Pytest Setup
Part 1, Chapter 6
Let's get our tests up and running for this endpoint with pytest.
Setup
Add a "tests" directory to the "project" directory, and then create the following files inside the newly created directory:
- __init__.py
- conftest.py
- test_ping.py
By default, pytest will auto-discover test files that start or end with test -- e.g., test_*.py or *_test.py. Test functions must begin with test_, and if you want to use classes they must also begin with Test.
Example:
# if a class is used, it must begin with Test
class TestFoo:
# test functions must begin with test_
def test_bar(self):
assert "foo" != "bar"
If this is your first time with pytest be sure to review the Installation and Getting Started guide.
Fixtures
Define a test_app fixture in conftest.py:
# project/tests/conftest.py
import os
import pytest
from starlette.testclient import TestClient
from app import main
from app.config import get_settings, Settings
def get_settings_override():
return Settings(testing=1, database_url=os.environ.get("DATABASE_TEST_URL"))
@pytest.fixture(scope="module")
def test_app():
# set up
main.app.dependency_overrides[get_settings] = get_settings_override
with TestClient(main.app) as test_client:
# testing
yield test_client
# tear down
Here, we imported Starlette's TestClient, which uses the Requests library to make requests against the FastAPI app.
To override the dependencies, we used the dependency_overrides attribute:
main.app.dependency_overrides[get_settings] = get_settings_override
dependency_overrides is a dict of key/value pairs where the key is the dependency name and the value is what we'd like to override it with:
- key:
get_settings - value:
get_settings_override
Fixtures are reusable objects for tests. They have a scope associated with them, which indicates how often the fixture is invoked:
- function - once per test function (default)
- class - once per test class
- module - once per test module
- session - once per test session
Check out All You Need to Know to Start Using Fixtures in Your pytest Code for more on pytest fixtures.
Take note of the test_app fixture. In essence, all code before the yield statement serves as setup code while everything after serves as the teardown.
For more on this review Fixture finalization / executing teardown code.
Then, add pytest and Requests to the requirements file:
pytest==6.2.5
requests==2.26.0
We need to re-build the Docker images since requirements are installed at build time rather than run time:
$ docker-compose up -d --build
With the containers up and running, run the tests:
$ docker-compose exec web python -m pytest
You should see:
=============================== test session starts ===============================
platform linux -- Python 3.10.1, pytest-6.2.5, py-1.11.0, pluggy-1.0.0
rootdir: /usr/src/app
plugins: anyio-3.4.0
collected 0 items
============================== no tests ran in 0.07s ==============================
Tests
Moving on, let's add a quick test to test_ping.py:
# project/tests/test_ping.py
from app import main
def test_ping(test_app):
response = test_app.get("/ping")
assert response.status_code == 200
assert response.json() == {"environment": "dev", "ping": "pong!", "testing": True}
While unittest requires test classes, pytest just requires functions to get up and running. In other words, pytest tests are just functions that either start or end with test.
You can still use classes to organize tests in pytest if that's your preferred pattern.
To use the fixture, we passed it in as an argument.
Run the tests again:
$ docker-compose exec web python -m pytest
You should see:
=============================== test session starts ===============================
platform linux -- Python 3.10.1, pytest-6.2.5, py-1.11.0, pluggy-1.0.0
rootdir: /usr/src/app
plugins: anyio-3.4.0
collected 1 item
tests/test_ping.py . [100%]
================================ 1 passed in 0.13s ================================
Given-When-Then
When writing tests, try to follow the Given-When-Then framework to help make the process of writing tests easier and faster. It also helps communicate the purpose of your tests better so it should be easier to read by your future self and others.
| State | Explanation | Code |
|---|---|---|
| Given | the state of the application before the test runs | setup code, fixtures, database state |
| When | the behavior/logic being tested | code under test |
| Then | the expected changes based on the behavior | asserts |
Example:
def test_ping(test_app):
# Given
# test_app
# When
response = test_app.get("/ping")
# Then
assert response.status_code == 200
assert response.json() == {"environment": "dev", "ping": "pong!", "testing": True}
✓ Mark as Completed