Test duration values ​​are now presented in seconds with six digits of precision. Tests without time measurements have been edited.

[KaganCanSit]

Pull Request: Systematically Time All Test Suites (#4417) - Test duration values ​​are now presented in seconds with six digits of precision. Tests without time measurements have been edited.

Summary

• Enhanced Time Precision:
  Adjusted the printing of time values to display with 6 digits of precision in seconds. This improvement ensures more accurate and consistent timing information across all tests.

• Added Time Measurement
  Implemented timing calls for tests that previously did not measure execution time. This update affected several files to achieve full coverage of time measurements across all test
  While considering where to set these timing calls, I explored alternatives like placing the time measurement in the Result object's constructor and destructor. However, given that the Result objects are stored in a vector and altering the existing structure might have unforeseen side effects, I opted to integrate the timing calls in a way that preserves the current architecture.

• Additional Considerations
  During testing, I observed that time values are not always printed due to the if (m_ns_taken > 0) condition in the std::string Test::Result::result_string() function located in "test.cpp." After careful consideration, I decided to retain this condition, as it appears to function as expected and aligns with the intended logic.

Best regards

Note: I am aware that this change affects a lot of files and their content. However, I have tried to make the impact minimal while remaining true to the existing architecture. If you have any content that you would like to see edited, I would be happy to help you out as much as I can.

[coveralls]

coverage: 91.175% (+0.1%) from 91.063%
when pulling e9e7810 on KaganCanSit:feature_test
into e5ec408 on randombit:master.

[KaganCanSit]

Even though I checked, I think there are some points I missed. I will review why the checks failed and edit to fix it.

[randombit]

@KaganCanSit thank you but this isn't really at all how I planned on addressing this; the problem is precisely that right now the timing start/stop has to have manual calls inserted, and that is error prone and noisy.

[KaganCanSit]

@randombit I thought about this too. But I didn't want to change the current code structure. If we can exchange small tips/ideas for this, I would like to work on it.