pytest-django-patterns（pytest Django 模式）

pytest-django Testing Patterns

TDD Workflow (RED-GREEN-REFACTOR)

Always follow this cycle:

1. RED: Write a failing test first that describes desired behavior
2. GREEN: Write minimal code to make the test pass
3. REFACTOR: Clean up code while keeping tests green
4. REPEAT: Never write production code without a failing test

Critical rule: If implementing a feature or fixing a bug, write the test BEFORE touching production code.

Essential pytest-django Patterns

Database Access

• Use @pytest.mark.django_db on any test touching the database
• Apply to entire module: pytestmark = pytest.mark.django_db
• Transactions roll back automatically after each test

Fixtures for Test Data

Use Factory Boy for models, pytest fixtures for setup:

• Factories: Create model instances with realistic data (UserFactory())

- Use factory.Sequence() for unique fields - Use factory.Faker() for realistic fake data - Use factory.SubFactory() for foreign keys - Use @factory.post_generation for M2M relationships

• Fixtures: Setup clients, auth state, or shared resources

- client fixture: Django test client - Create auth_client fixture: client.force_login(user) for authenticated requests - Define in conftest.py for reuse across test files

Test Organization

Structure tests to mirror app structure:

tests/
├── apps/
│   └── posts/
│       ├── test_models.py
│       ├── test_views.py
│       └── test_forms.py
├── factories.py
└── conftest.py

Group related tests in classes:

• Name classes TestComponentName (e.g., TestPostListView)
• Name test methods descriptively: test_<action>_<expected_outcome>
• Use @pytest.mark.parametrize for testing multiple scenarios

What to Test

Views

• Status codes: Correct HTTP responses (200, 404, 302)
• Authentication: Authenticated vs anonymous behavior
• Authorization: User can only access their own data
• Context data: Correct objects passed to template
• Side effects: Database changes, emails sent, tasks queued
• HTMX: Check HTTP_HX_REQUEST header returns partial template

Forms

• Validation: Valid data passes, invalid data fails with correct errors
• Edge cases: Empty fields, max lengths, unique constraints
• Clean methods: Custom validation logic works
• Save behavior: Objects created/updated correctly

Models

• Methods: __str__, custom methods return expected values
• Managers/QuerySets: Custom filtering works correctly
• Constraints: Database-level validation enforced
• Signals: Pre/post save hooks execute correctly

Celery Tasks

• Mock external calls: Patch HTTP requests, email sending, etc.
• Test logic only: Don't test actual async execution
• Idempotency: Running task multiple times is safe

Django-Specific Testing Patterns

Testing HTMX Responses

Check partial template rendered when HX-Request header present:

• Pass HTTP_HX_REQUEST="true" to client request
• Assert response.templates contains partial template name

Testing Permissions

Create authenticated vs anonymous client fixtures:

• Test redirect/403 for unauthorized access
• Test success for authorized access

Testing QuerySets

Verify efficient queries:

• Create test data with factories
• Execute query
• Assert correct objects returned/excluded
• Verify related objects loaded with select_related()/prefetch_related()

Testing Forms with Model Instances

Pass instance to form for updates:

• form = MyForm(data=new_data, instance=existing_obj)
• Verify form.save() updates, doesn't create

Common Patterns

Parametrize multiple scenarios: Use @pytest.mark.parametrize("input,expected", [...]) for testing various inputs

Mock external services: Use mocker.patch() to avoid actual HTTP calls, emails, file operations

Check database changes:

• Assert Model.objects.filter(...).exists() after creation
• Assert Model.objects.count() == expected for deletions
• Use refresh_from_db() to verify updates

Test error handling:

• Invalid form data produces correct errors
• Failed operations return error responses
• User sees appropriate error messages

Running Tests

uv run pytest                    # All tests
uv run pytest -x                 # Stop on first failure
uv run pytest --lf               # Run last failed
uv run pytest -x --lf            # Stop first, last failed only
uv run pytest -k "test_name"     # Run tests matching pattern
uv run pytest tests/apps/posts/  # Specific directory
uv run pytest --cov=apps         # With coverage report

Common Pitfalls

• Forgetting @pytest.mark.django_db: Results in "Database access not allowed" errors
• Not using factories: Creating instances manually is verbose and brittle
• Testing implementation: Test behavior and outcomes, not internal implementation details
• Skipping TDD: Writing tests after code means tests follow implementation, missing edge cases
• Over-mocking: Mock external dependencies, not your own code
• Testing framework code: Don't test Django's ORM, form validation, etc. Test YOUR logic

Setup Requirements

In pyproject.toml:

[tool.pytest.ini_options]
DJANGO_SETTINGS_MODULE = "config.settings.test"
python_files = ["test_*.py"]
addopts = ["--reuse-db", "-ra"]

In conftest.py: Define shared fixtures (auth_client, common factories, etc.)

Integration with Other Skills

• systematic-debugging: When fixing bugs, write failing test first to reproduce
• django-models: Test custom managers, QuerySets, and model methods
• django-forms: Test form validation, clean methods, and save behavior
• celery-patterns: Test task logic with mocked external dependencies