demo-python-test-examples

A collection of examples for testing python code using pytest.

How to structure the test suite

demo-python-testing-examples/
├── examples/
│   ├── example_1.py
│   ├── ...
│   ├── example_N.py
│   └── further_examples/
│       ├── example_Nplus1.py
│       └── ...
├── tests/
│   ├── conftest.py
│   ├── test_example_1.py
│   ├── ...
│   ├── test_example_N.py
│   └── further_examples/
│       ├── test_example_Nplus1.py
│       └── ...
├── pyproject.toml
└── ...

Consider in the file example/further_examples/example_Nplus1.py there is a function named complex_function_15. All the tests concerning this function will be found in the file named tests/further_examples/test_example_Nplus1.py in a class named TestComplexFunction15.

By structuring your code and tests in this manner it is trivial for anyone looking at the code to be able to locate all the tests for any part of the code with ease.

1. Simple testing of a function

code tests

A basic example which demonstrates how a test suite for a function should be structured.

2. Simple testing of an expected error

code tests

This example demonstrates how to test a function which has a known raise statement within it. The example contains 2 functions:

1. raise_runtime_error simply raises a RuntimeError and the corresponding tests ensure that this is the raised exception
2. raise_value_error can raise two different ValueErrors depending on the value passed to the function. To ensure that the wrong excepction isn't being raised for each case, the tests make use of the match parameter in the pytest.raises context manager This ensures that not only the specified excpetion is raised, but also with the matching message.

(See also the pytest documentation on expected exceptions)

3. Using marks - TO DO

4. Using pytest.fixture to reuse data across tests - TO DO

5. Builtin parameterize v custom parametrize_cases to test multiple scenarios

This example shows how a single function can be tested with more than one set of inputs in a succinct DRY manner. The test code demonstrates using

1. the built in pytest parametrize functionality
2. the use of custom parametrize_cases and Case code which is kept in conftest.py

The benefit of using option 2 is that it provides more readable test cases compared to the built in functionality.

code tests

6. Combining fixtures and parameterised tests using pytest-lazy-fixture

This example demonstrates how to combine the reusability of fixtures with parameterised tests. In particular using built in functionality as well as an additional pytest add in helper to simplify the usage.

See this for an example of using the built in usage in more detail.

code tests

7. Usage of mock v pytest-mock for mocking function calls

code tests

8. Mocking a single function call within a function - TODO

9. Mocking multiple calls to same function within a function - TODO

To add

• Positive and negative
• skipped test
• xfail test
• passing xfail test
• test coverage
• pytest output