pyinstrument: Given that we have a function that can turn all of our markdown files into html files ... what might be the best path forward for a speedup?
It may be not to change the function at all, but to instead change what files it updates. Maybe we should only update the files that updated such that we can leave all the other files alone. In a project like calmcode with 600+ files, this can yield a massive speedup!
The Big Picture
The main point we're trying to make here is that profiles won't solve everything for you. They will tell you how long your code takes to run, but they won't tell you how to make your code faster. They may even distract you a little bit!
That said, having a profiler around is certainly useful. And sofar pyinstrument seems to be a very likeable tool indeed!