I'm doing a kind of archeology for my team's Git repo. The goal is to find all overlapping commits, e.g. all pairs of commits touching the same lines of code.
I tried to do this with diff and blame commands. The first one is not that reliable since code line ranges it provides have a sort of relative addressing. And it's hardly to track these addresses through 1000+ commits. The second option gives nice per file output but without any intermediate information (especially on deleting of a code lines).
I believe there must be a better way to achieve my goal. Looking forward for any advice!