Testing with pytest

The testing fundamentals lesson covered why and what to test. pytest is how it's done in Python — the de-facto standard, loved for letting you write tests as ordinary functions with plain assert.

A test is just a function

Put functions starting with test_ in a file starting with test_, and assert what you expect (arrange–act–assert from the testing lesson):

# test_math.py
from mymath import sum_to

def test_sums_small():
    assert sum_to(3) == 6      # arrange + act + assert in one line

def test_zero():
    assert sum_to(0) == 0

Run the whole suite by just typing pytest. It discovers your tests, runs them, and on failure shows the exact values that didn't match — no boilerplate assertion methods to memorise.

Testing that errors are raised

Confirm code fails when it should, using pytest.raises:

import pytest

def test_rejects_negative():
    with pytest.raises(ValueError):
        set_age(-1)

Fixtures and parametrization

Two features remove most test repetition:

• Fixtures provide reusable setup (a temp file, a database connection, sample data) to any test that asks for it by name — clean arrange steps, shared across tests.
• Parametrization runs one test over many inputs:

@pytest.mark.parametrize("n, expected", [(1, 1), (3, 6), (5, 15)])
def test_sum_to(n, expected):
    assert sum_to(n) == expected

That's three cases (the boundary-hunting habit from the testing lesson) in a few lines, each reported separately.

Where to go next

That completes the Python Advanced tier. The JavaScript/TS track has its own Advanced tier covering the runtime's internals, builds, and advanced types.