cli-test-helpers 1.0.1

Useful helpers for writing tests for your Python CLI program.

cli-test-helpers

Useful helpers for writing tests for your Python CLI program.

Writing tests for a command line interface (CLI) tool may not seem strictly straight-forward when you think in terms of unit tests. Especially, when you use the argparse module or the click package, control of the application entry point is a bit taken away from you.

But it’s not all that bad. This package is here to help. The examples give you some guidance on how to get started, and the helpers allow you to deal with common cases, such as mocking CLI arguments and environment variable values.

Installation

pip install cli-test-helpers

Preferably, though, you add cli-test-helpers as a dependency to your Tox environment (see example).

Usage

Let’s assume you use pytest for running your tests, which is certainly a good idea. Your CLI program is called foobar. You have prepared a setup.py with a CLI entrypoint. For the tests you have prepared a tests/ folder (outside of foobar/, because you don’t want your tests to be packaged up with your application code). Then your directory layout looks somewhat like our example.

Functional tests

Start with a simple set of functional tests:

• Is the entrypoint script installed? (tests the configuration in your setup.py)
• Can this package be run as a Python module? (i.e. without having to be installed)
• Is command XYZ available? etc. Cover your entire CLI usage here!

This is almost a stupid exercise: Run the command as a shell command and inspect the status code of the exiting process (see example). The trick is that you run a non-destructive command, e.g. by using the usual --help option of every command. This should cover your entire CLI user interface definition.

Unit tests

Then you’re ready to take advantage of our helpers.

ArgvContext allows you to mimic the use of specific CLI arguments:

@patch('foobar.command.baz')
def test_cli_command(mock_command):
    """Is the correct code called when invoked via the CLI?"""
    with ArgvContext('foobar', 'baz'), pytest.raises(SystemExit):
        foobar.cli.main()

    assert mock_command.called

EnvironContext allows you to mimic the presence of environment values:

def test_fail_without_secret():
    """Must fail without a ``SECRET`` environment variable specified"""
    message_regex = "Environment value SECRET not set."

    with EnvironContext(SECRET=None):
        with pytest.raises(SystemExit, match=message_regex):
            foobar.command.baz()
            pytest.fail("CLI doesn't abort with missing SECRET")

See example.

TDD

Remember to stick to the test-driven mantra:

1. Write one line of test code. Make the test fail.
2. Write one line of application code. Make the test pass.
3. Goto 1.