Mastering Python pytest: A Technical Guide to Effective Testing

Python pytest: Write Tests That Actually Help You

German Yamil introduces a practical framework for implementing pytest in Python development. The system utilizes auto-discovery of test_*.py files to accelerate the testing lifecycle.

Why This Matters

Manual testing is insufficient for modern software delivery; tests must be fast enough for developers to actually execute them. Without automated validation of edge cases and error conditions, bugs reach production where the cost of remediation is significantly higher than during the development phase.

Key Insights

• Fixture Isolation: Using @pytest.fixture ensures each test receives a fresh object (e.g., UserDB), preventing state leaks between test cases.
• Input Parametrization: The @pytest.mark.parametrize decorator allows a single test function to validate multiple input/output sets, reporting each case separately.
• Dependency Mocking: monkeypatch and pytest-mock allow engineers to replace real network calls (e.g., requests.get) with fake responses, enabling offline, high-speed test execution.

Working Examples

Testing exception handling using pytest.raises.

import pytest

def divide(a: float, b: float) -> float:
    if b == 0:
        raise ValueError("Cannot divide by zero")
    return a / b

def test_divide_by_zero():
    with pytest.raises(ValueError, match="Cannot divide by zero"):
        divide(10, 0)

def test_divide_normal():
    assert divide(10, 2) == 5.0

Using parametrize to run four distinct test cases through one function.

import pytest

@pytest.mark.parametrize("a, b, expected", [
    (2, 3, 5),
    (-1, 1, 0),
    (0, 0, 0),
    (100, -50, 50),
])
def test_add(a, b, expected):
    assert add(a, b) == expected

Practical Applications

References:

• https://dev.to/german_yamil_e021eef8710d/python-pytest-write-tests-that-actually-help-you-3546