Closed holgerbrandl closed 2 years ago
Yes, I will look at all regressions.
The root cause of this issue is easy to fix. I will still have to work on the test case, which is not that easy.
I am putting here partial PR, which can be merged if the fix is required quickly: https://github.com/holgerbrandl/kscript/pull/350
Issue resolved - code fixed and test case added.
I've discovered an imho major regression.
Now the same with 3.1
So it seems that it's using some stale cache even if the script compilation failed. The correct behavior would be to show the same compiler error in the second run as well.
There was/is no test coverage for this process, so I guess have to blame myself my missing it in the first place. Anyway, @aartiPl would you have and idea and time to look into this problem? I still have to understand the revised impl better first.