Mercurial > ppgen
annotate ppgen.py @ 1:00ed7df30fe4
Checking for 8k entries now. Comment improvements.
| author | Bernhard Reiter <bernhard@intevation.de> |
|---|---|
| date | Mon, 22 Feb 2016 16:14:24 +0100 |
| parents | 7558ecd1cbf1 |
| children | a099246680ae |
| rev | line source |
|---|---|
| 0 | 1 #!/usr/bin/env python3 |
| 2 """Create a passphrase from a few random words. DRAFT | |
| 3 | |
| 4 Relies on the entropy of python's | |
| 5 random.SystemRandom class | |
| 6 which (according to the documentation) calls os.urandom() | |
| 7 which (according to the documentation) calls the operating system | |
| 8 specific randomness source which "should be unpredictable | |
| 9 enough for cryptographic applications" | |
| 10 | |
| 11 Requires: | |
| 12 * Python v>=3.2 | |
| 13 * a dictionary, Ding's trans-de-en by default. | |
| 14 E.g. on a Debian/Ubuntu system in package "trans-de-en". | |
| 15 or from http://ftp.tu-chemnitz.de/pub/Local/urz/ding/de-en/ | |
| 16 | |
|
1
00ed7df30fe4
Checking for 8k entries now. Comment improvements.
Bernhard Reiter <bernhard@intevation.de>
parents:
0
diff
changeset
|
17 Uses a hardcoded filepath and language. |
| 0 | 18 Search for **customize** below to change it. |
| 19 | |
| 20 Copyright 2016 by Intevation GmbH. | |
|
1
00ed7df30fe4
Checking for 8k entries now. Comment improvements.
Bernhard Reiter <bernhard@intevation.de>
parents:
0
diff
changeset
|
21 Author: Bernhard E. Reiter <bernhard@intevation.de> |
| 0 | 22 |
| 23 This file is Free Software under the Apache 2.0 license and thus | |
| 24 comes without any warranty (to extend permissible under applicable law). | |
| 25 """ | |
| 26 | |
| 27 import math | |
| 28 import re | |
| 29 import sys | |
| 30 | |
| 31 from random import SystemRandom | |
| 32 _srandom = SystemRandom() | |
| 33 | |
| 34 tainted = False # to be set if we find a hint that the passphrase may be weak | |
| 35 | |
| 36 def buildDictionary(): | |
| 37 """Build up a dictionary of unique words, calculate stats.""" | |
| 38 global tainted | |
| 39 d = [] | |
| 40 | |
| 41 # dictionary for testing | |
| 42 #d = ["abc", "aBc", "cde", "efg", "hij", "blubber", "jikf", "zug", "lmf", "opq"] | |
| 43 | |
| 44 # Using the dictionary from Ding **customize** | |
| 45 d = readDingDict(filename="/usr/share/trans/de-en", useLeft=True) | |
| 46 | |
| 47 ## for debugging purpuses, dump dictionary | |
| 48 #dumpfilename = "ddump.txt" | |
| 49 #print("Writing out {}.".format(dumpfilename)) | |
| 50 #with open(dumpfilename, "w") as f: | |
| 51 # for i in d: | |
| 52 # f.write("{}\n".format(i)) | |
| 53 | |
| 54 # Print some stats on the dictionary to be used | |
| 55 dl = len(d) | |
|
1
00ed7df30fe4
Checking for 8k entries now. Comment improvements.
Bernhard Reiter <bernhard@intevation.de>
parents:
0
diff
changeset
|
56 print("Found {:d} dictionary entries.".format(dl)) |
|
00ed7df30fe4
Checking for 8k entries now. Comment improvements.
Bernhard Reiter <bernhard@intevation.de>
parents:
0
diff
changeset
|
57 if dl < 8000: |
|
00ed7df30fe4
Checking for 8k entries now. Comment improvements.
Bernhard Reiter <bernhard@intevation.de>
parents:
0
diff
changeset
|
58 print("!Your dictionary is below 8k entries, that is quite small!") |
| 0 | 59 tainted = True |
| 60 | |
| 61 print("|= Number of words |= possibilities |") | |
| 62 for i in range(1,5): | |
| 63 print("| {:2d} | 2^{:4.1f} |".format( | |
| 64 i, math.log(dl**i,2))) | |
| 65 return d | |
| 66 | |
| 67 | |
| 68 def readDingDict(filename = "/usr/share/trans/de-en", useLeft=False): | |
| 69 """Read dictionary with unique words from file in Ding format. | |
| 70 | |
| 71 useLeft: Boolean to control which language to use | |
| 72 | |
| 73 TODO: add option to use both languages for people that speak them both? | |
| 74 """ | |
| 75 | |
| 76 dset = set() #using the datatype 'set' to aviod duplicates | |
| 77 | |
| 78 splitter = re.compile(r"""\ \|\ # first pattern ' | ' | |
| 79 |;\ # second pattern '; ' | |
| 80 |(?<=\S)/(?=\S) # 3.: '\' surrounded by chars | |
| 81 |\s+ # by whitespace | |
| 82 """,re.VERBOSE) | |
| 83 | |
| 84 print("Reading entries from {}.".format(filename), end='') | |
| 85 counter = 0 # for progress or stopping early | |
| 86 with open(filename, "r") as f: | |
| 87 for line in f: | |
| 88 if line[0] == '#': continue | |
| 89 | |
| 90 # languages are separated by " :: " | |
| 91 p = line.partition(" :: ") | |
| 92 languageEntry = p[0] if useLeft else p[2] | |
| 93 | |
| 94 for word in splitter.split(languageEntry): | |
| 95 word = word.strip('(",.)\'!:;').rstrip('/') | |
| 96 if len(word) > 2 and not word[0] in '[{/': | |
| 97 dset.add(word) | |
| 98 | |
| 99 #TODO: check for very common words and remove them? | |
| 100 | |
| 101 counter += 1 | |
| 102 ## stop early when debugging | |
| 103 #if counter > 10: break | |
| 104 if not counter % 10000: | |
| 105 print('.', end='') | |
| 106 sys.stdout.flush() | |
| 107 print() | |
| 108 | |
| 109 return list(dset) | |
| 110 | |
| 111 def main(): | |
| 112 global tainted | |
| 113 dictionary = buildDictionary() | |
| 114 | |
| 115 howMany = 4 | |
| 116 | |
| 117 # use a dictionary with lower case words for a simple check if | |
| 118 # our random source is okay | |
| 119 print("\nGenerated passphrase with {} randomly selected words:\n".format( | |
| 120 howMany)) | |
| 121 print(" ", end='') | |
| 122 words = {} | |
| 123 for x in range(howMany): | |
| 124 word = _srandom.choice(dictionary) | |
| 125 words[word.lower]= True | |
| 126 print(word, end='\n ') | |
| 127 print("\n") | |
| 128 | |
| 129 if len(words) < howMany: | |
| 130 print("! Your random generator is weak") | |
| 131 print("! or you are being very lucky.") | |
| 132 tainted = True | |
| 133 | |
| 134 if tainted: | |
| 135 print("!!! Don't use the resulting passphrase !!!") | |
| 136 | |
| 137 if __name__ == "__main__": | |
| 138 main() |