From bdb9269756709dc830cc943d7d28e8340ca79049 Mon Sep 17 00:00:00 2001 From: Neil Smith Date: Mon, 23 Jun 2014 22:29:50 +0100 Subject: [PATCH] Merged updated cipherbreak.py from development --- cipherbreak.py | 145 ++++++++++++++++++++----------------------------- 1 file changed, 58 insertions(+), 87 deletions(-) diff --git a/cipherbreak.py b/cipherbreak.py index 5e5956a..e89407c 100644 --- a/cipherbreak.py +++ b/cipherbreak.py @@ -2,7 +2,7 @@ import string import collections import norms import logging -from itertools import zip_longest, cycle, permutations +from itertools import zip_longest, cycle, permutations, starmap from segment import segment from multiprocessing import Pool from math import log10 @@ -183,31 +183,6 @@ def keyword_break_worker(message, keyword, wrap_alphabet, fitness): wrap_alphabet, fit, sanitise(plaintext)[:50])) return (keyword, wrap_alphabet), fit -def scytale_break(message, fitness=Pbigrams): - """Breaks a Scytale cipher - - >>> scytale_break('tfeulchtrtteehwahsdehneoifeayfsondmwpltmaoalhikotoere' \ - 'dcweatehiplwxsnhooacgorrcrcraotohsgullasenylrendaianeplscdriioto' \ - 'aek') # doctest: +ELLIPSIS - (6, -281.276219108...) - """ - best_key = 0 - best_fit = float("-inf") - for key in range(1, 20): - if len(message) % key == 0: - plaintext = scytale_decipher(message, key) - fit = fitness(sanitise(plaintext)) - logger.debug('Scytale break attempt using key {0} gives fit of ' - '{1} and decrypt starting: {2}'.format(key, - fit, sanitise(plaintext)[:50])) - if fit > best_fit: - best_fit = fit - best_key = key - logger.info('Scytale break best fit with key {0} gives fit of {1} and ' - 'decrypt starting: {2}'.format(best_key, best_fit, - sanitise(scytale_decipher(message, best_key))[:50])) - return best_key, best_fit - def column_transposition_break_mp(message, translist=transpositions, fitness=Pbigrams, chunksize=500): @@ -265,33 +240,44 @@ def column_transposition_break_worker(message, transposition, return (transposition, fillcolumnwise, emptycolumnwise), fit +def scytale_break_mp(message, max_key_length=20, + fitness=Pbigrams, chunksize=500): + """Breaks a scytale cipher using a range of lengths and + n-gram frequency analysis -def vigenere_keyword_break(message, wordlist=keywords, fitness=Pletters): - """Breaks a vigenere cipher using a dictionary and - frequency analysis - - >>> vigenere_keyword_break(vigenere_encipher(sanitise('this is a test ' \ - 'message for the vigenere decipherment'), 'cat'), \ - wordlist=['cat', 'elephant', 'kangaroo']) # doctest: +ELLIPSIS - ('cat', -52.947271216...) + >>> scytale_break_mp(scytale_encipher(sanitise( \ + "It is a truth universally acknowledged, that a single man in \ + possession of a good fortune, must be in want of a wife. However \ + little known the feelings or views of such a man may be on his \ + first entering a neighbourhood, this truth is so well fixed in the \ + minds of the surrounding families, that he is considered the \ + rightful property of some one or other of their daughters."), \ + 5)) # doctest: +ELLIPSIS + (5, -709.4646722...) + >>> scytale_break_mp(scytale_encipher(sanitise( \ + "It is a truth universally acknowledged, that a single man in \ + possession of a good fortune, must be in want of a wife. However \ + little known the feelings or views of such a man may be on his \ + first entering a neighbourhood, this truth is so well fixed in the \ + minds of the surrounding families, that he is considered the \ + rightful property of some one or other of their daughters."), \ + 5), \ + fitness=Ptrigrams) # doctest: +ELLIPSIS + (5, -997.0129085...) """ - best_keyword = '' - best_fit = float("-inf") - for keyword in wordlist: - plaintext = vigenere_decipher(message, keyword) - fit = fitness(plaintext) - logger.debug('Vigenere break attempt using key {0} ' - 'gives fit of {1} and decrypt starting: {2}'.format( - keyword, fit, - sanitise(plaintext)[:50])) - if fit > best_fit: - best_fit = fit - best_keyword = keyword - logger.info('Vigenere break best fit with key {0} gives fit ' - 'of {1} and decrypt starting: {2}'.format(best_keyword, - best_fit, sanitise( - vigenere_decipher(message, best_keyword))[:50])) - return best_keyword, best_fit + with Pool() as pool: + helper_args = [(message, trans, False, True, fitness) + for trans in + [[col for col in range(math.ceil(len(message)/rows))] + for rows in range(1,max_key_length+1)]] + # Gotcha: the helper function here needs to be defined at the top level + # (limitation of Pool.starmap) + breaks = pool.starmap(column_transposition_break_worker, + helper_args, chunksize) + best = max(breaks, key=lambda k: k[1]) + return math.trunc(len(message) / len(best[0][0])), best[1] +scytale_break = scytale_break_mp + def vigenere_keyword_break_mp(message, wordlist=keywords, fitness=Pletters, chunksize=500): @@ -310,6 +296,7 @@ def vigenere_keyword_break_mp(message, wordlist=keywords, fitness=Pletters, # (limitation of Pool.starmap) breaks = pool.starmap(vigenere_keyword_break_worker, helper_args, chunksize) return max(breaks, key=lambda k: k[1]) +vigenere_keyword_break = vigenere_keyword_break_mp def vigenere_keyword_break_worker(message, keyword, fitness): plaintext = vigenere_decipher(message, keyword) @@ -321,7 +308,7 @@ def vigenere_keyword_break_worker(message, keyword, fitness): -def vigenere_frequency_break(message, fitness=Pletters): +def vigenere_frequency_break(message, max_key_length=20, fitness=Pletters): """Breaks a Vigenere cipher with frequency analysis >>> vigenere_frequency_break(vigenere_encipher(sanitise("It is time to " \ @@ -333,26 +320,19 @@ def vigenere_frequency_break(message, fitness=Pletters): "sure"), 'florence')) # doctest: +ELLIPSIS ('florence', -307.5473096791...) """ - best_fit = float("-inf") - best_key = '' - sanitised_message = sanitise(message) - for trial_length in range(1, 20): - splits = every_nth(sanitised_message, trial_length) + def worker(message, key_length, fitness): + splits = every_nth(sanitised_message, key_length) key = ''.join([chr(caesar_break(s)[0] + ord('a')) for s in splits]) - plaintext = vigenere_decipher(sanitised_message, key) + plaintext = vigenere_decipher(message, key) fit = fitness(plaintext) - logger.debug('Vigenere key length of {0} ({1}) gives fit of {2}'. - format(trial_length, key, fit)) - if fit > best_fit: - best_fit = fit - best_key = key - logger.info('Vigenere break best fit with key {0} gives fit ' - 'of {1} and decrypt starting: {2}'.format(best_key, - best_fit, sanitise( - vigenere_decipher(message, best_key))[:50])) - return best_key, best_fit - -def beaufort_frequency_break(message, fitness=Pletters): + return key, fit + sanitised_message = sanitise(message) + results = starmap(worker, [(sanitised_message, i, fitness) + for i in range(1, max_key_length+1)]) + return max(results, key=lambda k: k[1]) + + +def beaufort_frequency_break(message, max_key_length=20, fitness=Pletters): """Breaks a Beaufort cipher with frequency analysis >>> beaufort_frequency_break(beaufort_encipher(sanitise("It is time to " \ @@ -364,25 +344,16 @@ def beaufort_frequency_break(message, fitness=Pletters): "that he is sure"), 'florence')) # doctest: +ELLIPSIS ('florence', -307.5473096791...) """ - best_fit = float("-inf") - best_key = '' - sanitised_message = sanitise(message) - for trial_length in range(1, 20): - splits = every_nth(sanitised_message, trial_length) + def worker(message, key_length, fitness): + splits = every_nth(sanitised_message, key_length) key = ''.join([chr(-caesar_break(s)[0] % 26 + ord('a')) for s in splits]) - plaintext = beaufort_decipher(sanitised_message, key) + plaintext = beaufort_decipher(message, key) fit = fitness(plaintext) - logger.debug('Beaufort key length of {0} ({1}) gives fit of {2}'. - format(trial_length, key, fit)) - if fit > best_fit: - best_fit = fit - best_key = key - logger.info('Beaufort break best fit with key {0} gives fit ' - 'of {1} and decrypt starting: {2}'.format(best_key, - best_fit, sanitise( - beaufort_decipher(message, best_key))[:50])) - return best_key, best_fit - + return key, fit + sanitised_message = sanitise(message) + results = starmap(worker, [(sanitised_message, i, fitness) + for i in range(1, max_key_length+1)]) + return max(results, key=lambda k: k[1]) def plot_frequency_histogram(freqs, sort_key=None): -- 2.34.1