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
Rucio used to have test coverage some time ago, but it was toggled off. I think it would be a good tool to have for PR review purposes; if there is a portion of code that is not covered, it might be a good indicator that some tests need to be written for that part of the code.
There are some criticisms for using coverage as a metric: the main reason behind this is that coverage only tells you whether code is covered or not, but it doesn't tell you much about the quality of the tests. I somewhat agree with this, and I would suggest only using coverage as a PR review tool, but we could also consider having a fairly achievable coverage threshold (60-70%) and aiming to always be over that threshold.
The text was updated successfully, but these errors were encountered:
Rucio used to have test coverage some time ago, but it was toggled off. I think it would be a good tool to have for PR review purposes; if there is a portion of code that is not covered, it might be a good indicator that some tests need to be written for that part of the code.
As we use pytest for testing, we can use https://github.com/pytest-dev/pytest-cov for coverage.
There are some criticisms for using coverage as a metric: the main reason behind this is that coverage only tells you whether code is covered or not, but it doesn't tell you much about the quality of the tests. I somewhat agree with this, and I would suggest only using coverage as a PR review tool, but we could also consider having a fairly achievable coverage threshold (60-70%) and aiming to always be over that threshold.
The text was updated successfully, but these errors were encountered: