Re: [Devel] Enumerating font encoding

[Keith Packard]

Around 16 o'clock on Feb 1, George Williams wrote:

> Remember that format 2 is for CJK encodings like SJIS & Wansung where you 
> have latin (and katakana) characters encoded in a single byte and Kanji or 
> Hangul encoded with double bytes.

Now all is clear -- SJIS uses the low 128 spots for ASCII and the upper 
128 spots for 2-byte characters.  Thanks for the clarification; I matched 
the other format 2 example so my new code should "just work".

With the new code, I can now build a boolean array of available encoded 
chars in a tiny fraction of the original time.  Useful for selecting fonts 
based on what chars are supported.

Keith Packard        XFree86 Core Team        Compaq Cambridge Research Lab