Noriaki Inouye wrote:
> I found a PDF file written in Oriya as follows:
> http://www.wbtc.com/articles/bibles/oriya/oriya_nt/Ori40Mt.pdf
>
> I can see some kinds of uniq ligatures on this file.
> I think these are not designed in the fonts like Arial Unicode MS,
> Lucida Sans Unicode,TITUS Bitstream Unicode,
I think that Lucida and Bitstream do not have ligatures for Indian
languages.
Arial Unicode MS has some them (at least for Devanagari, not sure about
Oriya) but, in order to display them, you need a new font technology called
OpenType. In practice, you need to install a component called UniScribe.
I have this component on my PC, but I don't know exactly how it arrived
here. I think that it was installed automatically together with Office 2000.
> Are these ligature encoded in the area from U+0B01 to U+0B6F
> of the Unicode 3.x Chart Map ?
No. In Unicode terminology, these ligatures are not characters: they are
"glyphs" that represent combinations or variations of characters.
Unicode text is encoded only by the "logical" characters that you see in the
charts. An "Unicode rendering engine" (the software that display the text)
is responsible for showing the ligature character where it is appropriate.
OpenType and UniScribe, that I mentioned above, are Microsoft's
implementation of an "Unicode rendering engine".
See the sections "Variant Shapes" and "Sequences" in:
http://www.unicode.org/unicode/standard/where/
For a more specific description of Indian scripts, see the Chapter 9 from
the book:
http://www.unicode.org/unicode/uni2book/ch09.pdf
The sections about Devanagari and Tamil explain quite well how the encoding
works, and the relationship between logical characters and displayed glyphs.
The sections about Oriya and other languages is not so accurate,
unfortunately, but the mechanism is analogous to the two "major" scripts.
HTH
_ Marco
This archive was generated by hypermail 2.1.2 : Fri Jul 06 2001 - 00:17:18 EDT