neatest 0.0.6

Runs standard unittest discovery and testing, requiring less rain dance.

neatest

Runs standard Python unittest discovery and testing. Provides a more convenient way of configuring the tests.

It replaces the shell command python -m unittest discover ... with a brief programmatic call from Python code.

Why

Testing should be simple. One line command. A really short line.

python -m unittest discover ... is too long.

run_tests.sh is better. But not pythonic and not cross-platform.

run_tests.py is much better.

For example, if your command looks like this

$ cd project_dir && python -m unittest discover -s ./mymodule -p '*_test.py' --buffer

you can replace it with run_tests.py:

import neatest
neatest.pattern = "*_test.py"
neatest.start_dir = "./module"
neatest.run()

This script can be run with python3 run_tests.py. Now the command is short and cross-platform.

Install

pip3 install neatest

Run

Run tests with .py script

project_dir / run_tests.py

import neatest

# optionally setting parameters
neatest.pattern = '*_test.py'
neatest.verbocity = 2

if __name__ == "__main__":
    neatest.run()

Terminal

$ python3 run_tests.py

Run tests from terminal

$ neatest

is equivalent to running the script

import neatest
neatest.run()

Run tests with setup.py

project_dir / run_tests.py

import neatest

# optionally setting parameters. 
# Setup.py will only respect parameters related to
# the test discovery (TestSuite initialization)
neatest.pattern = '*_test.py'  # will be used
neatest.verbocity = 2          # will be ignored 

project_dir / setup.py

setup(
  ...
  test_suite='run_tests.neatest.suite',
)

Terminal

$ cd project_dir
$ python3 setup.py test

Differences from "-m unittest discover"

pattern

neatest searches for tests in all *.py files. Any TestCase in the code is considered a test to be run.

Standard unittest discover searches only for tests in files named test*.py.

start_dir and top_level_dir

If not specified, neatest find the first directory containing __init__.py and consider it the start_dir (the module containing the tests).

The top_level_dir by default is the current directory, and it does not depend on start_dir.

It works well with package structure like that:

project_dir
  my_module
    tests
      __init__.py
      test_a.py
      test_b.py 
    __init__.py
    code1.py
    code2.py
  setup.py

If we run testing with project_dir being the current directory, the tests will be found inside my_module. The project_dir will remain the current directory for imports.

---

Standard unittest discover assumes that the starting directory is the current directory. If run from the project_dir it will not find any tests, unless you explicitly specify that the start_dir is my_module. But when you specify start_dir, the top_level_dir also changes to the same value. If you want to keep top_level_dir as the project_dir, you will have to specify this explicitly as well.

For neatest to behave as standard, just set neatest.start_dir="." and neatest.top_level_dir=None