1) Pass a reference of the array in object2 to object1, so it can add it directly.

2) Have a method in object2 that allows you to add to this array; object1 repeatedly calls this function for each object3 it adds

Method 2 is much much much better from the object oriented stand point as it hides the details of object2 from other objects. Say in the future, you decide to replace that array in object2 with a hash for speeding up another part of the process. With Method 2, you only need to modify how to add things via that function, while method 1 requires modifying object1's code to handle the change. Not the best design here.

All you need to do is
(a) store the array of obj3's as an array reference.
(b) dereference to access them.

Sample code : in obj2_maker method:

my @obj3s;
while (CONDITION) {
   push @obj3s, Object3->new();
}
my $obj2 = Object2->new(-children=>\@obj3s); # edited ... had the drea
+ded "=" without the ">"
[download]

Note you'll have to write an appropriate constructor for Obj2.

In obj2's children method:

sub children {
    my $self = shift;
    @_ ? $self->{_children} = shift :
    $self->{_children};
}
[download]

That method will return the anonymous array of children (if nothing is passed to it), or will set the children of the object to the array that is passed to it.

Access the array by de-referencing. @{ $obj2->children() }

This is a *very basic* framework. Don't take it as gospel. Just remember 'references' =)

HTH

perl -e 'print "How sweet does a rose smell? "; chomp $n = <STDIN>; $r
+ose = "smells sweet to degree $n"; *other_name = *rose; print "$other
+_name\n"'
[download]