Testing in Python with pytest

After this lesson, you will be able to: Write tests with pytest: fixtures, parametrize, mocking, and coverage.

Tests are how you ship without fear. pytest is the standard.

Prerequisites:Python Standard Library

pytest in 30 lines

Drop test_*.py files alongside your code.

python

# test_math.py
import pytest
from mymodule import add, divide

def test_add():
    assert add(2, 3) == 5

def test_divide_by_zero():
    with pytest.raises(ZeroDivisionError):
        divide(10, 0)

@pytest.mark.parametrize("a,b,expected", [
    (1, 1, 2),
    (0, 0, 0),
    (-1, 1, 0),
])
def test_add_cases(a, b, expected):
    assert add(a, b) == expected

# Run: pytest
# Run with coverage: pytest --cov=mymodule

Fixtures + mocking

Fixtures = setup helpers. Mocks = replace dependencies during tests.

python

import pytest
from unittest.mock import patch, MagicMock

@pytest.fixture
def sample_user():
    return {"name": "Alex", "email": "[email protected]"}

def test_uses_fixture(sample_user):
    assert sample_user["name"] == "Alex"

# Mock external API
@patch("mymodule.requests.get")
def test_fetch(mock_get):
    mock_get.return_value = MagicMock(
        status_code=200,
        json=lambda: {"ok": True},
    )
    from mymodule import fetch
    assert fetch("http://x")["ok"] is True
    mock_get.assert_called_once_with("http://x")

Coverage + CI

pip install pytest-cov. `pytest --cov=mypackage --cov-report=term-missing` shows uncovered lines. Aim for 60-80% on logic code (not getters/types). In CI: fail builds when coverage drops by more than X%. Use pytest + coverage.py + codecov.

Common mistakes only experienced Python devs avoid

Test files in src/ folder rather than tests/. Mixes prod + test code. Mocking everything until tests pass but real system is broken. Mock at the boundary only. Skipping pytest.raises. Catching with try/except in tests loses the pytest assertion. Hardcoded sleep() in tests. Use freezegun or pytest-asyncio properly.

Quick Check

What's the cleanest way to test a function with many input/output cases?

Pick the canonical pytest pattern.

Many functions named test_case_1, test_case_2@pytest.mark.parametrizeif statements inside one testWrite a loop

←Python Standard Library Deep Dive

Back to Python (Professional)

Python for Scripting and Automation→