my $str = "((A,B,(C,D,(E,F),G),H),(I,J,K))";
our @nodes;
our $re;
$re = qr{
    \(
    (?:
        (?:
            ((?> [^()]+ ))    # Non-parens without backtracking
            (?{  /./ })    # <--- any regex here segfaults
        )+
    |
        (??{ $re })     # Group with matching parens
    )*
    \)
}x;
[download]

#!/usr/bin/perl

use warnings;
use strict;

use lib '.';

use Bio::Tree::Node;

my $str = "  \n   (\n(A:0.333,B,(C,D,(E,  \nF),G),H):0.456,(I,J,K):0.1
+23)    \n   ";

my $current = Bio::Tree::Node->new;
my $last;

our $re = qr{
    \(
    (?{ $current = $current->add_new_child })
    (?:
        (?:
            ((?> [^()]+ ))    # Non-parens without backtracking
            (?{ process_children($current, $^N) })
        )+
    |
        (??{ $re })     # Group with matching parens
    )*
    \)(?{ $current = $current->parent })
}x;

$str =~ $re;

sub process_children {
    my ($node_obj, $str) = @_;
    my @nodes = split '\s*,\s*', $str;
    for my $node (@nodes) {
        my ($tag, $contents) = split ':', $node;
        if ($tag) {
            $node_obj->add_new_child($tag, $contents) 
        }
        else {
            $node_obj->contents($contents);
        }
    }
}
[download]