convert_whole_foods_categories.py

#!/usr/bin/env python

import csv
import hashlib
import re
import sys

def slugify(s):
	s = s.lower()
	for c in [' ', '-', '.', '/']:
		s = s.replace(c, '_')
	s = re.sub('\W', '', s)
	s = s.replace('_', ' ')
	s = re.sub('\s+', ' ', s)
	s = s.strip()
	s = s.replace(' ', '-')
	return s

writer = csv.writer(sys.stdout)

chain = sys.argv[2]

with open(sys.argv[1], 'rb') as csvfile:
	reader = csv.reader(csvfile)

	seen = set([])
	for row in reader:
		if not slugify(row[3]) in seen:
			writer.writerow([chain, '', slugify(row[3]), row[3]])
			seen.add(slugify(row[3]))
		if not slugify(row[3])+'/'+slugify(row[4]) in seen:
			writer.writerow([chain, slugify(row[3]), slugify(row[3])+'/'+slugify(row[4]), row[4]])
			seen.add(slugify(row[3])+'/'+slugify(row[4]))