-
As part of CI/CD, all features under a directory get triggered for every code merged. Time to time, the overall Behave test fails with exit code 1. It stated that there is 1 feature failed, but 0 scenarios failed. This is very bothersome because the testing job gets marked as failed on the pipeline (which is not true because all scenarios are passing) I noticed this false negative behavior can happen when new scenarios or features get added, and it gets fixed by itself when another new scenarios/features gets added again. However, I could not figure out what is causing it. For sure, it's not the CI/CD related issue because I can reproduce the bug when I run locally on my laptop. Here is the snapshot of the behave summary which states 1 feature failed, but 0 scenario failed (Note that Behave will list the failed scenarios at the bottom summary, but in this case, none is listed because there is no scenario failed). is this a bug in Behave? Any help will be appreciated. Thank you! |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
Has one of the hooks failed that caused one feature to fail? HINT:
|
Beta Was this translation helpful? Give feedback.
Has one of the hooks failed that caused one feature to fail?
Like:
before_feature()
orafter_feature()
hookHINT:
untested
Scenarios and steps are hinting for that.Probably, a
before_feature()
or abefore_tag()
hook (applied to a feature) failed which caused to skip the scenarios (and their steps) of this feature. Meaning, the failed feature should contain 11 untested scenarios.HOOK-ERROR
.