Use include/1 directives instead of ensure_loaded/1 directives to
avoid some portability issues. Add unicode_data.lgt file to load all
the files using Logtalk as some Prolog systems have buggy implementations
of the include/1 directive.
The predicate unicode_category/2 is faster when checking a code point
for a specific category. When generating code point - category pairs, the
code points are now generated in increasing order.
Third public release featuring a faster implementation of the
unicode_alphabetic/1 predicate (moved from the unicode_alphabetic.pl
file to the new unicode_range_alphabetic.pl file).
Second public release featuring a faster implementation of the
unicode_script/2 predicate (moved from the unicode_scripts.pl
file to the new unicode_range_scripts.pl file).
First public release.