From: Hans Aberg (haberg@math.su.se)
Date: Sat May 14 2005 - 05:46:07 CDT
At 12:25 +1100 2005/05/14, Andrey V. Panov wrote:
>On Saturday 14 May 2005 05:53, Adam Twardoch wrote:
>> Such things are addressed on the font level, using OpenType Layout
>> features. You can include variant glyphs if the guillemots, and create a
>> "locl" OpenType feature for the Russian language.
>
>As for me it would be better to create an unicode font with Russian style
>guillemets as default, and French style ones as "locl" alternatives.
I think it would be possible to add abstract characters to Unicode
representing scripts (or languages, localities), as these clearly are
atomic linguistic semantic data. To this, one would need to add begin
and end abstract characters, used to represent stacked data.
Such a technique could be used to represent what is in Unicode lingo
called BiDi data: Most natural languages have a natural, inherent
reading direction, "forward". To focus on something explicit, the
parenthesizes "(" and ")" could be named, neutral with respect to
rendering direction, "OPENING PARENTHESIS" and "CLOSING PARENTHESIS".
There there are abstract characters indicating rendering direction;
alternatively, a default might be indicated via a script marker. This
way, a rendering program can compute the current rendering direction
and choose the correct glyph.
-- Hans Aberg
This archive was generated by hypermail 2.1.5 : Sat May 14 2005 - 05:47:16 CDT