'''
    Author: Alexander Hanel 
    Date: 20160729
    Purpose: Searches for ngrams in a file and then uses those as an XOR key. 
'''
import sys
from collections import Counter
from itertools import cycle
from itertools import product

def xor_mb(message, key):
    '''Multi-byte XOR of a string message and string key'''
    return''.join(chr(ord(m_byte)^ord(k_byte)) for m_byte,k_byte in zip(message, cycle(key)))


def run():
    data = open(sys.argv[1],'rb').read()
    c = 0 
    # source http://stackoverflow.com/a/25071991
    for n in range(1, len(data)):
        substr_counter = Counter(data[i: i+n] for i in range(len(data) - n))
        key, count = substr_counter.most_common(1)[0]
        if count == 1:
                break
        temp = xor_mb(data, key)
        if "program" in temp:
                print "Length: %s, Count: %s Key: %s" % (len(key), count, key)
                f = open(str(c) + ".bin", "wb")
                fk = open( "key-" + str(c) + ".bin", "wb")
                f.write(temp)
                fk.write(key)
                f.close()
                fk.close()
                c +=1

run()