Quick python code to parse mbox files, specifically those used by GMail. Extracts sender, date, plain text contents etc., ignores base64 attachments.

gmail_mbox_parser.py

#! /usr/bin/env python3
# ~*~ utf-8 ~*~

import mailbox
import bs4

def get_html_text(html):
    try:
        return bs4.BeautifulSoup(html, 'lxml').body.get_text(' ', strip=True)
    except AttributeError: # message contents empty
        return None

class GmailMboxMessage():
    def __init__(self, email_data):
        if not isinstance(email_data, mailbox.mboxMessage):
            raise TypeError('Variable must be type mailbox.mboxMessage')
        self.email_data = email_data

    def parse_email(self):
        email_labels = self.email_data['X-Gmail-Labels']
        email_date = self.email_data['Date']
        email_from = self.email_data['From']
        email_to = self.email_data['To']
        email_subject = self.email_data['Subject']
        email_text = self.read_email_payload() 

    def read_email_payload(self):
        email_payload = self.email_data.get_payload()
        if self.email_data.is_multipart():
            email_messages = list(self._get_email_messages(email_payload))
        else:
            email_messages = [email_payload]
        return [self._read_email_text(msg) for msg in email_messages]

    def _get_email_messages(self, email_payload):
        for msg in email_payload:
            if isinstance(msg, (list,tuple)):
                for submsg in self._get_email_messages(msg):
                    yield submsg
            elif msg.is_multipart():
                for submsg in self._get_email_messages(msg.get_payload()):
                    yield submsg
            else:
                yield msg

    def _read_email_text(self, msg):
        content_type = 'NA' if isinstance(msg, str) else msg.get_content_type()
        encoding = 'NA' if isinstance(msg, str) else msg.get('Content-Transfer-Encoding', 'NA')
        if 'text/plain' in content_type and 'base64' not in encoding:
            msg_text = msg.get_payload()
        elif 'text/html' in content_type and 'base64' not in encoding:
            msg_text = get_html_text(msg.get_payload())
        elif content_type == 'NA':
            msg_text = get_html_text(msg)
        else:
            msg_text = None
        return (content_type, encoding, msg_text)

######################### End of library, example of use below

mbox_obj = mailbox.mbox('path/to/your-mbox-file-from-gmail.mbox')

num_entries = len(mbox_obj)

for idx, email_obj in enumerate(mbox_obj):
    email_data = GmailMboxMessage(email_obj)
    email_data.parse_email()
    print('Parsing email {0} of {1}'.format(idx, num_entries))

email_text does not contain pure email text; there is noisy data (like css tags). Is there a method to get the pure text?

"Is there a way" yes; but it depends on what exactly you mean. A common solution to extracting just the text from HTML payloads is to run Beautifulsoup on the HTML. If you want to trim off quoted text from earlier messages in a thread, I don't know of any existing libraries for that (but that doesn't mean there aren't any). Similarly, you might want to trim signature blocks (honest-to-RFC signatures start with newline, dash, dash, space, newline; but very few modern signatures adhere to this convention).

See also:

• https://stackoverflow.com/questions/753052/strip-html-from-strings-in-python
• https://stackoverflow.com/questions/5420402/reliable-way-to-only-get-the-email-text-excluding-previous-emails (somewhat limited answers, I'm afraid)

For anyone who wants useful output out of the HTML: you'll want

msg.get_payload(decode=True).decode()

This would have saved me a lot of heartburn.