Nitpycker

NitPycker is a improved runner for your tests in Python. It allows your tests to run in parallel in a way totally transparent for you.

For a quick jump start, please see Quickstart, for more information please read Why NitPycker ?.

However, be warned that if you have tests that depend on each others, or tests fixtures that prevent the use of multiprocessing, some tests will fail. This is not a bug in NitPycker. You can still use nitpycker if you have only a few problems. For this, please see Avoiding problems.

Why NitPycker ?

I created NitPycker because running Python unittest in a multithreaded environment was impossible for me. I had a heavy test suite (>90 tests) with integration tests, and running them sequentially would take more than 8 hours. With NitPycker, time got down to 40 minutes.

Other Python testing framework do also provide multithreaded tests, such as py.test, nose and nose2. However, they all have some problems when running in multithreaded environment (code coverage incorrect, path manipulation, deadlocks) meaning that they are not perfect for it. NitPycker is created with multiprocess in mind, meaning a better integration and more reliable results. If tests are independent and work with Python’s unittest, Nitpycker will be fully compatible with them [1].

Quickstart

Getting started is trivial :

  1. Download NitPycker from the NitPycker page on the Python Package Index or by using pip install nitpycker.

  2. Use nitpycker to run your tests

    $ nitpycker [options] [args] directory

Getting help

Bug reports are welcome on the GitHub’s project’s issue tracker

More information

[1]if it’s not, please open an issue, this should get fixed !