Bob Hallissy wrote:
> Section 2.6 "Combining characters" states that "Some specific combining
> characters override the default stacking behavior...", and goes on to
> illustrate with the Greek example of breathing + acute or grave accent. Is
> there a definitive list of the "specific" combining characters that should
> exhibit such exceptional behavior? Or are implementors left to
> discover the exceptions for themselves?
> Thank you.
I think this text may also be intended to refer to the "fixed form"
combining consonant characters 0fba, 0fbb, and 0fbc and maybe to the "fixed
form" consonant 0f6a - which, although it is not itself a combining
character, has properties shared with those three characters when it is
followed by a combining consonant from the range 0f90 to 0fbc.
These four characters were encoded to resolve a particular problem in
classical Tibetan and Dzongkha chos skad text:
e.g the sequence
0f68 0f62 0f6b should be rendered with a half (yatag) form combining YA
while the sequence 0f68 0f62 0fbb should be rendered with a full form.
the sequence 0f62 0fa0 should be rendered with a (normal) half (rago) form
RA while the sequence 0f6a 0fa0 should be rendered with a full form RA
although it is followed by a combining consonant with which RA would normlly
take a half form.
(BTW These characters are not a case of encoding simple glyph variants since
in Dzongkha and Tibetan text some combinations like these represent attested
lexically distinct items which would otherwise have become confused.)
I don't know enough about the intricacies of individual scripts outside the
block 0f00 to 0fcf - and all the languages which they cover - to hazard a
guess as to which other characters may be covered in the paragraph you
quote. If this is not something which can be gleaned from the character
properties database, a list of these characters would be good idea - though
no matter how much information one may have on character properties and
rendering it is pretty risky to implement any script without the aid of an
expert or experts on that script and the languages which it is used for.
[For instance with the script used for the Dzongkha and Tibetan languages
you would still need to know which combining consonants normally would cause
a RA to take a half form and which normally wouldn't - and the rules for
this vary with different forms of the same script. Its probably not
reasonable to expect the Unicode or ISO 10646 standards to go into that
level of detail for each script.]
- Chris Fynn
-- Christopher J. Fynn <firstname.lastname@example.org> DDC Dzongkha Computing Project Thimphu, Bhutan.
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:15 EDT