Re: [Boost-users] find japanese character with boost regex++

Aside from the unified Han characters (kanji/hanzi), characters of the same category generally aren't neatly grouped together in Unicode. The 128-character blocks tend to correspond to locales, communities of use or specific legacy encodings and not to categories. You need a look-up table (or more efficiently two levels of table) to check character categories.

I was talking about the Unicode block ranges (defined in blocks.txt from the Unicode ftp site), and you are correct that a two stage table is required for general categories. John