Unit Testing

Unit testing is essential to ensure software quality. Given our software is heavily data based. We developed a minimal testing module to create unit tests. The idea behind the module is similar to some existing Python unit testing modules. For better compatibility, the new version of the unit_test class will simply be a wrapper of the existing Python unittest module, so that it can be easily detected by existing Python unit testing modules. This wrapper will ensure easier creation of tests and minimize redundant codes.

Creating a test

Below is an example of a unit test case. If pre_test method is defined, the unit test engine will execute the pre_test method to prepare the tests. Then if will find all the methods begins with test_ and execute them. To compare the test results, the engine will compare all the non-callable attributes that do not start with _, with a provided data source (A dictionary, a yaml file or another instance of a test). Assertion errors will be raised if the comparison failed.

from unit_test import unit_test

class this_is_a_test(unit_test):

    def pre_test(self):
        # Pre-testing set up
        pass

    def test_1(self):
        # Testing actions
        pass

Running tests

• One can run tests by instantiating a class and do comparison python   atest = this_is_a_test() By default, the option in the constructor check_after_run is True, and the code will try to open data source located at output/<class name>.yaml, and compare the results. Or the method assert_compare_with can be called manually providing the data source.
• Or, one can execute the code unit_test.py directly, and use the provided daemon to search all tests in the current directory recursively and execute them. Any tests with the name starting with _ will be ignored. One can utilize this to write secondary wrappers.