https://github.com/scrapy/loginform

Fill HTML login forms automatically
https://github.com/scrapy/loginform

Last synced: 8 months ago
JSON representation

Fill HTML login forms automatically

• Host: GitHub
• URL: https://github.com/scrapy/loginform
• Owner: scrapy
• Created: 2012-10-23T13:57:53.000Z (about 13 years ago)
• Default Branch: master
• Last Pushed: 2024-04-24T08:56:31.000Z (over 1 year ago)
• Last Synced: 2025-05-03T13:44:01.728Z (8 months ago)
• Language: Python
• Size: 90.8 KB
• Stars: 274
• Watchers: 25
• Forks: 80
• Open Issues: 14
• Metadata Files:
  • Readme: README.rst

Awesome Lists containing this project

README

          
loginform

=========

.. image:: https://secure.travis-ci.org/scrapy/loginform.png?branch=master

   :target: http://travis-ci.org/scrapy/loginform

.. image:: https://img.shields.io/codecov/c/github/scrapy/loginform/master.svg

   :target: http://codecov.io/github/scrapy/loginform?branch=master

   :alt: Coverage report

loginform is a library for filling HTML login forms given the login url,

username & password. Which form and fields to fill are inferred automatically.

It's implemented on top of `lxml form filling`_, and thus depends on lxml.

Usage

-----

Usage is very simple and best illustrated with an example::

    >>> from loginform import fill_login_form

    >>> import requests

    >>> url = "https://github.com/login"

    >>> r = requests.get(url)

    >>> fill_login_form(url, r.text, "john", "secret")

    ([('authenticity_token', 'FQgPiKd1waDL+pycPH8IGutirTnP69SiZgm0zXwn+VQ='),

      ('login', 'john'),

      ('password', 'secret')],

     u'https://github.com/session',

     'POST')

And it is possible to use it as a tool to quickly debug a login form::

    $ python -m loginform -u john -p secret https://github.com/login

    url: https://github.com/session

    method: POST

    payload:

    - authenticity_token: FQgPiKd1waDL+pycPH8IGutirTnP69SiZgm0zXwn+VQ=

    - login: john

    - password: secret

Testing

-------

A collection of real-world samples is used to keep this library tested. Those

samples are managed as follows:

First, you select a site to try, find out its login url, and run the following

command to try loginform on it::

    $ python test_samples.py https://github.com/login

    [

       "https://github.com/login", 

       [

          [

             [

                "authenticity_token", 

                "NsdVWGpzxKmn7zSJSOdgnDcLIzIdJlCTO754LiEv2W4="

             ], 

             [

                "login", 

                "USER"

             ], 

             [

                "password", 

                "PASS"

             ]

          ], 

          "https://github.com/session", 

          "POST"

       ]

    ]

From the output you can judge if it worked or not. If it worked, great. If it

didn't, you would hack ``loginform.py`` to make it work and then add the sample

with::

    $ python test_samples.py https://github.com/login -w github

Note that we gave the sample a name (``github`` in this case).

To list all available samples use::

    $ python test_samples.py -l

To run all tests, install tox and run::

    $ tox

.. _lxml form filling: http://lxml.de/lxmlhtml.html#forms

.. _tox: https://pypi.python.org/pypi/tox