test: Remove moq library

[askpt]

Remove Moq library

• This PR removes the Moq library from the test projects. Now it uses NSubstitute.

Related Issues

Fixes #140

Notes

Follow-up Tasks

How to test

[benjiro]

@askpt I'm fine with this going a head, thanks for doing it 👍🏻

Are you able to look into the failing steps

[codecov]

Codecov Report

Merging #141 (b0aa44a) into main (15473b6) will decrease coverage by 0.74%.
Report is 1 commits behind head on main.
The diff coverage is n/a.

@@            Coverage Diff             @@
##             main     #141      +/-   ##
==========================================
- Coverage   94.64%   93.91%   -0.74%     
==========================================
  Files          20       20              
  Lines         542      542              
  Branches       39       39              
==========================================
- Hits          513      509       -4     
- Misses         16       20       +4     
  Partials       13       13              

see 1 file with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[benjiro]

Aah looks like the build is breaking for other reasons, ill take a look this evening after work

[askpt]

@benjiro The library is not fully removed in this PR. There are still a few tests missing that have more advanced scenarios. I don't mind fixing the build issues and creating a new issue for these tests.

[benjiro]

If you rebase off main it will fix the warnings that are making E2E Test, and Test linux fail.

error EnableGenerateDocumentationFile: Set MSBuild property 'GenerateDocumentationFile' to 'true' in project file to enable IDE0005 (Remove unnecessary usings/imports) on build (https://github.com/dotnet/roslyn/issues/41640) [/home/runner/work/dotnet-sdk/dotnet-sdk/test/OpenFeature.E2ETests/OpenFeature.E2ETests.csproj::TargetFramework=net6.0]

[bacherfl]

thanks for doing this and great work with the adapted tests so far - I've just had a look and all the assertions we previously had in place are still present.

[bacherfl]

This looks good to me! For me it would be ok to tackle the remaining test cases in a follow up - @benjiro what do you think?

[askpt]

This looks good to me! For me it would be ok to tackle the remaining test cases in a follow up - @benjiro what do you think?

Pushed more tests now. Only a handful missing. We can merge as it is.

[bacherfl]

This looks good to me! For me it would be ok to tackle the remaining test cases in a follow up - @benjiro what do you think?

Pushed more tests now. Only a handful missing. We can merge as it is.

Thanks! Regarding the failing codecov check: This seems to be caused by a different behavior of NSubstitute vs Moq. The difference is in the Register_Hooks_Should_Be_Available_At_All_Levels in the OpenFeatureHookTest - This was previously covering the exception handling in the Error hook, as Moq mocks throw an exception for methods where no behavior has been defined, while NSubstitute does not seem to do that, therefore we do not reach this catch block in the aforementioned test.
Given that the purpose of this test is just to check if the Register method is available on all levels to register hooks, I think covering the exception handling of the Error hook can be part of a dedicated PR for that

[askpt]

@bacherfl

This looks good to me! For me it would be ok to tackle the remaining test cases in a follow up - @benjiro what do you think?

Pushed more tests now. Only a handful missing. We can merge as it is.

Thanks! Regarding the failing codecov check: This seems to be caused by a different behavior of NSubstitute vs Moq. The difference is in the Register_Hooks_Should_Be_Available_At_All_Levels in the OpenFeatureHookTest - This was previously covering the exception handling in the Error hook, as Moq mocks throw an exception for methods where no behavior has been defined, while NSubstitute does not seem to do that, therefore we do not reach this catch block in the aforementioned test. Given that the purpose of this test is just to check if the Register method is available on all levels to register hooks, I think covering the exception handling of the Error hook can be part of a dedicated PR for that

These tests are the ones that miss the transition to NSubstitute:

• Hook_Hints_May_Be_Optional
• When_Error_Occurs_In_Before_Hook_Should_Return_Default_Value
• When_Error_Occurs_In_After_Hook_Should_Invoke_Error_Hook

I was trying to play around but they are failing because the sequence is not respected.
I would suggest creating a new issue to migrate these three tests and I can try to fix it later.

See output:
`Expected to receive these calls in order:

Hook.Before<Boolean>(any HookContext<Boolean>, Dictionary<String, Object>)
FeatureProvider.ResolveBooleanValue(any String, any Boolean, any EvaluationContext)
Hook.After<Boolean>(any HookContext<Boolean>, any FlagEvaluationDetails<Boolean>, Dictionary<String, Object>)
Hook.Error<Boolean>(any HookContext<Boolean>, any Exception, Dictionary<String, Object>)
Hook.Finally<Boolean>(any HookContext<Boolean>, Dictionary<String, Object>)

Actually received matching calls in this order:

FeatureProvider.ResolveBooleanValue("test", True, EvaluationContext)`

[benjiro]

Thanks for your efforts @askpt, I'll look at the remaining tests tomorrow. I'll merge this or as is now