    header_message = email.message_from_string(raw_email_str,
                                                policy=email.policy.EmailPolicy(utf8=True))
    message_id = header_message.get('Message-ID')
    message_date = header_message.get('Date')
    message_from = ''.join(mail_body_decode(header_message.get('From')))
    message_to = header_message.get('To')
    message_subject = header_message.get('Subject')
    message_date = datetime.fromtimestamp(mktime(parsedate(header_message['Date'])))
    subject = ''.join(mail_body_decode(message_subject))
    logger.debug(subject)

    mail = Mail()
    mail.mail_message_id = message_id
    mail.mail_date = message_date
    mail.mail_from = message_from
    mail.mail_to = message_to
    mail.subject = subject
    mail.mail_message_id_hash = hashlib.md5(message_id.encode('utf8')).hexdigest()
    return mail, header_message

def extract_mail_attachements(message):
    for part in message.walk():
        attachment = parse_attachment(part)
        if attachment:
            yield attachment.name, part.get_payload(decode=True)

Tried chardet, it detected wrong if mail is in "GB2312" encoding as "ISO-9958-1", but UnicodeDammit works good.

a very helpful reference
https://www.ianlewis.org/en/parsing-email-attachments-python