Test Template¶

A test should be written for every feature.

1. Create a new python file to contain the feature test code.

   For example, if the feature concerns x_feature, the test case would be stored in testframework/test/../../test_x_feature.py.

2. Derive the test class from test.SingleNif and name it TestXFeature.

   test.SingleNif is designed upon the template pattern and takes care of all execution, loading of files and clean up.

   Note

   A template for a test class is available in testframework/test/template.py.

3. Each test needs to overwrite the following methods from test.SingleNif:

   • The n_create_data() used to create the physical .nif

   • The n_check_data() used to test the data of physical .nif

   • The b_create_data() used to mimic how the user would create the objects.

   • The b_check_data() check Blender data, used to check before export and after import.

These methods are intended to be delegate function, calling external methods in either n_gen_xxxx or b_gen_xxxx files.

Test Implementation¶

Each test has two paths of execution one importing/export a python generated nif, the other exporting/importing a nif created by mimicking how a user would create it.

Each test can be comprised of multiple required features and a lego block-building approach should be taken to building the nifs.

1. Recreate the nif using python:

   • Create a nif (say in nifskope, or with the old Blender nifscripts).

     Take care to make the file as simple as possible.

   • Use PyFFI’s dump_python spell to convert it to python code.

   • The n_create_data() method of the class will call the methods from n_gen_x_feature module to construct the physical nif.

     It may be required to call other reusable functions from other tests. Reusable functions should be created from the dump_python and stored in testframework/test/../../n_gen_x_feature.py.

   • Write code to test the desired features of the physical.

     The n_check_data() method will call the methods from n_gen_x_feature module to check the nif data.

2. Recreate the feature within Blender, using user functions:

   • Write Python code which recreates the corresponding data in the Blender scene in testframework/test/../../b_gen_x_feature.

   • Where possible make the test case as simple as possible.

     For instance, use primitives readily available in Blender. This code goes in the b_create_data() method of the test class.

   • Document the feature in docs/features/x_feature.rst as you write b_create_data():

     Explain what the user has to do in Blender in order to export the desired data, and wherein Blender the data ends up during import.

   • Write Python code which tests the Blender scene against the desired feature: b_check_data() method of the test class.

3. Implement the feature in the import and export plugin, until the regression test passes.

That’s it!

Execution Order¶

The tests will run like this: