pyspark flatten multi parent hierarchy into a single column

Question

I have a node - > parent_node dataset in the following format. I need to generate a node_map so that any query trying to select the \'node\' or \'any of it parents\' results