Multiple test functions rely on files that are not included in the repo by default

[liammegill]

Type of issue

• bug report

Description

A number of test functions (listed below) rely on files only created after running utils/create_test_files.py.

Current behaviour: On github, we have configured the automatic tests to first run this script, so there is never an error. However, if you run the tests locally and have not run the create_test_files script, you may be left wondering why there are so many errors (like I was).

Test functions that produce errors if the test files are not present:

• calc_response_test.py: all functions
• construct_conc_test.py: all functions
• interpolate_space_test.py: all functions
• read_config_test.py/TestLoadConfig: test_type and test_invalid
• read_netcdf_test.py/TestCheckSpecAttributes: all functions

Expected behaviour: The user should be able to run the tests immediately after cloning the repository.

Suggestions for fix: I can think of three options:

• Preferred option: Always use simple versions of the required files (config, response, inventory) that do not need to be saved to file, e.g. using a pytest fixture.
• Run the create_test_files.py script from the test scripts if the required files are not present. This is not ideal, because you do not want the test scripts to result in saved files.
• Write in the documentation/README that you need to run create_test_files.py before testing the program (currently implemented).

Environment:

• oac version: v2.8.3
• Operating system: Windows 11