The reason BrowserUk passed the node-printing code as an anonymous subroutine was to separate the traversal logic from the node-processing logic. With this separation, we can use the same traversal logic for other purposes, and we can use the same node-processing logic with other traversals.

Consider the following nested-array-style tree and logic to traverse it:

my $tree = [ 1
           , [ 2
             , 3
             , [ 4
               , [ 5 ]
               , 6     ] ]
           , 7             ] ;

sub traverse {
    my ($nodefn, $tree, $depth) = @_;
    $depth ||= 0;
    $depth++;
    for (@$tree) {
        ref() ? traverse( $nodefn, $_, $depth )
              : $nodefn->( $_, $depth );
    }
}
[download]

We can combine this traversal logic with task-specific node-processing logic. For example, to count the nodes in the tree, we might use logic like this:

sub count_nodes {
    my $count;
    traverse( sub { $count++ }, @_ );
    $count;
}

print count_nodes($tree), $/;
# 7
[download]

For other jobs we can easily "plug in" other node-processing code. Here is code to print a single indented node:

sub print_node {
    my ($leaf, $depth) = @_;
    print "  " x ($depth - 1), "* $leaf", $/;
}
[download]

We just plug it into traverse to print the tree's outline:

traverse( \&print_node, $tree );
# * 1
#   * 2
#   * 3
#     * 4
#       * 5
#     * 6
# * 7
[download]

Going further, we can plug in new traversal logic, too. Here is logic to do a breadth-first traversal – all the leaves at each depth are processed before going deeper.

sub bfs_traverse {
    my ($nodefn, $tree, $depth) = @_;
    $depth ||= 0;
    $depth++;
    my @leaves   = grep !ref(), @$tree;
    my @branches = grep  ref(), @$tree;
    $nodefn->( $_, $depth ) for @leaves;
    bfs_traverse( $nodefn, [map @$_, @branches], $depth )
        if @branches;
}
[download]

Now we can use this traversal logic with our existing print_node logic to outline the tree in breadth-first order:

bfs_traverse( \&print_node, $tree );
# * 1
# * 7
#   * 2
#   * 3
#     * 4
#     * 6
#       * 5
[download]

By this point, I hope you can see the answer to your second question: by keeping traversal and node-processing concerns separated and by combining them through a well-defined interface (here, function passing), we can mix and match logic. If we want a breadth-first outline, we don't need to write a breadth-first outline function; we can just combine a breadth-first traversal with a node-printing processor.

Cheers,
Tom