clone a string with a percentage of similarity

test_clone.py

# -*- coding: utf-8 -*-
from datetime import datetime, timedelta
import random
import unittest
import difflib

def ratio(original, changed):
	changes = 0
	u = zip(original, changed)
	
	for i,j in u:
		if i != j:
			changes+=1

	ratio = round(1.0 - (changes/len(original)), 2)
	return ratio

def clone_date(dt, percentage, comparison='='):
	fecha_original_str = dt.strftime("%y%m%d")
	fecha_str = None

	changed = False
	intents = 0
	max_intents = 10000

	if comparison == '=' and percentage == 1:
		return fecha_original_str

	while not changed and intents < max_intents:
		_dt = dt + timedelta(days=random.randrange(0, 3000))

		fecha_str = _dt.strftime("%y%m%d")

		print("Trying ..." + str(intents))

		print("Ratio : " + str(ratio(fecha_original_str, fecha_str)))

		if comparison == '=':
			if ratio(fecha_original_str, fecha_str) == percentage:
				changed = True
				break

		if comparison == '>':
			if ratio(fecha_original_str, fecha_str) > percentage:
				changed = True
				break

		if comparison == '<':
			if ratio(fecha_original_str, fecha_str) < percentage:
				changed = True
				break

		intents+=1

	if intents == max_intents:
		raise Exception('Max intents exceeded. Remenber that some percentages of similarity are imposible to get.')
		
	return fecha_str

def clone(string, percentage, only_numbers=False, start=0):
	length = len(string)
	lstring = list(string)
	res = round(length * (round(1 - percentage, 2)))
	words_to_change = int(res)

	index_characters_changed = []

	for i in range(0, words_to_change):
		
		found = False

		while not found:
			random_index = random.randrange(start, length)
			if random_index not in index_characters_changed:
				index_characters_changed.append(random_index)
				found = True
			if lstring[random_index] == '\n':
				found = False
			if lstring[random_index] == ' ':
				found = False

		original_character = lstring[random_index]

		if only_numbers:
			w = '1234567890'
		else:
			w = 'abcdefghijklmnopqrstuvwxyz0123456789'.upper()

		changed = False
		while not changed:
			lstring[random_index] = random.choice(w)
			if lstring[random_index] != original_character:
				changed = True

	return "".join(lstring)

class TestClone(unittest.TestCase):

	def test_clone_date(self):
		dt = datetime.now() + timedelta(days=random.randrange(0,3))
		fecha_original_str = dt.strftime("%y%m%d")
		percentage = 0.95
		comparison = '<'
		fecha_str = clone_date(dt, percentage, comparison=comparison)
		_ratio = ratio(fecha_original_str, fecha_str)

		print("========== TEST CLONE DATE ==========")
		print(fecha_original_str + " === " + fecha_str + " , ratio = " + str(_ratio))
		print("=====================================")

		if comparison == '=':
			self.assertTrue(_ratio == percentage)

		if comparison == '<':
			self.assertTrue(_ratio < percentage)

		if comparison == '>':
			self.assertTrue(_ratio > percentage)

	def test_clone(self):
		original_word = """
:50K:/191000695986399
IBM D5L PERU
IBMDELIERU S.A.C.
AV. PROLOCGACIHN JAVIER PRADO EETE
LIMA 10 PERU
"""
		percentage = 0.95
		#changed_word = clone(original_word, percentage, start=22)
		changed_word = """
:50K:/191000695986399
IBM D5L PERU
IBMDELIERU S.O.D.
AV. PROLOCKACIHN JAVIER PRADO EETE
LIM2 1D PERU
"""

		print("ORIGINAL : " + original_word)
		print("CHANGED  : " + changed_word)

		changes = 0
		u = zip(original_word, changed_word)
		
		for i,j in u:
			if i != j:
				changes+=1

		print("LEN : " + str(len(original_word)))
		print("CHANGES : " + str(changes))

		_ratio = round(1.0 - (changes/len(original_word)),2)

		print("RATIO : " + str(_ratio))

		self.assertEqual(percentage, _ratio)

if __name__ == '__main__':
		unittest.main()