You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently some tests share the same setup, e.g. read the same input files, which can potentially cause problems and make maintenance harder. It would be good for every test to be responsible for its input and output. It's probably better to also delete all files associated with a test after it's done.
The text was updated successfully, but these errors were encountered:
One issue that arises by not deleting test output files is that in some cases (e.g. mapping) we only go ahead with the process if the output file doesn't already exist. This means that if we change the relevant code between two test runs then the updated code won't be tested because the test output file will already exist.
Another issue is that in the mapping tests we also perform parsing, so the test will fail if there is something wrong with the parser, even if the mapping is fine. I think it's better for the mapping tests to receive the parser output as input directly so that they are independent of the parser's code.
Currently some tests share the same setup, e.g. read the same input files, which can potentially cause problems and make maintenance harder. It would be good for every test to be responsible for its input and output. It's probably better to also delete all files associated with a test after it's done.
The text was updated successfully, but these errors were encountered: