On 2/12/2012 7:47 AM, Zdenek Wagner wrote:
Another possibility is to use ucharclasses...
This is a small sample (no need to change fonts manually):
\documentclass[12pt]{article}
\usepackage{fontspec}
\usepackage[Devanagari]{ucharclasses}
\setmainfont{Charis SIL}
\setTransitionTo{Devanagari}{\fontspec[Script=Devanagari]{Nakula}}
\begin{document}
I can write in English और वाक्य का एक भाग हिंदी में हो सकता है।
\end{document}
This sort of thing came up several years ago on this mailing list (I
don't recall the date, but it was before the 19 Dec 2010 date on the
current ucharclasses documentation). Some people on the list said that
this idea would break where there were ambiguous characters. I presume
that would mostly be punctuation. Some Unicode blocks have their own
punctuation (there's a danda, roughly the equivalent of a period = full
stop, in the above text), but often other blocks use ASCII or extended
Latin punctuation. I don't suppose that's much of a problem with
left-to-right scripts, but I can at least imagine it messing up
right-to-left scripts. A period or paren or comma in the middle of
Arabic, Hebrew, or Thaana text, say.
I skimmed the documentation, but it doesn't mention this situation. (It
does mention potential problems with English and Vietnamese.) However,
there is a \setTransitionsForPunctuation (sxn 4.4), and maybe you could
use this to work around the problem (if it is indeed a problem).
Has anyone worked with mixing L2R and R2L scripts using this, and if so,
did you run into problems?
--
Mike Maxwell
[email protected]
"My definition of an interesting universe is
one that has the capacity to study itself."
--Stephen Eastmond
--------------------------------------------------
Subscriptions, Archive, and List information, etc.:
http://tug.org/mailman/listinfo/xetex
