i saw a video a few days ago that used "reverse a string" as a joke for "simple problem to solve" and oh god it's one of the most complicated problems i've thought about in a while.
there's like 3 layers of complexity
don't break utf8 (fine)
don't break combining characters (annoying, but text segmentation is already fairly well implemented)
properly reverse bidi control characters (oh god what the fuck)
i'm just imaginging that i'd have to reimplement the unicode bidi algorithm to be able to efficiently pair control characters and figure out which to swap
Add comment