alfonsrv/mail-parser-reply

📧 Mail reply parser library for Python with multi-language support

Mail Reply Parser 📧🐍

Multi-language email reply parsing for international environments 🌍

Mail clients handle reply formatting differently, making reliable parsing difficult. Thank god we have standards. This library splits text-based emails into separate replies based on common headers produced by different, multilingual clients usually indicating separation.

Replies can either present the whole mail message body, or strip headers, signatures and common disclaimers if required. Currently supported languages are: English (en), German (de), French (fr), Italian (it), Japanese (ja), Polish (pl) – adding more languages is quite easy.

This is an improved Python implementation of GitHub's Ruby-based email_reply_parser and an adaptation of Zapier's email-reply-parser which both split the mails in fragments instead of distinct replies. They also only support English.

⭐ Features

⭐ Easy to implement
⭐ Multilanguage Support
⭐ Text-based mail parsing
⭐ Detect headers, signatures and disclaimers
⭐ Fully type annotated
⭐ Easy-to-read code and well-tested

Overview 🔭

This library makes it easy to split an incoming mail into replies, making working with emails much more manageable and easily providing the text content for each reply – with or without signatures, disclaimers and headers.

For example, it can turn the following email:

Awesome! I haven't had another problem with it.

Thanks,
alfonsrv

On Wed, Dec 20, 2023 at 13:37, RAUSYS <[email protected]> wrote:

> The good news is that I've found a much better query for lastLocation.
> It should run much faster now. Can you double-check?

Into just the replied text content:

Awesome! I haven't had another problem with it.

Get started 👾

Installation

pip install mail-parser-reply

Parse Replies

from mailparser_reply import EmailReplyParser

mail_body = 'foobar'; languages = ['en', 'de']
mail_message = EmailReplyParser(languages=languages).read(text=mail_body)
print(mail_message.replies)

Or get only the latest reply using:

latest_reply = EmailReplyParser(languages=languages).parse_reply(text=mail_body)

Parser API

EmailMessage.text:              Mail body
EmailMessage.languages:         Languages to use for parsing headers
EmailMessage.replies:           List of EmailReply; single parsed replies
EmailMessage.include_english:   Always include English language for parsing
EmailMessage.default_language:  Default language to use if language dictionary 
                                doesn't include any other language codes

EmailMessage.HEADER_REGEX:      RegEx for identifying headers, separating mails
EmailMessage.SIGNATURE_REGEX:   RegEx for identifying signatures
EmailMessage.DISCLAIMERS_REGEX: RegEx for identifying disclaimers

EmailMessage.read(): Parse EmailMessage.text to EmailReply which are then stored 
                     in EmailMessage.replies

EmailReply.content:     Unprocessed mail body with headers, signatures, disclaimers
EmailReply.body:        Mail body without headers, signatures, disclaimers
EmailReply.full_body:   Mail body; just without headers

EmailReply.headers:     Identified Headers
EmailReply.signatures:  Identified Signatures
EmailReply.disclaimers: Identified disclaimers

---

mail-parser-reply's Issues

Unable to use the parser

I have a function where I am calling your library. By simple copying your example (even using the "foobar" as data) I am getting the following error:
mail_message = EmailReplyParser.read(text=mail_body, languages=languages) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ TypeError: EmailReplyParser.read() got an unexpected keyword argument 'languages'

Function
`def beautify_body(body):

# Remove HTML code
soup = BeautifulSoup(body, 'html.parser')
body = soup.get_text()

mail_body = 'foobar'; languages = ['en', 'de']
mail_message = EmailReplyParser.read(text=mail_body, languages=languages)
print(mail_message.replies)


return body`

Question - why some test cases are commented?

Hi,

First of all, thank you for the great work you did on this package.

I'm coming from the node world, and after spending some time on looking for 'the-best-solution' (free & open-source) to parse email, with their contents (replies), i found this one, and I think this offers great extensibility.

I was looking at the code, pretty-clean ! Wondering why some test cases are commented tho?

Thanks in advance.