Closed kant2002 closed 1 year ago
It seems to be related to issue #2252. So I close this for now. Please reopen if this is unrelated.
No, if you notice that's different project for mutation. I assume that you mutate only project which I specify in command line. Please re-open
For the record, cannot reopen. It can be done only by you or another maintainer.
Sorry for the mix up, I am looking into this
Here are my findings:
The xUnit bugs demonstrate that Stryker does not handle this situation well and silently fails when tests are no longer detected after the initial testrun (dedicated to list available tests). What I find surprising is that no one discovered this problem before.
My first idea was to run dotnet stryker -p Refit.csproj -f net6.0
only to findout that -f
is for configuration file. But at least mayby --framework
can be passed from command line?
Time estimation is bad, it starts with 50 minutes, then drop to ~15 minutes and after that finished after 25 minutes.
Other than that, Refit provide approximately good results, some parts are not covered "by design", some things may be improved. I would say results are positive.
Current time estimation has several flaws; the biggest one is that there is a significant initialization time that leads to overestimated run time when starting. There is also a bias because mutations with high coverage are run at the end, which may give an impression that things are slowing down. But, mutant coverage is also important: test runs are slower when the mutants survive. The purpose of this estimate is to give a rough idea of how long it will take
Describe the bug When run stryker over Refit I receive report where https://github.com/reactiveui/refit/blob/569108a65079d364647f06437f977df8145fc578/Refit/RestMethodInfo.cs#L68 is marked as NoCoverage. But if I remove this line, tests are fails.
That's puzzling.
Steps
Expected behavior I would like that either mutation of this line was enabled, or something else what is meaningful from your perspective.