Re: [OpenType] library for identifying equivalent sequences

Date: Wed Jul 31 2002 - 21:06:22 EDT

On 07/31/2002 05:46:02 PM Eric Muller wrote:

>Eg. don't you also want the strings that contain a sprinkling of ZWJ,
>ZWNJ, CGJ, SHY and various other things?

(Yuck.) Why, of course. (Bleecchh.) But it's easier to write an algorithm
to insert those than to derive the other. (Gag, choke.) So, I was just
asking for the harder part, but if you've got something to offer that
generates the myriad possibilities involving both... (Wretch.)

Ahem. Actually, I, like you, would much rather not have to mess with all
this within fonts -- would much rather have the software / font interface
deal with the equivalencies -- but the current state of our technologies
requires that, if we want our fonts to provide the same display for all of
the different possible sequences that ought to appear the same, then we
have to deal with each and every one. And when you're dealing with a Latin
font and want to support stacking of multiple-diacritic combinations (given
different possible orderings, different possible combinations encoded as
precomposed characters, and all those invisibles), that's a lot of

- Peter

Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <>

This archive was generated by hypermail 2.1.2 : Wed Jul 31 2002 - 19:20:49 EDT