Not-so-good example of mail header / body parsing

mail-parse.py

import os
import re
import sys
from email.parser import Parser
from email.header import decode_header

import logging

logger = logging.getLogger(__name__)


def setup_logger(logger, level, stdout=False):
    if stdout:
        handler = logging.StreamHandler()
        handler.setFormatter(logging.Formatter('%(asctime)-15s %(levelname)s: '
                                               '%(message)s'))
    else:
        import logging.handlers
        handler = logging.handlers.SysLogHandler(address='/dev/log')
        handler.setFormatter(logging.Formatter('%(filename)s %(levelname)s: '
                                               '%(message)s'))
    logger.addHandler(handler)
    logger.setLevel(level)


setup_logger(logger, logging.DEBUG)

"""
Le    :   ,   @    :
On    :   ,   @    :
2013-10-31 11:22 GMT+01:00     @     :

"""
quotation_regexs = (r'\n(Le|On)[^:]+:[^,]+,[^@]+@[^:]+:\n\n>',
                    r'\n[0-9-]+ [0-9:A-Z+-]+[^@]+@[^:]+:\n\n>')


def get_message_in_body(body):
    if body is None:
        return ''

    # Remove carriage return:
    body = body.replace('\r', '')

    # Remove quotations header in body:
    for quotation_regex in quotation_regexs:
        body = re.sub(quotation_regex, '\n>', body, flags=re.M)

    # Remove every quotations:
    body = re.sub('^>.*$', '', body, flags=re.M)

    # Remove signature:
    body = body.split('\n-- \n')[0]

    # Deduplicate new lines:
    body = re.sub('\n+', '\n', body)

    return body


def try_decode_header(header):
    """
    Try to decode a header, from:
    =?UTF-8?Q?Conseils_d=27utilisation_de_votre_nouvelle_bo=C3=AEte_de_r?=
    to:
    Conseils d'utilisation de votre nouvelle boîte de réception
    """
    if header is None:
        return ''
    try:
        header, encoding = decode_header(header)[0]
    except Exception as ex:
        print(ex, file=sys.stderr)
    if encoding is None:
        encoding = 'utf-8'
    if type(header) is bytes:
        return header.decode(encoding)
    return header


def get_attachments(mail):
    if mail.is_multipart():
        for payload in mail.get_payload():
            if 'multipart/alternative' in payload.get('Content-Type', ''):
                for attachement in get_attachments(payload):
                    yield attachement
            if 'image/' in payload.get('Content-Type', ''):
                params = dict(payload.get_params())
                mime = payload.get_params()[0][0]
                yield (mime, try_decode_header(params.get('name')),
                       payload.get_payload(decode=True))


def get_text_body(mail):
    if mail.is_multipart():
        for payload in mail.get_payload():
            if 'multipart/alternative' in payload.get('Content-Type', ''):
                return get_text_body(payload)
            if 'text/plain' in payload.get('Content-Type', ''):
                decoded_payload = payload.get_payload(decode=True)
                try:
                    charset = 'UTF-8'
                    if 'charset=' in payload['Content-Type']:
                        charset = payload['Content-Type'].split('charset=')[1]
                    return decoded_payload.decode(charset)
                except UnicodeDecodeError:
                    print("Failed to decode a mail with Content-Type:" +
                          payload['Content-Type'])
                    raise
    else:
        try:
            return mail.get_payload(decode=True).decode('utf-8')
        except UnicodeDecodeError:
            return mail.get_payload(decode=True).decode('latin-1')


ticket = Parser().parse(sys.stdin)
subject = try_decode_header(ticket['Subject']).strip()
sender = try_decode_header(ticket['From'])
body = get_message_in_body(get_text_body(ticket))