webunit2 is a spiritual successor to Richard Jones’ webunit, but is by no means a fork. This library has been built from the ground up, complete with documentation and testing. It’s also hosted on GitHub, making it a cinch to fork.
I developed the need for webunit when I decided to standardize a lot of the testing for the company I work for. Since many of the applications managed by our team were backend services, having an easy way of wrapping web calls was a must, and that’s where I stumbled upon the original webunit.
However, I found that many things were missing–documentation was sparse, header manipulation wasn’t available in the library, and the repository housing the sources was nowhere to be found. Additionally, it only supported a subset of the HTML protocols, due to the httplib backend. This was fine for much of the frontend usage that webunit was designed for, but not so much for the backend services I was trying to test.
Thus, webunit2 was born one rainy Saturday. In its current form, it’s not quite up to par with webunit2, although it does support the features I immediately needed. Notably missing is cookie support, caching, web resource fetching (CSS, images, etc.), and DOM parsing.
To run unittests:
# Standard faire
$ nosetests
# With coverage reports
$ nosetests --with-coverage --cover-package=webunit2
To build this documentation:
$ easy_install sphinx
[...]
$ python setup.py build_sphinx
The built documentation is available (by default) in build/sphinx/html, and can be viewed locally on your web browser without the need for a separate web server.