Code coverage and mocks messing the analysis
Created by: jhchabran
As per https://sourcegraph.slack.com/archives/C07KZF47K/p1647903813941219?thread_ts=1647901745.907199&cid=C07KZF47K the current hypothesis is that by moving around 43k LOC of mocks, the code coverage took a dip because it started to be considered as code that should be covered.
- Check that it's the case.
- If yes, find a way for it to not be counted.