ROHIT SANJAY Unit Testing in Python BELPY 2021

Hi I’m Rohit Sanjay. - Final Year Engineering Undergraduate at MIT Manipal - Author and maintainer of testbook (under nteract org) - Incoming Software Engineer Intern at Goldman Sachs, Bengaluru rohitsanjay.com

--Boris Beizer, in "Software testing techniques" [source] More than the act of testing, the act of designing tests is one of the best bug preventers known.

Agenda • Introduction to unit testing • Writing unit tests ❖ unittest vs pytest ❖ Mocking ❖ Using context managers ❖ PyTest features - f i xtures and parametrized tests

Introduction to Unit Testing

• A unit test is a test that checks that a single component operates in the right way. • A unit test helps you to isolate what is broken in your application and fix it faster. Unit Testing in two sentences That’s it! Moving on..

unittest vs pytest

unittest pytest

• Too much boilerplate code is needed. • The API can make the test code's intent hard to understand • camelCase naming unittest

• Detailed info on failing assert statements • Auto-discovery of test modules and functions (searches for test_*.py or *_test.py files) • Modular fixtures for managing small or parametrized long-lived test resources • Tests parametrization • Can run unittest test suites out of the box • Rich plugin architecture, with over 315+ external plugins and thriving community Why pytest is the better option..

Also, you have to admit, pytest looks so much better

Mocking Source: https://www.mathworks.com/help/matlab/mocking-framework.html Simulates the look and feel of real world objects.

unittest.mock Mock()

No content

unittest.mock Mock() patch()

No content

Mock an object where it is used, not where it came from.

When to use mocks

System calls jovian/utils/environment.py jovian/tests/utils/test_environment.py

Network/API calls jovian/utils/api.py jovian/tests/utils/test_api.py

Network/API calls jovian/tests/utils/test_api.py Using patch as decorator

Network/API calls jovian/tests/utils/test_api.py What is side_effect?

A little about side_effect When side_effect is an exception

A little about side_effect When side_effect is a function (or any callable)

A little about side_effect When side_effect is an iterable Let us get back to when we should use mocks

• System calls • Network/API calls • I/O operations • Unpredictable results (random) When to use mocks

Using context managers

• It’s a simple “protocol” (or interface) that your object needs to follow so it can be used with the with statement. • Basically all you need to do is add __enter__ and __exit__ methods to an object if you want it to function as a context manager. • Python will call these two methods at the appropriate times in the resource management cycle. Context Managers in Python Are you with me?

No content

contextlib

With that context managed, we can now move on to the next part

Context managers in unit tests jovian/tests/resources/shared.py @contextmanager def temp_directory(): orig_path = os.getcwd() try: with TemporaryDirectory() as dir: os.chdir(dir) yield dir finally: os.chdir(orig_path)

@contextmanager def mock_git_repo(): orig_dir = os.getcwd() with temp_directory() as dir: os.mkdir("mock_git_repo") os.chdir("mock_git_repo") commands = textwrap.dedent(""" git init git remote add origin https://github.com/JovianML/mock_repo.git touch notebook.ipynb git add . git commit -m "initialcommit" """).splitlines()[1:] for command in commands: check_call(command.split()) yield dir os.chdir(orig_dir) jovian/tests/resources/shared.py

pytest

• capsys - Capture, as text, output to sys.stdout and sys.stderr. • testdir - Provide a temporary test directory to aid in running, and testing, pytest plugins. • tmp_path - Provide a pathlib.Path object to a temporary directory which is unique to each test function. • tmpdir - Provide a py.path.local object to a temporary directory which is unique to each test function; replaced by tmp_path. pytest fixtures

Custom pytest fixtures import pytest @pytest.fixture def smtp(): import smtplib return smtplib.SMTP("smtp.gmail.com") def test_ehlo(smtp): response, msg = smtp.ehlo() assert response == 250 assert 0 # for demo purposes

Custom pytest fixtures import pytest from click.testing import CliRunner @pytest.fixture(scope='module') def runner(): # Get instance of CliRunner runner = CliRunner() return runner jovian/tests/utils/test_cli.py

pytest parametrize import pytest @pytest.mark.parametrize( "test_input, expected", [ ("3+5", 8), ("2+4", 6), ("6*9", 42) ] ) def test_eval(test_input, expected): assert eval(test_input) == expected

pytest parametrize @pytest.mark.parametrize( "urls, expected_result", [ (["https://jovian.ml"], "https://jovian.ml"), (["https://jovian.ml/"], "https://jovian.ml/"), (["///user/siddhant"], "user/siddhant"), (["https://jovian.ml", "user/siddhant"], "https://jovian.ml/user/siddhant"), (["https://jovian.ml", "/user/siddhant"], "https://jovian.ml/user/siddhant"), ] ) def test_urljoin(urls, expected_result): assert urljoin(*urls) == expected_result jovian/tests/utils/test_misc.py

Memes

No content

No content

No content

No content

That’s all folks.

• To mock, or not to mock - PyCon 2016 • Context Managers and the “with” Statement in Python References