Sounds good.
If you want to check your Bidi code for compliance to the Unicode Bidi
algorithm, then you may want to try against the implementation in the IBM
Classes for Unicode, see
http://www10.software.ibm.com/developerworks/opensource/icu/ . It is
checked against an extensive test suite and behaves the same as the very
direct Java reference implementation that is going to be on the Unicode 3.0
CD.
This implements in C the full algorithm with implicit as well as explicit
rules and includes quite some optimizations, too. By nature of the
algorithm, the character database is the only real dependency on the rest
of the library here, too. It is also prepared to handle surrogate pairs,
and the comments describe how to use it with UTF-8.
Today, we released our latest version 1.3.1 of the IBM Classes for Unicode.
You can get the library source code directly from the CVS server (the path
for the Bid code is icu/source/common/*bidi*), or in the zip file with the
entire source tree for icu at http://www.alphaworks.ibm.com/tech/icu
(->download).
Best regards,
markus
Markus Scherer IBM Cupertino, CA schererm@us.ibm.com
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:54 EDT