This ticket (and branch) are a place to capture my May 2024 Skunkworks project.
We currently have code coverage data for WT, generated by running a collection of tests, and achieving just over 50% code coverage. It would be very useful if it were possible to answer the questions such as the following very quickly/easily:
- I’m modifying function X, and/or these lines of code: which existing tests cover this? This would help make a developer’s life easier.
- Does a test (or set of tests) add code coverage compared to the rest of the set of tests? And if so, what? This would support optimising our testing.
I think that gathering the data should not be too hard (and builds on the work to parallelise code coverage) - the real challenge is how to report it to users in an easy to understand/use way (this will involve building our own versions of the code coverage HTML reports).
Note: the coverage data would have limitations due to some of our current code coverage tests being non-deterministic.