Testing with Fixtures

Goals

• Understand why putting a fixture in one test file makes it unavailable to other test files.
• Use conftest.py to define fixtures that pytest shares automatically across all test files in a directory.
• Split tests into multiple files, each focused on one part of the application.

The Duplication Problem

The previous lesson put small_db directly in test_server_db.py. What happens if you want to write a second test file that also needs a small database?

• A fixture defined inside a test file is private to that file
  • Importing it from another test file would work, but it breaks pytest's fixture discovery
  • Copying the fixture into every new test file means maintaining the same setup code in multiple places
  • If the database schema changes, every copy has to be updated separately
• The fixture is doing real work: creating a temporary database, inserting rows, and returning a path
  • That work belongs in one place, shared by all the tests that need it

Where does pytest look when it needs a fixture that is not defined in the current test file?

• pytest looks for a file named conftest.py in the same directory as the test file, then in each parent directory up to the project root
  • No import statement is required: fixtures defined in conftest.py are available to every test file in the same directory and all subdirectories
  • pytest finds conftest.py by convention, not by any explicit reference in the test files

Extracting the Fixture

Move the SMALL list, the SQL constants, and the small_db fixture from test_server_db.py into a new file called conftest.py in the same directory.

• Everything the fixture needs moves with it: SMALL, CREATE_TABLE, and INSERT_ROW
  • Test files that use small_db do not import anything from conftest.py; pytest injects the fixture automatically based on the parameter name
• The fixture itself is unchanged: it still uses tmp_path to get a fresh directory, creates the database, inserts the two rows, and returns the path

import sqlite3

import pytest

from schema import CREATE_TABLE
from small import SMALL

INSERT_ROW = "insert into sightings values (?, ?, ?, ?, ?, ?, ?, ?)"


@pytest.fixture
def small_db(tmp_path):
    db_path = tmp_path / "test.db"
    conn = sqlite3.connect(db_path)
    conn.execute(CREATE_TABLE)
    for s in SMALL:
        conn.execute(
            INSERT_ROW,
            [
                s["id"],
                s["species"],
                s["sex"],
                s["weight"],
                s["color"],
                s["datetime"],
                s["latitude"],
                s["longitude"],
            ],
        )
    conn.commit()
    conn.close()
    return db_path

Testing in Separate Files

Write a file called test_index.py with tests that check the index route, and a file called test_detail.py with tests that check the detail route. Both files should use small_db.

• test_index.py imports only TestClient and make_app
  • small_db arrives as a parameter
  • test_index_ok checks that the home page responds with 200 and contains the site title
  • test_index_shows_both_species confirms both species from SMALL appear in the page
• test_detail.py also imports only TestClient and make_app
  • test_detail_ok visits sighting 1 and confirms the species name appears
  • test_missing_sighting confirms that a nonexistent ID returns 404
  • test_none_displayed_as_empty confirms that a sighting with null fields does not show the word "None" in the page

from litestar.testing import TestClient

from server_db import make_app


def test_index_ok(small_db):
    with TestClient(app=make_app(small_db)) as client:
        response = client.get("/")
    assert response.status_code == 200
    assert "Sasquatch Sightings" in response.text


def test_index_shows_both_species(small_db):
    with TestClient(app=make_app(small_db)) as client:
        response = client.get("/")
    assert "G. canadensis" in response.text
    assert "G. horribilus" in response.text

from litestar.testing import TestClient

from server_db import make_app


def test_detail_ok(small_db):
    with TestClient(app=make_app(small_db)) as client:
        response = client.get("/sighting/1")
    assert response.status_code == 200
    assert "G. canadensis" in response.text


def test_missing_sighting(small_db):
    with TestClient(app=make_app(small_db)) as client:
        response = client.get("/sighting/9999")
    assert response.status_code == 404


def test_none_displayed_as_empty(small_db):
    with TestClient(app=make_app(small_db)) as client:
        response = client.get("/sighting/2")
    assert response.status_code == 200
    assert "None" not in response.text

• Run both files together from the testdatabase/ directory:

pytest test_index.py test_detail.py -v

• pytest loads conftest.py once and makes small_db available to every test in both files
• Each test that lists small_db as a parameter gets its own fresh database, so a write in one test cannot affect another

Check Understanding

See [pytest2025] for the full pytest documentation, including the reference page on fixtures and conftest.py.

from conftest import small_db

def test_index_ok(small_db):
    ...

def test_missing_sighting(small_db):
    with TestClient(app=make_app(small_db)) as client:
        response = client.get("/sighting/9999")
    assert response.status_code != 200

Exercises

Add a Test for the CSS Route

Add a test to test_index.py that sends GET /style.css. It must check that the response status code is 200 and that the response text contains "table".

Test a Specific Detail

Add a test to test_detail.py that visits /sighting/2 and confirms that "G. horribilus" appears in the response.

Add a Third Test File

Create test_links.py that uses small_db to check that the index page contains two links inside the table and that the detail page for sighting 1 contains a link whose text is "Back to all sightings". You will need to look for the right strings in response.text.

Change the Test Data

Add a third row to SMALL in conftest.py with a species of "G. canadensis" and a different color. Without changing any test assertions, run the suite and confirm that all tests still pass, then explain why test_index_shows_both_species still passes even though the number of rows changed.

Confirm Isolation

Add a test to test_detail.py that inserts a third row directly into small_db using sqlite3, then checks that the index page shows three links. Run the full suite and confirm that test_index_ok still passes with only two links, which shows that each test received its own database.