Ah yes! Good catch graff. The big problem, as you noted was that I was parsing the whole text. Simply adding another last after the chop(...) fixes that. The other was splitting on one space -- normalizing whitespace like you've done fixes that. Lastly, calling return text all those times is really bad like you've noted, so i just assigned it to a var (and used as_is). Now, while i still like yours better, here's an updated version of mine :)

while ( my $token = $p->get_token ) {
    if ($token->is_text) {
        my $text = $token->as_is;
        $text =~ s/\s+/ /g;
        if (length($text) + $total <= 200) {
            $doc2 .= $text;
            $total += length($text);
        }
        else {
            for (split / /, $text) {
                if ($total + length($_) <= 200) {
                    $doc2 .= $_ . ' ';
                    $total += length($_) + 1;
                }
                else {
                    last;
                }
            }
            chop($doc2) if $doc2 =~ /\s$/;
            last;
        }
    }
    else {
        $doc2 .= $token->as_is;
    }
}
[download]

Thanks for the feedback!

Minor fix to yours:

$doc2 .= substr( $tkntext, 0, rindex( $tkntext, ' ', $maxlen );
[download]

Should be:

$doc2 .= substr( $tkntext, 0, rindex( $tkntext, ' ', $maxlen ) );
[download]

it was missing a bracket. =)

--
"To err is human, but to really foul things up you need a computer." --Paul Ehrlich