Lua Unicode normalization data. It's kind of similar to Skeleton algorithm from Unicode tr39, while it considers readability and cases.
Any unicode that looks similar to a latin letter is normalized to latin letters, even if it's a number or a punctuation. Characters are normalized by shape for latin letters, so Greek letter ν (lower case Nu) is normalized to latin letter V.
Chinese characters (a.k.a kanji) are normalized to Simplified Chinese as much as possible. The normalized Chinese sentence should be readable by native Chinese people.
The ccnorm.lua is automatically generated, so please report bugs in Issues. Do not send pull requests.