python-testing

name	python-testing
description	Python testing strategies using pytest, TDD methodology, fixtures, mocking, and parametrization. Core testing fundamentals.

Python Testing Patterns

Comprehensive testing strategies for Python applications using pytest, TDD methodology, and best practices.

When to Activate

• Writing new Python code (follow TDD: red, green, refactor)
• Designing test suites for Python projects
• Reviewing Python test coverage
• Setting up testing infrastructure
• Choosing the correct scope (function, module, session) for a pytest fixture
• Debugging flaky tests caused by shared mutable state or incorrect mock namespaces
• Parametrizing tests to cover multiple inputs without duplicating test logic
• Enforcing 80%+ coverage with pytest --cov and understanding which lines are uncovered

Core Testing Philosophy

Test-Driven Development (TDD)

Always follow the TDD cycle:

1. RED: Write a failing test for the desired behavior
2. GREEN: Write minimal code to make the test pass
3. REFACTOR: Improve code while keeping tests green

# Step 1: Write failing test (RED)
def test_add_numbers():
    result = add(2, 3)
    assert result == 5

# Step 2: Write minimal implementation (GREEN)
def add(a, b):
    return a + b

# Step 3: Refactor if needed (REFACTOR)

Coverage Requirements

• Target: 80%+ code coverage
• Critical paths: 100% coverage required
• Use pytest --cov to measure coverage

pytest --cov=mypackage --cov-report=term-missing --cov-report=html

pytest Fundamentals

Basic Test Structure

import pytest

def test_addition():
    """Test basic addition."""
    assert 2 + 2 == 4

def test_string_uppercase():
    """Test string uppercasing."""
    text = "hello"
    assert text.upper() == "HELLO"

def test_list_append():
    """Test list append."""
    items = [1, 2, 3]
    items.append(4)
    assert 4 in items
    assert len(items) == 4

Assertions

# Equality
assert result == expected

# Inequality
assert result != unexpected

# Truthiness
assert result  # Truthy
assert not result  # Falsy
assert result is True  # Exactly True
assert result is False  # Exactly False
assert result is None  # Exactly None

# Membership
assert item in collection
assert item not in collection

# Comparisons
assert result > 0
assert 0 <= result <= 100

# Type checking
assert isinstance(result, str)

# Exception testing (preferred approach)
with pytest.raises(ValueError):
    raise ValueError("error message")

# Check exception message
with pytest.raises(ValueError, match="invalid input"):
    raise ValueError("invalid input provided")

# Check exception attributes
with pytest.raises(ValueError) as exc_info:
     ValueError()
 (exc_info.value) == 

Fixtures

Basic Fixture Usage

import pytest

@pytest.fixture
def sample_data():
    """Fixture providing sample data."""
    return {"name": "Alice", "age": 30}

def test_sample_data(sample_data):
    """Test using the fixture."""
    assert sample_data["name"] == "Alice"
    assert sample_data["age"] == 30

Fixture with Setup/Teardown

@pytest.fixture
def database():
    """Fixture with setup and teardown."""
    # Setup
    db = Database(":memory:")
    db.create_tables()
    db.insert_test_data()

    yield db  # Provide to test

    # Teardown
    db.close()

def test_database_query(database):
    """Test database operations."""
    result = database.query("SELECT * FROM users")
    assert len(result) > 0

Fixture Scopes

# Function scope (default) - runs for each test
@pytest.fixture
def temp_file():
    with open("temp.txt", "w") as f:
        yield f
    os.remove("temp.txt")

# Module scope - runs once per module
@pytest.fixture(scope="module")
def module_db():
    db = Database(":memory:")
    db.create_tables()
    yield db
    db.close()

# Session scope - runs once per test session
@pytest.fixture(scope="session")
def shared_resource():
    resource = ExpensiveResource()
    yield resource
    resource.cleanup()

Fixture with Parameters

@pytest.fixture(params=[1, 2, 3])
def number(request):
    """Parameterized fixture."""
    return request.param

def test_numbers(number):
    """Test runs 3 times, once for each parameter."""
    assert number > 0

Using Multiple Fixtures

@pytest.fixture
def user():
    return User(id=1, name="Alice")

@pytest.fixture
def admin():
    return User(id=2, name="Admin", role="admin")

def test_user_admin_interaction(user, admin):
    """Test using multiple fixtures."""
    assert admin.can_manage(user)

Autouse Fixtures

@pytest.fixture(autouse=True)
def reset_config():
    """Automatically runs before every test."""
    Config.reset()
    yield
    Config.cleanup()

def test_without_fixture_call():
    # reset_config runs automatically
    assert Config.get_setting("debug") is False

Conftest.py for Shared Fixtures

# tests/conftest.py
import pytest

@pytest.fixture
def client():
    """Shared fixture for all tests."""
    app = create_app(testing=True)
    with app.test_client() as client:
        yield client

@pytest.fixture
def auth_headers(client):
    """Generate auth headers for API testing."""
    response = client.post("/api/login", json={
        "username": "test",
        "password": "test"
    })
    token = response.json["token"]
    return {"Authorization": f"Bearer {token}"}

Parametrization

Basic Parametrization

@pytest.mark.parametrize("input,expected", [
    ("hello", "HELLO"),
    ("world", "WORLD"),
    ("PyThOn", "PYTHON"),
])
def test_uppercase(input, expected):
    """Test runs 3 times with different inputs."""
    assert input.upper() == expected

Multiple Parameters

@pytest.mark.parametrize("a,b,expected", [
    (2, 3, 5),
    (0, 0, 0),
    (-1, 1, 0),
    (100, 200, 300),
])
def test_add(a, b, expected):
    """Test addition with multiple inputs."""
    assert add(a, b) == expected

Parametrize with IDs

@pytest.mark.parametrize("input,expected", [
    ("[e-mail]", True),
    ("invalid", False),
    ("@no-domain.com", False),
], ids=["valid-email", "missing-at", "missing-domain"])
def test_email_validation(input, expected):
    """Test email validation with readable test IDs."""
    assert is_valid_email(input) is expected

Parametrized Fixtures

@pytest.fixture(params=["sqlite", "postgresql", "mysql"])
def db(request):
    """Test against multiple database backends."""
    if request.param == "sqlite":
        return Database(":memory:")
    elif request.param == "postgresql":
        return Database("postgresql://localhost/test")
    elif request.param == "mysql":
        return Database("mysql://localhost/test")

def test_database_operations(db):
    """Test runs 3 times, once for each database."""
    result = db.query("SELECT 1")
    assert result is not None

Markers and Test Selection

Custom Markers

# Mark slow tests
@pytest.mark.slow
def test_slow_operation():
    time.sleep(5)

# Mark integration tests
@pytest.mark.integration
def test_api_integration():
    response = requests.get("https://api.example.com")
    assert response.status_code == 200

# Mark unit tests
@pytest.mark.unit
def test_unit_logic():
    assert calculate(2, 3) == 5

Run Specific Tests

# Run only fast tests
pytest -m "not slow"

# Run only integration tests
pytest -m integration

# Run integration or slow tests
pytest -m "integration or slow"

# Run tests marked as unit but not slow
pytest -m "unit and not slow"

Configure Markers in pytest.ini

[pytest]
markers =
    slow: marks tests as slow
    integration: marks tests as integration tests
    unit: marks tests as unit tests
    django: marks tests as requiring Django

Mocking and Patching

Mocking Functions

from unittest.mock import patch, Mock

@patch("mypackage.external_api_call")
def test_with_mock(api_call_mock):
    """Test with mocked external API."""
    api_call_mock.return_value = {"status": "success"}

    result = my_function()

    api_call_mock.assert_called_once()
    assert result["status"] == "success"

Mocking Return Values

@patch("mypackage.Database.connect")
def test_database_connection(connect_mock):
    """Test with mocked database connection."""
    connect_mock.return_value = MockConnection()

    db = Database()
    db.connect()

    connect_mock.assert_called_once_with("localhost")

Mocking Exceptions

@patch("mypackage.api_call")
def test_api_error_handling(api_call_mock):
    """Test error handling with mocked exception."""
    api_call_mock.side_effect = ConnectionError("Network error")

    with pytest.raises(ConnectionError):
        api_call()

    api_call_mock.assert_called_once()

Mocking Context Managers

@patch("builtins.open", new_callable=mock_open)
def test_file_reading(mock_file):
    """Test file reading with mocked open."""
    mock_file.return_value.read.return_value = "file content"

    result = read_file("test.txt")

    mock_file.assert_called_once_with("test.txt", "r")
    assert result == "file content"

Using Autospec

@patch("mypackage.DBConnection", autospec=True)
def test_autospec(db_mock):
    """Test with autospec to catch API misuse."""
    db = db_mock.return_value
    db.query("SELECT * FROM users")

    # This would fail if DBConnection doesn't have query method
    db_mock.assert_called_once()

Mock Class Instances

class TestUserService:
    @patch("mypackage.UserRepository")
    def test_create_user(self, repo_mock):
        """Test user creation with mocked repository."""
        repo_mock.return_value.save.return_value = User(id=1, name="Alice")

        service = UserService(repo_mock.return_value)
        user = service.create_user(name="Alice")

        assert user.name == "Alice"
        repo_mock.return_value.save.assert_called_once()

Mock Property

@pytest.fixture
def mock_config():
    """Create a mock with a property."""
    config = Mock()
    type(config).debug = PropertyMock(return_value=True)
    type(config).api_key = PropertyMock(return_value="test-key")
    return config

def test_with_mock_config(mock_config):
    """Test with mocked config properties."""
    assert mock_config.debug is True
    assert mock_config.api_key == "test-key"

Anti-Patterns

Patching the Wrong Namespace

Wrong:

# Patching where the function is defined, not where it is used
@patch("mypackage.utils.requests.get")
def test_fetch(mock_get):
    mock_get.return_value.json.return_value = {"ok": True}
    result = mypackage.service.fetch_data()  # imports requests.get directly
    mock_get.assert_called_once()  # never called — wrong target

Correct:

# Patch the name as it is imported in the module under test
@patch("mypackage.service.requests.get")
def test_fetch(mock_get):
    mock_get.return_value.json.return_value = {"ok": True}
    result = mypackage.service.fetch_data()
    mock_get.assert_called_once()

Why: @patch replaces the name in the namespace where it is looked up, not where it is defined.

---

Sharing Mutable State Between Tests via Module-Level Variables

Wrong:

SHARED_LIST = []  # module-level mutable state

def test_append_item():
    SHARED_LIST.append("a")
    assert "a" in SHARED_LIST

def test_list_is_empty():
    assert SHARED_LIST == []  # fails if test_append_item ran first

Correct:

@pytest.fixture
def items():
    return []

def test_append_item(items):
    items.append("a")
    assert "a" in items

def test_list_is_empty(items):
    assert items == []  # always a fresh list

Why: Module-level mutable state leaks between tests, making order-dependent failures that are hard to diagnose.

---

Asserting on Exceptions Without Checking the Message

Wrong:

def test_invalid_email_raises():
    with pytest.raises(ValueError):
        create_user(email="not-an-email")
    # passes even if ValueError is raised for a completely different reason

Correct:

def test_invalid_email_raises():
    with pytest.raises(ValueError, match="invalid email"):
        create_user(email="not-an-email")

Why: Without match=, any ValueError — including ones from unrelated bugs — makes the test pass silently.

---

Using assert Inside a Loop Without Identifying the Failing Case

Wrong:

def test_all_users_active():
    for user in get_users():
        assert user.is_active  # failure output: "assert False" — no context

Correct:

@pytest.mark.parametrize("user", get_users())
def test_user_is_active(user):
    assert user.is_active, f"Expected user {user.id!r} to be active"

Why: Parametrized tests identify exactly which input failed; a loop gives no context and stops at the first failure.

---

Using time.sleep in Tests to Wait for Async Behaviour

Wrong:

def test_background_job_runs():
    trigger_job()
    time.sleep(2)  # hope it finishes in time — flaky on slow CI
    assert job_completed()

Correct:

def test_background_job_runs():
    trigger_job()
    # poll with a timeout, or use pytest-asyncio / mock the scheduler
    deadline = time.monotonic() + 5
    while not job_completed():
        assert time.monotonic() < deadline, "Job did not complete within 5s"
        time.sleep(0.1)

Why: Fixed sleeps cause flaky tests on slow machines and waste time on fast ones; polling with a deadline is both reliable and fast.

For advanced testing — async code (pytest-asyncio), exception attribute testing, file/side-effect testing, test organization (directory structure, test classes), best practices, API endpoint patterns, database testing, pytest configuration, and CLI reference — see skill: python-testing-advanced.