Well actually.....it doesn't break the counting code. Only the first instance of sub foo { is removed courtesy of the ^ The \n are gone by the time the sub bit is removed and even if they were not there is no /m. Although the sub removed is sub r the result is correct. To satisfy the spec sub r shoud be contained within sub n as the spec was for a single sub to do the task not a whole app! The code says 82 BTW :-)

use strict;

my $code = <<'CODE';
 
sub r{$_[0]eq reverse$_[0]}
sub n{foreach(1..$_[0]){$a{$_}=$_*$_ if r($_)&&r($_*$_)};\%a}  
}

CODE

$code =~ s/^\s+|\s+$//gm;
$code =~ s/\n//g;
$code =~ s/}$//;
$code =~ s/^sub\s+(\w+)\s*{//;

printf "Length of sub '$1' => %d chars", length $code;

print "\n\n$code\n";
__END__
[download]

cheers

tachyon

s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print