Python Testing With pytest

A test is a small piece of code that checks your real code gives the right answer, and pytest runs all your tests with one command.

Why Write Tests?

Every time you change a program, something old might break. Checking everything by hand is slow and easy to forget. Automatic tests check it for you in seconds.

You find bugs before your users do
You can change and improve code without fear
Tests show other developers how your functions are meant to work
Many freelance clients and companies expect tested code

Install pytest

Inside your virtual environment:

Example

bash

pip install pytest

The Code to Test

Here is a small module for a shop. Save it as shop.py:

Example

python

# shop.py
def price_with_tax(price, rate=0.18):
    if price < 0:
        raise ValueError("price cannot be negative")
    return round(price * (1 + rate))


def apply_discount(total, percent):
    return total - total * percent / 100

Your First Tests

Create test_shop.py in the same folder. Each test is a normal function whose name starts with test_. Inside it, use assert to state what must be true.

Example

python

# test_shop.py
import pytest
from shop import price_with_tax, apply_discount


def test_tax_is_added():
    assert price_with_tax(1000) == 1180


def test_custom_rate():
    assert price_with_tax(1000, rate=0.05) == 1050


def test_discount():
    assert apply_discount(2000, 10) == 1800


def test_negative_price_is_rejected():
    with pytest.raises(ValueError):
        price_with_tax(-50)

pytest.raises(ValueError) checks that the code inside does raise that error. If it does not, the test fails.

Running the Tests

In the terminal, in the project folder, type:

Example

bash

pytest

Output

============================= test session starts ==============================
platform linux -- Python 3.13.16, pytest-9.1.1, pluggy-1.7.0
rootdir: /home/ali/shop
collected 4 items

test_shop.py ....                                                        [100%]

============================== 4 passed in 0.02s ===============================

Each dot is a passing test. pytest found the tests by itself, because the file name starts with test_ and so do the function names.

Add -v (verbose) to see each test by name:

Example

bash

pytest -v

Output

test_shop.py::test_tax_is_added PASSED                                   [ 25%]
test_shop.py::test_custom_rate PASSED                                    [ 50%]
test_shop.py::test_discount PASSED                                       [ 75%]
test_shop.py::test_negative_price_is_rejected PASSED                     [100%]

============================== 4 passed in 0.02s ===============================

Reading a Failure

Suppose we write a test with a wrong expected value. 20% off Rs 1500 is Rs 1200, but we wrote 1300:

Example

python

def test_discount_wrong():
    assert apply_discount(1500, 20) == 1300

Output

test_fail.py F                                                           [100%]

=================================== FAILURES ===================================
_____________________________ test_discount_wrong ______________________________

    def test_discount_wrong():
>       assert apply_discount(1500, 20) == 1300
E       assert 1200.0 == 1300
E        +  where 1200.0 = apply_discount(1500, 20)

test_fail.py:5: AssertionError
=========================== short test summary info ============================
FAILED test_fail.py::test_discount_wrong - assert 1200.0 == 1300
============================== 1 failed in 0.02s ===============================

F means failed. The > points at the failing line, and the E lines show the real value (1200.0) next to the expected one (1300). Here the test was wrong, not the code. Sometimes it is the other way round, and that is a bug you just caught.

Many Inputs With parametrize

Instead of copying a test for every value, give pytest a table of inputs and expected results.

Example

python

import pytest
from shop import price_with_tax


@pytest.mark.parametrize(
    "price, expected",
    [
        (100, 118),
        (250, 295),
        (0, 0),
        (999, 1179),
    ],
)
def test_many_prices(price, expected):
    assert price_with_tax(price) == expected

pytest runs the test once for each row:

Output

test_more.py::test_many_prices[100-118] PASSED                           [ 16%]
test_more.py::test_many_prices[250-295] PASSED                           [ 33%]
test_more.py::test_many_prices[0-0] PASSED                               [ 50%]
test_more.py::test_many_prices[999-1179] PASSED                          [ 66%]

Fixtures: Shared Test Data

A fixture prepares something that several tests need. Mark a function with @pytest.fixture, then use its name as a parameter in a test. pytest passes in a fresh copy every time.

Example

python

import pytest


@pytest.fixture
def cart():
    return {"tea": 60, "samosa": 40, "juice": 120}


def test_cart_total(cart):
    assert sum(cart.values()) == 220


def test_cart_has_tea(cart):
    assert "tea" in cart

Output

test_more.py::test_cart_total PASSED                                     [ 83%]
test_more.py::test_cart_has_tea PASSED                                   [100%]

Useful Commands

Command	What it does
`pytest`	Runs every test in the folder
`pytest -v`	Shows each test by name
`pytest test_shop.py`	Runs one file
`pytest -k discount`	Runs tests whose name contains "discount"
`pytest -x`	Stops at the first failure
`pytest --lf`	Runs only the tests that failed last time

Write a test every time you fix a bug. The test proves the bug is gone and makes sure it never quietly comes back.

Exercise

How does pytest know which functions are tests?

They must be inside a class called TestTheir names start with test_ (in files named test_*.py)They must return TrueYou list them in requirements.txt