Hey guys, I am building a new test coverage service that calculates score based on how well the test coverage overlaps with how the codebase is used in production.
When I was working on upgrading Rails in a company I working at, the best practice as suggested by Rails guide is to increase test coverage to a level we can be confident.
But the traditional test coverage metric as tracked by Simplecov has some limitations in this particular use case: There was no way to tell which codes are being used irl — dead codes need not to be tested anymore and can be removed.
So I started working on OyenCov, a new test coverage service that factors in partial usages of our Rails applications, and tells us which parts of the code are being heavily used (implying commercial importance), which ones have low test coverage and overlaps, and which codes may be safe to be deleted.
Do give it a try, let me know what you think and how to improve it further!