Closed ilaflott closed 1 month ago
hooray! we have a test that will catch this now:
https://github.com/NOAA-GFDL/fre-cli/actions/runs/10291762258/job/28484761124
normally i would say do-not merge because the tests are failing, but this is an exception 🤣. @Ciheim is working on the fix- i just wanted to make sure the test we define fails under the right conditions.
This actually was not complete. i should have put it in draft mode, that was my mistake.
This test was designed to catch a current issue (CatalogBuilder
CI update), the name of the branch, "test_test_should_fail" indicated it's pipeline should have failed, but in fact, it succeeded. ergo, this code will not catch future errors, yet, and wasn't really ready for merge.
I'm not sure what kind of real review is being given to PRs, but i'd appreciate it if we all had a more careful eye
About line 155 on this page sounds like it describes that behavior:
https://github.com/NOAA-GFDL/fre-cli/actions/runs/10291762258/job/28484761124
=========================== short test summary info ============================ FAILED fre/tests/test_fre_cli.py::test_cli_fre_catalog_builder - assert 1 == 0
- where 1 = <Result AttributeError("module 'catalogbuilder.scripts.gen_intake_gfdl' has no attribute 'main'")>.exit_code ======== 1 failed, 64 passed, 2 skipped, 8 warnings in 85.71s (0:01:25) ======== Error: Process completed with exit code 1.
Is there something else that I should have been looking for?
catalogbuilder
's API changed, causingfre catalog builder
to throw an error. the tests should catch this but do not currently.updating with a test that will fail, which should show itself in the pipeline.