The logging implementation #103, has revealed that many of the tests rely on standard out.
As per meeting 18/12/23 we should probably be testing the functionality (where possible) than the messages (print statements) that are emitted from the tested functions.
We can make use of temporary files/directories in GitHub actions, and can use a combination of subprocess mocking and dummy data to test functionality that would otherwise rely on an actual run of CABLE.
The logging implementation #103, has revealed that many of the tests rely on standard out.
As per meeting 18/12/23 we should probably be testing the functionality (where possible) than the messages (print statements) that are emitted from the tested functions.
For example: https://github.com/CABLE-LSM/benchcab/blob/0d2e415137f6338412742b85d0ecd481aaf3fc6b/tests/test_model.py#L173
We can make use of temporary files/directories in GitHub actions, and can use a combination of subprocess mocking and dummy data to test functionality that would otherwise rely on an actual run of CABLE.