Test Parametrization
Parameterized testing is crucial for enhancing test coverage without writing numerous test functions for scenarios that only differ by their inputs and expected outputs. This approach not only saves time but also increases the clarity and maintainability of test suites by:
- Reducing code duplication and the potential for errors in test code;
- Enabling easy addition of new test cases by simply adding parameters;
- Facilitating comprehensive testing across a wide range of inputs.
Let's consider a simple function that calculates the average of two numbers. There can be a logical error if to forget to put braces. We'll use @pytest.mark.parametrize
to test this function with several pairs of numbers.
# Function to testdef calculate_average(num1, num2):return (num1 + num2) / 2# Test function using pytest's parametrizeimport pytest@pytest.mark.parametrize("num1, num2, expected",[(10, 20, 15),(20, 30, 25),(5, 5, 5)])def test_calculate_average(num1, num2, expected):assert (calculate_average(num1, num2) == expected), f"Sum of {num1} and {num2} should be equal to {expected}"
This example demonstrates how to run test_calculate_average with three different sets of numbers, checking if the function correctly computes their average.
Syntax with id
The id
parameter is used to assign an identifier to each set of inputs, which helps to identify tests more easily in the test output:
@pytest.mark.parametrize("num1, num2, expected", [pytest.param(10, 20, 15, id="integers"),pytest.param(20, 30, 25, id="more integers"),pytest.param(5, 5, 5, id="equal numbers")])def test_calculate_average_with_ids(num1, num2, expected):assert calculate_average(num1, num2) == expected
Using pytest.param
allows for more granular control over each test case, including setting individual IDs directly.
Syntax with ids
You can also provide a unique ID for each test case using ids
, which makes test reports more readable:
@pytest.mark.parametrize("num1, num2, expected", [(10, 20, 15),(20, 30, 25),(5, 5, 5)], ids=["integers", "more integers", "equal numbers"])def test_calculate_average_with_ids(num1, num2, expected):assert calculate_average(num1, num2) == expected
In this code, each set of parameters has an associated ID that describes the test case, which will appear in the test output.
Here's what the output might look like:
===================== test session starts ======================collected 3 itemstest_example.py::test_calculate_average_with_ids[integers] FAILEDtest_example.py::test_calculate_average_with_ids[more integers] PASSEDtest_example.py::test_calculate_average_with_ids[equal numbers] PASSED=========================== FAILURES ===========================________________ test_calculate_average_with_ids[integers] ________________num1 = 10, num2 = 20, expected = 15def test_calculate_average_with_ids(num1, num2, expected):> assert calculate_average(num1, num2) == expectedE assert 20 == 15E + where 20 = calculate_average(10, 20)test_example.py:10: AssertionError================== 1 failed, 2 passed in 0.23s ==================
Obrigado pelo seu feedback!