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
For a file with parser error; external-test-runner is better
git checkout 03af25561e00a71b44cadc39c4020e68d9b356fa
clojure -M:poly test
clojure -M:poly test with:external-test-runner
Scenario 2: empty file
Both runners are not very helpful in case of an empty file. Unfortunately this is a realworld problem; my emacs creates empty files when I open a new buffer. If I forget to save it stays empty.
git checkout bc8409004707644f1af5fc0f8a0c83ce5e8fc001
clojure -M:poly test
clojure -M:poly test with:external-test-runner
For the empty file case both runners do not help (much), it would require some extra checks I guess.
Expected behavior
In general Polylith knows about the syntax problem and it gives a warning as a first output. However, in a big project this is easy to miss since the runner just continues with other output. The final error has nothing to do with the actual error. If instead Polylith would do a hard exit when running into the problem (by default or even as option to configure) this would be a big improvement.
The main improvement from the external-test-runner is that it is showing you the error at the end of the run.
Unfortunately for the empty file, both runners are not helping much, but arguably the external-test-runner is slightly more helpful.
Screenshots
Polylith default test runner in both scenario's
Polylith external test runner in scenario 1
Polylith external test runner in scenario 2
Versions (please complete the following information):
Describe the bug
Certain type of errors in my project are not clearly reported by Polylith. I've demostrated two of them in this project:
First clone the repo
Scenario 1: parser error
For a file with parser error; external-test-runner is better
Scenario 2: empty file
Both runners are not very helpful in case of an empty file. Unfortunately this is a realworld problem; my emacs creates empty files when I open a new buffer. If I forget to save it stays empty.
For the empty file case both runners do not help (much), it would require some extra checks I guess.
Expected behavior
In general Polylith knows about the syntax problem and it gives a warning as a first output. However, in a big project this is easy to miss since the runner just continues with other output. The final error has nothing to do with the actual error. If instead Polylith would do a hard exit when running into the problem (by default or even as option to configure) this would be a big improvement.
The main improvement from the external-test-runner is that it is showing you the error at the end of the run.
Unfortunately for the empty file, both runners are not helping much, but arguably the external-test-runner is slightly more helpful.
Screenshots
Polylith default test runner in both scenario's
Polylith external test runner in scenario 1
Polylith external test runner in scenario 2
Versions (please complete the following information):
The text was updated successfully, but these errors were encountered: