gathers a hierarchical tree of entities in a list of open civic data identifiers

gather-ocdid-entities.py

import csv
import json
import pprint
pp = pprint.PrettyPrinter(indent=4)

# setup:
# download the identifiers file for target country directly from the
# open civic identifiers public data repo:
# https://github.com/opencivicdata/ocd-division-ids/tree/master/identifiers
COUNTRY_SOURCE_FILE = 'country-us.csv'

def get_nested_key_from_strings(data, keys):
	if not keys:
		return data
	if keys[0] not in data:
		return False
	return get_nested_key_from_strings(data[keys[0]], keys[1:])

def set_nested_key_from_strings(data, keys, val):
	if len(keys) == 1:
		data[keys[0]] = val
		return
	if keys[0] in data:
		return set_nested_key_from_strings(data[keys[0]], keys[1:], val)
	if keys[0] not in data:
		data[keys[0]] = {}
		set_nested_key_from_strings(data,keys,val)

def open_file():
	with open(COUNTRY_SOURCE_FILE,'r') as f:
		d = list(csv.DictReader(f))
		d = list(set([i['id'] for i in d]))
		d = [i.split('/') for i in d]
	return d

def process(d):
	c = {}
	for i in d:
		if i[0] != 'ocd-division':
			raise Exception('unexpected format')
		path = ['ocd-division']
		count = 1
		while True:
			try:
				lev = i[count].split(':')[0]
			except:
				break
			path += [lev]

			if not get_nested_key_from_strings(c,path):
				set_nested_key_from_strings(c,path,{})
			count += 1
	return c

def view_collection(c):
	pp.pprint(c)

def write_output(c):
	with open('output.json','w') as f:
		json.dump(c,f)

def main():
	d = open_file()
	c = process(d)
	view_collection(c)
	write_output(c)

main()