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
I came accross this project as it is mentioned in HaFla and two Ducklings PRs are used there as Datapoints for buggy (and then fixed) programs.
However, I don't fully understand how Duckling runs it's tests / build.
I used GHCUP to set my GHC to 8.2 and cabal to 3.10.1.0, checking out Duckling at Commit ddbb6fd (Issue #196 and PR #199 ).
It touches 3 Files: A Classifier, a Rule and a Test.
I try to reproduce a failing CI by doing the following:
Which also fails after re-running the duckling-regen-exe.
If I revert both files, I get a passing TestSuite again.
In a similar manner, the other PR used (Commit d051632 , Issue #273) touches a Rule, Corpus, Classifier and Test.
If I only revert the Classifier, the expected Tests fail. If I revert any other element, tests pass.
I would expect that I have to revert the Rule and Regenerate the classifiers for a test-failure.
This behaviour is a bit opaque to me, and any explanation would be greatly appreciated.
The text was updated successfully, but these errors were encountered:
[ Info ] downloading: https://raw.githubusercontent.com/haskell/ghcup-metadata/master/ghcup-0.0.8.yaml as file /Users/kellybrower/.ghcup/cache/ghcup-0.0.8.yaml
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
[ Warn ] Assuming you meant version 8.2.2
[ Error ] [GHCup-00010] Unable to find a download for 8.2.2
[ Error ] Also check the logs in /Users/kellybrower/.ghcup/logs
Hello,
I came accross this project as it is mentioned in HaFla and two Ducklings PRs are used there as Datapoints for buggy (and then fixed) programs.
However, I don't fully understand how Duckling runs it's tests / build.
I used GHCUP to set my GHC to
8.2
and cabal to3.10.1.0
, checking out Duckling at Commit ddbb6fd (Issue #196 and PR #199 ).It touches 3 Files: A Classifier, a Rule and a Test.
I try to reproduce a failing CI by doing the following:
But it passes.
Instead I have to do:
git checkout ddbb6fd git revert ddbb6fd --no-commit git restore --staged ./Duckling/Ranking/Classifiers/DE_XX.hs git restore ./Duckling/Ranking/Classifiers/DE_XX.hs cabal test
Which also fails after re-running the duckling-regen-exe.
If I revert both files, I get a passing TestSuite again.
In a similar manner, the other PR used (Commit d051632 , Issue #273) touches a Rule, Corpus, Classifier and Test.
If I only revert the Classifier, the expected Tests fail. If I revert any other element, tests pass.
I would expect that I have to revert the Rule and Regenerate the classifiers for a test-failure.
This behaviour is a bit opaque to me, and any explanation would be greatly appreciated.
The text was updated successfully, but these errors were encountered: