Am 20.06.2011 00:03, schrieb Peter Baker:
It seems to me (having worked with OpenType fonts for some years) that
while it might be possible to make an Arabic-->Roman converter at the
font level, that's going to be one of the most inefficient possible ways
to handle it.
Then why is Unicode proposing / imposing it that way?
With OT you can make a set of rules that says
Here's a 1 followed by two digits; substitute a C;
Here's a 2 followed by one digit; substitute XX
Here's a 3 followed by a something other than a digit; substitute III.
But it can't understand numbers the way a programming language can do.
Does it have to? I thought arabic->roman to merely be a (somewhat
complex) substitution of glyphs.
If you want to be able to write XC for 90, the task gets somewhat more
complex, because OT definitely can't say
For a number in the range 90-99, do the following . . .
9 followed by one digit: substitute XC
9 followed by two digits: substitute CM
9 followed by something other than a digit: IX
(9 followed by three digits: MMMMMMMMM)
Surely a programmatic solution would be better; and (La)TeX has an
understanding of roman numbers built in. With a little Googling I was
able to come up with this file, which works:
%&program=xelatex
%&encoding=UTF-8 Unicode
\documentclass[11pt,letterpaper,twoside,openany]{book}
\usepackage{fontspec}
\makeatletter
\newcommand{\rmnum}[1]{\romannumeral #1}
\newcommand{\Rmnum}[1]{\expandafter\@slowromancap\romannumeral #1@}
\makeatother
\begin{document}
There are \rmnum{123}\ fish in the sea.
And there are \Rmnum{5123}\ leaves on the tree.
\end{document}
That one uses letters to represent roman numerals. As was said before,
that's not my intention, as roman numerals have a different meaning than
letters.
But thanks for the macro recommendation. Until now, I only knew of
\roman{counter}.
bye
Toscho
--------------------------------------------------
Subscriptions, Archive, and List information, etc.:
http://tug.org/mailman/listinfo/xetex
