Sander van Rossen🇺🇦 on Nostr: I used Unicode's own ICU (which is a library I'm not too happy about) for analyzing ...

I used Unicode's own ICU (which is a library I'm not too happy about) for analyzing unicode text, subdividing it into different language pieces & figuring out the direction of those pieces (words, if you will.. which is not an uniform language concept). I also used harfbuzz for text shaping those individual pieces.