Defining the pytest fixtures

Fixtures are used in pytest to prepare the context in which a test should be executed, preparing it and cleaning it at the end. The application fixture is expected by pytest-flask, as seen in the documentation. The plugin generates a client fixture that we can use to send requests in test mode. We see this fixture in action in the thoughts_fixture fixture, which generates three thoughts through the API and deletes everything after our test has run. 

The structure, simplified, is as follows:

1. Generate three thoughts. Store its thought_id:

@pytest.fixture
def thought_fixture(client):

    thought_ids = []
    for _ in range(3):
        thought = {
            'text': fake.text(240),
        }
        header = token_validation.generate_token_header(fake.name(),
                                                        PRIVATE_KEY)
        headers = {
            'Authorization': header,
        }
        response = client.post('/api/me/thoughts/', data=thought,
                               headers=headers)
        assert http.client.CREATED == response.status_code
        result = response.json
        thought_ids.append(result['id'])

1. Then, add yield thought_ids to the test:

yield thought_ids

1. Retrieve all thoughts and delete them one by one:

# Clean up all thoughts
response = client.get('/api/thoughts/')
thoughts = response.json
for thought in thoughts:
    thought_id = thought['id']
    url = f'/admin/thoughts/{thought_id}/'
    response = client.delete(url)
    assert http.client.NO_CONTENT == response.status_code

Note that we use the faker module to generate fake names and text. You can check its full documentation at https://faker.readthedocs.io/en/stable/. It is a great way of generating random values for your tests that avoid reusing test_user and test_text over and over. It also helps to shape your tests, by checking the input independently and not blindly copying a placeholder.

Also note the usage of the client fixture, which is provided by pytest-flask.