From: Christopher Fynn (cfynn@gmx.net)
Date: Fri Jan 07 2005 - 10:01:56 CST
Andrew C. West wrote:
> Well it all depends. A text editor might import a GB18030 document with BrdaRten
> SetA characters, and using the code point mapping tables convert AAA1 etc. to
> U+E000 etc. The user then selects a BrdaRten font that maps precomposed BrdaRten
> glyphs to U+E000 etc. and everything is displayed correctly. This kind of
> support does not need any modifications to the mapping tables as the mapping of
> U+E000 to <0F40, 0F74> is irrelevant ... PUA characters are just PUA characters,
> and if you have the right font these PUA character will be rendered as
> precomposed BrdaRten glyphs.
>
> Of course if you then want to treat these PUA characters as real Unicode Tibetan
> you need to know the character mapping, but from my perspective character
> mapping is something that is optionally applied on top of the code point
> mapping.
As soon as you want to edit the text in a Unicode based application
you'd probably need to convert (or "character map") the BrdaRten PUA
characters to "real Unicode" [or you might end up with the horrors of a
kind of mixed encoding]. Comparing text from "real Unicode" with
precomposed Tibetan (PUA or GB18030), and collation would be difficult
without conversion as well.
- Chris
This archive was generated by hypermail 2.1.5 : Fri Jan 07 2005 - 10:08:15 CST