code sketch: grab a CSV file with a column called "OTV SKU#" and split the characters in all the values into a character tree

skutree.py

from CsvFileModel import CsvFileModel  # available from https://gist.github.com/devvyn/759f08e3d83e4cf8f7ef
import json
import yaml
import collections


class StringTree(object):
	def __init__(self):
		c = CsvFileModel('csp.csv')
		skus = list(x.strip() for x in c['OTV SKU#'])
		print("SKU count: %i" % len(skus))
		charset = set(''.join(skus))
		print("Character set: %s" % charset)
		d = dict()
		j = d
		for sku in skus:
			for position, char in enumerate(sku):
				j = j.setdefault(char, dict())
			j = d
		print(d.keys())
		print(yaml.dump(d))
		self.skus = skus
		self.data = c
		self.tree = d
		self.charset = charset

if __name__ == '__main__':
	d = StringTree()
	data = d.tree
	with open('sku_codes.yml', 'w') as outfile:
		yaml.dump(data, outfile, default_flow_style=False)

This is pretty old, but for what it's worth, I detect some code smell. Hard coded file name in the constructor is probably the worst.

Hard to read. Looks like j is being reassigned multiple times without read.