Building an undirected graph from depth-first sorted node list in Python

Question

I have an array called nodes that contains a list of Node objects. The list is sorted in depth-first order. I have a function called build_undirected_tree that assigns the parent value and creates a list of children for each Node in the list.

I'm not satisfied with the need for using the idx array to circumvent Python's lack of pass-by-reference on primitive-type arguments and I'm wondering if there's a conceptually simpler way to accomplish the same thing.

class Node(object):
 def __init__(self, child_count):
 self.child_count = child_count
'''
This is a depth-first iterator. The weird `idx` array is used because
integers cannot be pass-by-reference.
'''
def node_iterator(nodes, idx=None, parent=None):
 if idx is None:
 idx = [0]
 node = nodes[idx[0]]
 yield (node, parent)
 for i in range(node.child_count):
 idx[0] += 1
 yield from node_iterator(nodes, idx, node)
def build_undirected_tree(nodes):
 for (node, parent) in node_iterator(nodes):
 node.parent = parent
 node.children = []
 if parent is not None:
 parent.children.append(node)
nodes = [...]
# Populate nodes from a file.
build_undirected_tree(nodes)

Answer 1

I believe you can get a similar result by creating and passing around an iterator on the nodes list:

def node_iterator(nodes, nit=None, parent=None):
 if nit is None:
 nit = iter(nodes)
 node = next(nit)
 yield (node, parent)
 for i in range(node.child_count):
 yield from node_iterator(nodes, nit, parent=node)