Something like the following may help (encode_monk and decode_monk are only there to save you potential character set problems with your text editor -- don't let them distract you). Cheers.

use strict;
use warnings;

use Encode 'decode_utf8';

my $suspect = decode_utf8( decode_monk('v~C3~A4hicule') );
warn '$suspect='.$suspect;

if ($suspect =~ /^([\p{IsAlnum}]+)$/) {
    warn "MATCHES $1\n";
}
exit( 0 );

##------------------------------------------------------------------+
## WARNING: Using sprintf here is too expensive.  A
##          lookup table may be the best solution.
sub encode_monk {
    my ($txt) = @_;

    join q(),
    map {
        ( $_ >= 0x20 && $_ <= 0x7D )
            ? ( sprintf "%c", $_ )
            : ( sprintf "~%02X", $_ )
        }
    unpack "C*", $txt;
}

##------------------------------------------------------------------+
## WARNING: Consider err policy here.
sub decode_monk {
    my ($code) = @_;

    if (! defined $code or length( $code ) == 0) {
        return;
    }
    $code =~ s/\~([\da-fA-F]{2})/chr( hex( $1 ) )/eg;
    return $code;
}
[download]