I am pleased to announce a new Python project aimed to bridge Python to broader testing areas. The new project is called tappy and its goal is to generate test information in the Test Anything Protocol (TAP).

tappy converts Python unitest.TestCase based results into TAP results with either a TAPTestRunner that behaves just like the standard library TextTestRunner or with a nose plugin.

tappy popped into existence because I work in a very heterogeneous programming environment at work. The details of that work won’t excite the average reader, but the job consists of C++, Python, Perl, and plenty of shell scripts. All of these languages have various tests to report status and each testing environment is different. Unfortunately, the data from multiple languages cannot be entirely aggregated into one view.

The Test Anything Protocol can solve the problem of test aggregation by providing a lingua franca for tests. If each language generates results in the TAP format, the data can be normalized into that common view. The common view can give a sense of the health of the whole system. In such an environment, tappy acts as the TAP source for Python based test suites.

My long term goal for tappy is to make it available to as many versions of Python as possible and as many different Python test tools as possible (e.g., nose and trial).

You can follow tappy development on GitHub or get tappy from PyPI. The tappy name was taken on PyPI by the “Tidal Analysis Program” so I had to use tap.py instead.

$ pip install tap.py