How to Create a JSON Tree from a Tabulated Hierarchy in Python

I'm sure there's an elegant way of doing this in Python (or at a push, Javascript), but for the life of me I can't see it...

I have a CSV file of the form:

ID, Name, Description
A, A-name,
A100, A100-name, A100-desc
A110, A110-name, A110-desc
A111, A111-name, A111-desc
A112, A112-name, A112-desc
A113, A113-name, A113-desc
A120, A120-name, A120-desc
A131, A131-name, A131-desc
A200, A200-name, A200-desc
B, B-name,
B100, B100-name, B100-desc
B130, B130-name, B130-desc
B开发者_高级运维131, B131-name, B131-desc
B140, B140-name, B140-desc

and I want to generate a hierarchical JSON structure so I can visualise the data in theJIT.

var json = {  
  "id": "aUniqueIdentifier",  
  "name": "usually a nodes name",  
  "data": {  
    "some key": "some value",  
    "some other key": "some other value"  
   },  
  "children": [ *other nodes or empty* ]  
}; 

My plan was to map ID to id, Name to name, Description to data.desc, and organise the hierarchy so that:

• Root is parent of A and B
• A is parent of A100 and A200
• A100 is parent of A110 and A120
• A110 is parent of A111, A112 and A113
• B is parent of B100
• B100 is parent of B130 and B140
• B130 is parent of B131

There is also a pathological case in the otherwise regular ordering by ID, where A100 is parent of A131 (the expected A130 is not present).

I was hoping to find an elegant Python solution to this, but it's defeating me at the moment, even ignoring the pathological case...

This does it...

import csv
import json

class Node(dict):
    def __init__(self, (nid, name, ndescr)):
        dict.__init__(self)
        self['id'] = nid
        self['name'] = name.lstrip() # you have badly formed csv....
        self['description'] = ndescr.lstrip()
        self['children'] = []

    def add_node(self, node):
        for child in self['children']:
            if child.is_parent(node):
                child.add_node(node)
                break
        else:
            self['children'].append(node)

    def is_parent(self, node):
        if len(self['id']) == 4 and self['id'][-1] == '0':
            return node['id'].startswith(self['id'][:-1])
        return node['id'].startswith(self['id'])

class RootNode(Node):
    def __init__(self):
        Node.__init__(self, ('Root', '', ''))

    def is_parent(self, node):
        return True

def pretty_print(node, i=0):
    print '%sID=%s NAME=%s %s' % ('\t' * i, node['id'], node['name'], node['description'])
    for child in node['children']:
        pretty_print(child, i + 1)

def main():
    with open('input.csv') as f:
        f.readline() # Skip first line
        root = RootNode()
        for node in map(Node, csv.reader(f)):
            root.add_node(node)

    pretty_print(root)
    print json.dumps(root)

if __name__ == '__main__':
    main()