We're using a mono-repo on our project (please don't try this at home), and running pylint vanilla on all folders on pre-push hook. Because the mono-repo is large, pylint takes around 47 (!) seconds. Obviously the best solution is to seperate the mono-repo, but in our use-case it's currently too hard. Is there a way a strategy to optimize pylint run on large repos? The best solution we came up with is to run it only on changed files, but this isn't good enough, since it won't detect files that imports deprecated functions. What would you do? Have you came across this issue? Continue reading...
