gvasold / pv

Repository for the "Programmieren vertieft" course

pv

Repository for the "Programmieren vertieft" course.

Topic: Testing

The purpose of the testing directory and all branches starting with 'testen-' is to demonstrate different (and improved) ways to test code.

Branch testen-1

This branch contains simpletests.py containing simple tests without using a test framework. This is only to demonstrate the core idea of automated tests. Normally you would always prefer to use a test framework like shown in the other branches.

Branch testen-2

This branch shows how to use the unittest Framework contained in the standard library. Personally, like most Python developers, I prefer the pytest framwork over the unittest framework.

The big difference to testen-1 is, the we use the unittest framework (which is heavily inspired by the junit Framework fpr Java). The framwork provides some test classes and assertion methods. It also contains a test runner, which is able to collect tests from different files, run these and presents the outcome.

Branch testen-3a

This branch uses the pytest framework. As it is not part of the standard library, there is a chance that you have to install it (pip install pytest).

This branch uses the framework in a very minimalistic way. All test modules are stored in a tests folder, which ist not strictly necessary, but a good practice. Notice that I put two test modules there which do not follow a common naming scheme: filter_test.py and test_reader.py. Normally you would settle on either test_xxx.py or xxx_test.py. To have test_ or _test.py in the file name is very important, because this is how the pytest test collector identifies test modules to run.

Branch testen-3b

This branch improves the pytest based tests in a first step:

• it unifies the names schema of the test files
• it moves the ugly test csv files to a temporary directory. This also demonstrates how to use pytest fixtures, a very powerful alternative to unittests setup() method

Branch testen-3c

• Create a conftest.py module. All fixtures defined in this module can be used directly in any test module
• Move the test data to fixture functions in the conftest.py file. The advantage of this is, that this data will be recreated by default for each test function - it is very important, that a test function cannot influence other test functions. Using a fixture for test data creation (or anything else which has might be changed by tests) solves this problem. Do not get confused: Here this is not really necessary, as the data is never changed during tests, but I take it as an example to show how to create data in fixtures.

Branch testen-3d

• This branch contains a new example person.py and test_person.py. It demonstrates how to test a class.
• Add the end of test_person.py I have added a small example to show how using a fixture guarantees, that the Person object is recreated for each test function (test_recreate1 and test_recreate2).

Coverage

In branch-3d we could also test the coverage function. Coverage means, that we get information about the percentage of code which is tested.

To use this functionality, we must install a pytest plugin called pytest-cov (pip install pytest-cov)

Then we can run this command:

pytest --cov

To find out which parts of the code are not tested, we use this command:

pytest --cov --cov-report=term-missing