Flanker - email address and MIME parsing for Python

Flanker is an open source parsing library written in Python by the Mailgun Team. Flanker currently consists of an address parsing library (flanker.addresslib) as well as a MIME parsing library (flanker.mime).

Detailed documentation is provided in the User Manual as well as the API Reference. A Quickstart Guide is provided below.

Installing

Flanker was built and tested with Python 2.7.2.

You can install flanker via pip or clone the repo from GitHub.

You'll need Python headers files before you start working with flanker, so install them first:

# ubuntu
sudo apt-get install python-dev
# fedora
sudo yum install python-devel

If you are using pip, simply type:

pip install flanker

If you are cloning from GitHub, you can type:

git clone [email protected]:mailgun/flanker.git
cd flanker
pip install -e .

Address Parsing

To parse a single mailbox (display name as well as email address):

>>> from flanker.addresslib import address
>>>
>>> address.parse('Foo [email protected]')
Foo <foo@example.com>

An invalid address is returned as None:

>>> from flanker.addresslib import address
>>>
>>> print address.parse('@example.com')
None

To parse a single email address (no display name):

>>> from flanker.addresslib import address
>>>
>>> address.parse('[email protected]', addr_spec_only=True)
foo@example.com

To parse an address list:

>>> from flanker.addresslib import address
>>>
>>> address.parse_list('[email protected], [email protected], @example.com')
[foo@example.com, bar@example.com]

To parse an address list as well as return a tuple containing the parsed addresses and the unparsable portions

>>> from flanker.addresslib import address
>>>
>>> address.parse_list('[email protected], [email protected], @example.com', as_tuple=True)
[foo@example.com, bar@example.com], ['@example.com']

To parse an address list in strict mode:

>>> from flanker.addresslib import address
>>>
>>> address.parse_list('[email protected], [email protected], @example.com', strict=True)
[foo@example.com, bar@example.com]

To validate an email address (parse as well as DNS, MX existence, and ESP grammar checks):

>>> from flanker.addresslib import address
>>>
>>> address.validate_address('[email protected]')
foo@mailgun.com

To validate an address list:

>>> from flanker.addresslib import address
>>>
>>> address.validate_list('[email protected], [email protected], @mailgun.com', as_tuple=True)
([foo@mailgun.com, bar@mailgun.com], ['@mailgun.com'])

MIME Parsing

For the following examples, message_string will be set to the following MIME message:

MIME-Version: 1.0
Content-Type: multipart/alternative; boundary=001a11c1d71697c7f004e6856996
From: Bob <[email protected]>
To: Alice <[email protected]>
Subject: hello, world
Date: Mon, 16 Sep 2013 12:43:03 -0700

--001a11c1d71697c7f004e6856996
Content-Type: text/plain; charset=us-ascii

Hello, *Alice*

--001a11c1d71697c7f004e6856996
Content-Type: text/html; charset=us-ascii

<p>Hello, <b>Alice</b></p>

--001a11c1d71697c7f004e6856996--

To parse a MIME message:

>>> from flanker import mime
>>>
>>> msg = mime.from_string(message_string)

MIME message headers (unicode multi-value dictionary with headers):

>>> from flanker import mime
>>>
>>> msg = mime.from_string(message_string)
>>> msg.headers.items()
[('Mime-Version', '1.0'),
 ('Content-Type',
  ('multipart/alternative', {'boundary': u'001a11c1d71697c7f004e6856996'})),
 ('From', 'Bob <[email protected]>'),
 ('To', 'Alice <[email protected]>'),
 ('Subject', 'hello, world'),
 ('Date', 'Mon, 16 Sep 2013 12:43:03 -0700')]

Useful content_type member with predicates:

>>> from flanker import mime
>>> msg = mime.from_string(message_string)
>>>
>>> msg.content_type.is_multipart()
True
>>>
>>> msg.content_type.is_singlepart()
False
>>>
>>> msg.content_type.is_message_container()
False

Decoded body of a message:

>>> from flanker import mime
>>> msg = mime.from_string(message_string)
>>>
>>> # None because message is multipart
>>> print msg.body
None
>>>
>>> for part in msg.parts:
       print 'Content-Type: {} Body: {}'.format(part, part.body)

Content-Type: (text/plain) Body: Hello, *Alice*
Content-Type: (text/html) Body: <p>Hello, <b>Alice</b></p>

>>> # None because no enclosed messages exist
>>> print msg.enclosed
None