From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Mon Jul 09 2007 - 21:13:47 CDT
Kenneth Whistler wrote:
> So I agree with John Hudson -- if there is a particular
> Latin letter (or symbol) that Lao vowels need to be
> displayed on for particular dictionary or pedagogical
> purposes, then that should be treated as a special case,
> rather than expecting such behavior as the general rule.
Isn't such exception already used with the dotted circle symbol?
May be one way to solve this problem is to abstain completely to use
reordering across scripts when they are letters, and instead define a
general rule for characters that are symbols shared by all scripts, i.e.
assigned with "Common" script, and general category "S*".
This way, one could use any other symbol as the glyph holding diacritics,
for example a square symbol, a dotted square, a circle symbol, a
mathematical cross.
One problem is that fonts (at least with TrueType/OpenType) are not designed
to support reordering and positioning with an unbound number of base
characters. For example the GSUB/GPOS tables in TrueType require listing
somewhere the complete list of codepoints where such reordering and
positioning may be applied, something that can't be performed in fonts with
the current format, because they don't allow defining character classes in
them, and assigning them pseudo-glyph IDs that can be used in GSUB tables.
It seems then more reasonable that renderers implement these character
classes, and recognize which fonts support such reordering : the renderer
for example could be looking for rules based on the dotted circle symbol,
and automatically infer the other applicable rules for other Common symbols,
and existing fonts don't need then to be updated: their existing rules for
reordering and positioning glyphs will work (notably with left-side Indic
vowels, or two-parts Indic vowels)
This archive was generated by hypermail 2.1.5 : Mon Jul 09 2007 - 21:17:18 CDT