Regexp is definitely not the place to post character sets, definitely ... except that it is :) um, I changed my mind

well actually UTF8 definitely belongs nowhere , characters ranges aren't UTF8, this module doesn't deal with UTF8 ... sure the OP wrote

 
 $char = "...";  # some UTF8 string
[download]

So after some more research

Regexp::Cherokee - Regular Expressions Support for Cherokee Script. 
Regexp::Ethiopic - Regular Expressions Support for Ethiopic Script.
Regexp::CharClasses - Provide character classes
Regexp::EN::NumVerbage - Regex pattern to match English number verbage
+ in text 
Regexp::CharClasses - Provide character classes
[download]

I propose the one new namespace to rule them all Regexp::CharProps - User Defined Character Properties like \p{InKona}

So for the OP Regexp::CharProps::Thai - \p(InThaiDigit} and other thai language character properties definitions

So maybe

#!/usr/bin/perl --

=head1 NAME

Regexp::CharProps - User Defined Character Properties 

=head1 SYNOPSIS

    use Regexp::CharProps qw/ Thai /; ## like use Regexp::CharProps::T
+hai qw/ :all /;
                                      ## imports all exports like sub 
+InThaiPunct...
    
    print "\$_ has got Thai" if m{
        \p{InThai}
        |\p{InThaiCons}
        |\p{InThaiHCons}
        |\p{InThaiMCons}
        |\p{InThaiLCons}
        |\p{InThaiVowel}
        |\p{InThaiPreVowel}
        |\p{InThaiPostVowel}
        |\p{InThaiCompVowel}
        |\p{InThaiDigit}
        |\p{InThaiTone}
        |\p{InThaiPunct}
    }x;
    
    use Regexp::CharProps::Thai qw/ InThaiPunct /; ## not import :all 
+just sub InThaiPunct
    print "got Thai punctuation\n" if m/\p{InThaiPunct}/;

=cut


package Regexp::CharProps;
sub import {
    my( $class, @modsforall ) = @_;
    
    return if not @modsforall;
    
    my $target  = scalar caller;
    require Import::Into;
    for my $mod( @modsforall ){
        my $package = $class."::".$mod;
        $package ->import::into( $target , ':all' );
    }
}
1;
[download]

with accompanying EXPORT_TAGS

Read more... (6 kB)

"but he meant some unicode string"

Yes. It definitely wouldn't work on an upper-ascii-type encoding such as Thai originally began with, without some form of encoding/decoding going on. I guess I put "UTF8" because that is what gets used most with Thai, and what I knew would work having developed strictly with that. I presume any Unicode type should work equally well, though I don't claim to be an expert on Unicode.

In your code example:

print "\$_ has got Thai" if m{
        \p{InThai}
        |\p{InThaiCons}
        |\p{InThaiHCons}
        |\p{InThaiMCons}
        |\p{InThaiLCons}
        |\p{InThaiVowel}
        |\p{InThaiPreVowel}
        |\p{InThaiPostVowel}
        |\p{InThaiCompVowel}
        |\p{InThaiDigit}
        |\p{InThaiTone}
        |\p{InThaiPunct}
    }x;
[download]

...only the first item in the OR'ed list should ever see action. All of the subsequent categories are already "InThai", and the "InThai" token already comes standard with Perl, AFAIK (see pg. 172 of "Programming Perl, Third Edition"), so that code would do little to test additional functionality. If the first line (\p{InThai}) failed, none of the others should succeed either.

NOTE: I've updated my list to reflect your proposed name, but I've adapted it slightly to one that seems a better fit to me.

Blessings,

~Polyglot~