Tibor Arpas

That is not unthinkable but quite difficult. Coverage data refers to the source code by line numbers, so it becomes invalid whenever you add or remove a line. I would...

As I wrote: > Coverage data refers to the source code by line numbers, so it becomes invalid whenever you add or remove a line. Also to have complete picture:...

This has been implemented and release to the extent it's technically practical/feasible. The results are not perfect, see discussion here: https://github.com/tarpas/pytest-testmon/issues/86

Yeah, unfortunatelly coveragepy doesn't handle reporting coverage of the same run to more independent callers. P.S. there is --testmon-off for exactly your use case. P.P.S.: I doubt using testmon in...

Which brings me to revelation that making coveragepy properly "re-entrant" would help also with issue #11 "dogfooding" (using testmon when developing testmon) which would be beautiful. A quick search doesn't...

#97 would solve the issue in general way

@webknjaz pytest-cov doesn't solve it either. You have to pick one of pytest-cov, pytest-testmon, Pycharm debugger. I renamed the issue, because it is a requested feature but, I don't have...

With testmon v 1.4.0+ it's easy to produce .testmondata and .coverage within the same run. However, there are still 2 issues, which are more on the side of coveragepy: *...

> Then is it possible to run the tests for the altered lines "first" and then run the remaining tests ? Yes. --testmon-noselect does exactly that. > That way CI...

What's the difference between reference run with and without testmon? As for your ideas, I don't know. I think first is finding out what the problem is. How many test...