Md5.py
Revision as of 13:57, 29 February 2008 by PeterHarding (talk | contribs)
Documentation
http://www.performiq.com.au/kb/images/Md5py.pdf
Script
#!/usr/bin/env python
"""A sample implementation of MD5 in pure Python.
This is an implementation of the MD5 hash function, as specified by
RFC 1321, in pure Python. It was implemented using Bruce Schneier's
excellent book "Applied Cryptography", 2nd ed., 1996.
Surely this is not meant to compete with the existing implementation
of the Python standard library (written in C). Rather, it should be
seen as a Python complement that is more readable than C and can be
used more conveniently for learning and experimenting purposes in
the field of cryptography.
This module tries very hard to follow the API of the existing Python
standard library's "md5" module, but although it seems to work fine,
it has not been extensively tested! (But note that there is a test
module, test_md5py.py, that compares this Python implementation with
the C one of the Python standard library.
BEWARE: this comes with no guarantee whatsoever about fitness and/or
other properties! Specifically, do not use this in any production
code! License is Python License!
Special thanks to Aurelian Coman who fixed some nasty bugs!
Dinu C. Gherman
"""
__date__ = '2001-10-1'
__version__ = 0.9
import struct, string, copy
# ======================================================================
# Bit-Manipulation helpers
#
# _long2bytes() was contributed by Barry Warsaw
# and is reused here with tiny modifications.
# ======================================================================
def _long2bytes(n, blocksize=0):
"""Convert a long integer to a byte string.
If optional blocksize is given and greater than zero, pad the front
of the byte string with binary zeros so that the length is a multiple
of blocksize.
"""
# After much testing, this algorithm was deemed to be the fastest.
s = ''
pack = struct.pack
while n > 0:
### CHANGED FROM '>I' TO '<I'. (DCG)
s = pack('<I', n & 0xffffffffL) + s
### --------------------------
n = n >> 32
# Strip off leading zeros.
for i in range(len(s)):
if s[i] <> '\000':
break
else:
# Only happens when n == 0.
s = '\000'
i = 0
s = s[i:]
# Add back some pad bytes. This could be done more efficiently
# w.r.t. the de-padding being done above, but sigh...
if blocksize > 0 and len(s) % blocksize:
s = (blocksize - len(s) % blocksize) * '\000' + s
return s
def _bytelist2long(list):
"Transform a list of characters into a list of longs."
imax = len(list)/4
hl = [0L] * imax
j = 0
i = 0
while i < imax:
b0 = long(ord(list[j]))
b1 = (long(ord(list[j+1]))) << 8
b2 = (long(ord(list[j+2]))) << 16
b3 = (long(ord(list[j+3]))) << 24
hl[i] = b0 | b1 |b2 | b3
i = i+1
j = j+4
return hl
def _rotateLeft(x, n):
"Rotate x (32 bit) left n bits circularly."
return (x << n) | (x >> (32-n))
# ======================================================================
# The real MD5 meat...
#
# Implemented after "Applied Cryptography", 2nd ed., 1996,
# pp. 436-441 by Bruce Schneier.
# ======================================================================
# F, G, H and I are basic MD5 functions.
def F(x, y, z):
return (x & y) | ((~x) & z)
def G(x, y, z):
return (x & z) | (y & (~z))
def H(x, y, z):
return x ^ y ^ z
def I(x, y, z):
return y ^ (x | (~z))
def XX(func, a, b, c, d, x, s, ac):
"""Wrapper for call distribution to functions F, G, H and I.
This replaces functions FF, GG, HH and II from "Appl. Crypto.
Rotation is separate from addition to prevent recomputation
(now summed-up in one function).
"""
res = 0L
res = res + a + func(b, c, d)
res = res + x
res = res + ac
res = res & 0xffffffffL
res = _rotateLeft(res, s)
res = res & 0xffffffffL
res = res + b
return res & 0xffffffffL
class MD5:
"An implementation of the MD5 hash function in pure Python."
def __init__(self):
"Initialisation."
# Initial 128 bit message digest (4 times 32 bit).
self.A = 0L
self.B = 0L
self.C = 0L
self.D = 0L
# Initial message length in bits(!).
self.length = 0L
self.count = [0, 0]
# Initial empty message as a sequence of bytes (8 bit characters).
self.input = []
# Length of the final hash (in bytes).
self.HASH_LENGTH = 16
# Length of a block (the number of bytes hashed in every transform).
self.DATA_LENGTH = 64
# Call a separate init function, that can be used repeatedly
# to start from scratch on the same object.
self.init()
def init(self):
"Initialize the message-digest and set all fields to zero."
self.length = 0L
self.input = []
# Load magic initialization constants.
self.A = 0x67452301L
self.B = 0xefcdab89L
self.C = 0x98badcfeL
self.D = 0x10325476L
def _transform(self, inp):
"""Basic MD5 step transforming the digest based on the input.
Note that if the Mysterious Constants are arranged backwards
in little-endian order and decrypted with the DES they produce
OCCULT MESSAGES!
"""
a, b, c, d = A, B, C, D = self.A, self.B, self.C, self.D
# Round 1.
S11, S12, S13, S14 = 7, 12, 17, 22
a = XX(F, a, b, c, d, inp[ 0], S11, 0xD76AA478L) # 1
d = XX(F, d, a, b, c, inp[ 1], S12, 0xE8C7B756L) # 2
c = XX(F, c, d, a, b, inp[ 2], S13, 0x242070DBL) # 3
b = XX(F, b, c, d, a, inp[ 3], S14, 0xC1BDCEEEL) # 4
a = XX(F, a, b, c, d, inp[ 4], S11, 0xF57C0FAFL) # 5
d = XX(F, d, a, b, c, inp[ 5], S12, 0x4787C62AL) # 6
c = XX(F, c, d, a, b, inp[ 6], S13, 0xA8304613L) # 7
b = XX(F, b, c, d, a, inp[ 7], S14, 0xFD469501L) # 8
a = XX(F, a, b, c, d, inp[ 8], S11, 0x698098D8L) # 9
d = XX(F, d, a, b, c, inp[ 9], S12, 0x8B44F7AFL) # 10
c = XX(F, c, d, a, b, inp[10], S13, 0xFFFF5BB1L) # 11
b = XX(F, b, c, d, a, inp[11], S14, 0x895CD7BEL) # 12
a = XX(F, a, b, c, d, inp[12], S11, 0x6B901122L) # 13
d = XX(F, d, a, b, c, inp[13], S12, 0xFD987193L) # 14
c = XX(F, c, d, a, b, inp[14], S13, 0xA679438EL) # 15
b = XX(F, b, c, d, a, inp[15], S14, 0x49B40821L) # 16
# Round 2.
S21, S22, S23, S24 = 5, 9, 14, 20
a = XX(G, a, b, c, d, inp[ 1], S21, 0xF61E2562L) # 17
d = XX(G, d, a, b, c, inp[ 6], S22, 0xC040B340L) # 18
c = XX(G, c, d, a, b, inp[11], S23, 0x265E5A51L) # 19
b = XX(G, b, c, d, a, inp[ 0], S24, 0xE9B6C7AAL) # 20
a = XX(G, a, b, c, d, inp[ 5], S21, 0xD62F105DL) # 21
d = XX(G, d, a, b, c, inp[10], S22, 0x02441453L) # 22
c = XX(G, c, d, a, b, inp[15], S23, 0xD8A1E681L) # 23
b = XX(G, b, c, d, a, inp[ 4], S24, 0xE7D3FBC8L) # 24
a = XX(G, a, b, c, d, inp[ 9], S21, 0x21E1CDE6L) # 25
d = XX(G, d, a, b, c, inp[14], S22, 0xC33707D6L) # 26
c = XX(G, c, d, a, b, inp[ 3], S23, 0xF4D50D87L) # 27
b = XX(G, b, c, d, a, inp[ 8], S24, 0x455A14EDL) # 28
a = XX(G, a, b, c, d, inp[13], S21, 0xA9E3E905L) # 29
d = XX(G, d, a, b, c, inp[ 2], S22, 0xFCEFA3F8L) # 30
c = XX(G, c, d, a, b, inp[ 7], S23, 0x676F02D9L) # 31
b = XX(G, b, c, d, a, inp[12], S24, 0x8D2A4C8AL) # 32
# Round 3.
S31, S32, S33, S34 = 4, 11, 16, 23
a = XX(H, a, b, c, d, inp[ 5], S31, 0xFFFA3942L) # 33
d = XX(H, d, a, b, c, inp[ 8], S32, 0x8771F681L) # 34
c = XX(H, c, d, a, b, inp[11], S33, 0x6D9D6122L) # 35
b = XX(H, b, c, d, a, inp[14], S34, 0xFDE5380CL) # 36
a = XX(H, a, b, c, d, inp[ 1], S31, 0xA4BEEA44L) # 37
d = XX(H, d, a, b, c, inp[ 4], S32, 0x4BDECFA9L) # 38
c = XX(H, c, d, a, b, inp[ 7], S33, 0xF6BB4B60L) # 39
b = XX(H, b, c, d, a, inp[10], S34, 0xBEBFBC70L) # 40
a = XX(H, a, b, c, d, inp[13], S31, 0x289B7EC6L) # 41
d = XX(H, d, a, b, c, inp[ 0], S32, 0xEAA127FAL) # 42
c = XX(H, c, d, a, b, inp[ 3], S33, 0xD4EF3085L) # 43
b = XX(H, b, c, d, a, inp[ 6], S34, 0x04881D05L) # 44
a = XX(H, a, b, c, d, inp[ 9], S31, 0xD9D4D039L) # 45
d = XX(H, d, a, b, c, inp[12], S32, 0xE6DB99E5L) # 46
c = XX(H, c, d, a, b, inp[15], S33, 0x1FA27CF8L) # 47
b = XX(H, b, c, d, a, inp[ 2], S34, 0xC4AC5665L) # 48
# Round 4.
S41, S42, S43, S44 = 6, 10, 15, 21
a = XX(I, a, b, c, d, inp[ 0], S41, 0xF4292244L) # 49
d = XX(I, d, a, b, c, inp[ 7], S42, 0x432AFF97L) # 50
c = XX(I, c, d, a, b, inp[14], S43, 0xAB9423A7L) # 51
b = XX(I, b, c, d, a, inp[ 5], S44, 0xFC93A039L) # 52
a = XX(I, a, b, c, d, inp[12], S41, 0x655B59C3L) # 53
d = XX(I, d, a, b, c, inp[ 3], S42, 0x8F0CCC92L) # 54
c = XX(I, c, d, a, b, inp[10], S43, 0xFFEFF47DL) # 55
b = XX(I, b, c, d, a, inp[ 1], S44, 0x85845DD1L) # 56
a = XX(I, a, b, c, d, inp[ 8], S41, 0x6FA87E4FL) # 57
d = XX(I, d, a, b, c, inp[15], S42, 0xFE2CE6E0L) # 58
c = XX(I, c, d, a, b, inp[ 6], S43, 0xA3014314L) # 59
b = XX(I, b, c, d, a, inp[13], S44, 0x4E0811A1L) # 60
a = XX(I, a, b, c, d, inp[ 4], S41, 0xF7537E82L) # 61
d = XX(I, d, a, b, c, inp[11], S42, 0xBD3AF235L) # 62
c = XX(I, c, d, a, b, inp[ 2], S43, 0x2AD7D2BBL) # 63
b = XX(I, b, c, d, a, inp[ 9], S44, 0xEB86D391L) # 64
A = (A + a) & 0xffffffffL
B = (B + b) & 0xffffffffL
C = (C + c) & 0xffffffffL
D = (D + d) & 0xffffffffL
self.A, self.B, self.C, self.D = A, B, C, D
# Down from here all methods follow the Python Standard Library
# API of the md5 module.
def update(self, inBuf):
"""Add to the current message.
Update the md5 object with the string arg. Repeated calls
are equivalent to a single call with the concatenation of all
the arguments, i.e. m.update(a); m.update(b) is equivalent
to m.update(a+b).
"""
leninBuf = long(len(inBuf))
# Compute number of bytes mod 64.
index = (self.count[0] >> 3) & 0x3FL
# Update number of bits.
self.count[0] = self.count[0] + (leninBuf << 3)
if self.count[0] < (leninBuf << 3):
self.count[1] = self.count[1] + 1
self.count[1] = self.count[1] + (leninBuf >> 29)
partLen = 64 - index
if leninBuf >= partLen:
self.input[index:] = map(None, inBuf[:partLen])
self._transform(_bytelist2long(self.input))
i = partLen
while i + 63 < leninBuf:
self._transform(_bytelist2long(map(None, inBuf[i:i+64])))
i = i + 64
else:
self.input = map(None, inBuf[i:leninBuf])
else:
i = 0
self.input = self.input + map(None, inBuf)
def digest(self):
"""Terminate the message-digest computation and return digest.
Return the digest of the strings passed to the update()
method so far. This is a 16-byte string which may contain
non-ASCII characters, including null bytes.
"""
A = self.A
B = self.B
C = self.C
D = self.D
input = [] + self.input
count = [] + self.count
index = (self.count[0] >> 3) & 0x3fL
if index < 56:
padLen = 56 - index
else:
padLen = 120 - index
padding = ['\200'] + ['\000'] * 63
self.update(padding[:padLen])
# Append length (before padding).
bits = _bytelist2long(self.input[:56]) + count
self._transform(bits)
# Store state in digest.
digest = _long2bytes(self.A << 96, 16)[:4] + \
_long2bytes(self.B << 64, 16)[4:8] + \
_long2bytes(self.C << 32, 16)[8:12] + \
_long2bytes(self.D, 16)[12:]
self.A = A
self.B = B
self.C = C
self.D = D
self.input = input
self.count = count
return digest
def hexdigest(self):
"""Terminate and return digest in HEX form.
Like digest() except the digest is returned as a string of
length 32, containing only hexadecimal digits. This may be
used to exchange the value safely in email or other non-
binary environments.
"""
d = map(None, self.digest())
d = map(ord, d)
d = map(lambda x:"%02x" % x, d)
d = string.join(d, '')
return d
def copy(self):
"""Return a clone object.
Return a copy ('clone') of the md5 object. This can be used
to efficiently compute the digests of strings that share
a common initial substring.
"""
return copy.deepcopy(self)
# ======================================================================
# Mimick Python top-level functions from standard library API
# for consistency with the md5 module of the standard library.
# ======================================================================
def new(arg=None):
"""Return a new md5 object.
If arg is present, the method call update(arg) is made.
"""
md5 = MD5()
if arg:
md5.update(arg)
return md5
def md5(arg=None):
"""Same as new().
For backward compatibility reasons, this is an alternative
name for the new() function.
"""
return new(arg)
test_md5.py
"""A test script to compare MD5 implementations.
A note about performance: the pure Python MD5 takes roughly
160 sec. per MB of data on a 233 MHz Intel Pentium CPU.
"""
import string, unittest
import md5 # Python's implementation in C.
import md5py # The pure Python implementation.
# Helpers...
def formatHex(str):
"Print a string's HEX code in groups of two digits."
d = map(None, str)
d = map(ord, d)
d = map(lambda x:"%02x" % x, d)
return string.join(d, " ")
def format(str):
"Print a string as-is in groups of two characters."
s = ''
for i in range(0, len(str)-1, 2):
s = s + "%03s" % str[i:i+2]
return s[1:]
def printDiff(message, d1, d2, expectedResult=None):
"Print different outputs for same message."
print "Message: '%s'" % message
print "Message length: %d" % len(message)
if expectedResult:
print "%-48s (expected)" % format(expectedResult)
print "%-48s (Std. lib. MD5)" % formatHex(d1)
print "%-48s (Pure Python MD5)" % formatHex(d2)
print
# The real comparison function.
def compareImp(message):
"""Compare two MD5 implementations, C vs. pure Python module.
For equal digests this returns None, otherwise it returns
a tuple of both digests.
"""
# Use Python's standard library MD5 compiled C module.
m1 = md5.md5()
m1.update(message)
d1 = m1.digest()
d1h = m1.hexdigest()
# Use MD5 module in pure Python.
m2 = md5py.md5()
m2.update(message)
d2 = m2.digest()
d2h = m2.hexdigest()
# Return None if equal or the different digests if not equal.
if d1 == d2 and d1h == d2h:
return
else:
return d1, d2
class MD5CompareTestCase(unittest.TestCase):
"Compare pure Python MD5 against Python's std. lib. version."
def test1(self):
"Test cases with known digest result."
cases = (
("",
"d41d8cd98f00b204e9800998ecf8427e"),
("a",
"0cc175b9c0f1b6a831c399e269772661"),
("abc",
"900150983cd24fb0d6963f7d28e17f72"),
("message digest",
"f96b697d7cb7938d525a2f31aaf161d0"),
("abcdefghijklmnopqrstuvwxyz",
"c3fcd3d76192e4007dfb496cca67e13b"),
("ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789",
"d174ab98d277d9f5a5611c2c9f419d9f"),
("1234567890"*8,
"57edf4a22be3c955ac49da2e2107b67a"),
)
for i in xrange(len(cases)):
res = compareImp(cases[i][0])
try:
assert res == None
except AssertionError, details:
d1, d2 = res
message, expectedResult = cases[i][0], None
if len(cases[i]) == 2:
expectedResult = cases[i][1]
printDiff(message, d1, d2, expectedResult)
def test2(self):
"Test cases without known digest result."
cases = (
"123",
"1234",
"12345",
"123456",
"1234567",
"12345678",
"123456789 123456789 123456789 ",
"123456789 123456789 ",
"123456789 123456789 1",
"123456789 123456789 12",
"123456789 123456789 123",
"123456789 123456789 1234",
"123456789 123456789 123456789 1",
"123456789 123456789 123456789 12",
"123456789 123456789 123456789 123",
"123456789 123456789 123456789 1234",
"123456789 123456789 123456789 12345",
"123456789 123456789 123456789 123456",
"123456789 123456789 123456789 1234567",
"123456789 123456789 123456789 12345678",
)
for i in xrange(len(cases)):
res = compareImp(cases[i][0])
try:
assert res == None
except AssertionError, details:
d1, d2 = res
message = cases[i][0]
printDiff(message, d1, d2)
def test3(self):
"Test cases with long messages (can take a while)."
cases = (
(2**10*'a',),
(2**10*'abcd',),
## (2**20*'a',), ## 1 MB, takes about 160 sec. on a 233 Mhz Pentium.
)
for i in xrange(len(cases)):
res = compareImp(cases[i][0])
try:
assert res == None
except AssertionError, details:
d1, d2 = res
message = cases[i][0]
printDiff(message, d1, d2)
def test4(self):
"Test cases with increasingly growing message lengths."
i = 0
while i < 2**5:
message = i * 'a'
res = compareImp(message)
try:
assert res == None
except AssertionError, details:
d1, d2 = res
printDiff(message, d1, d2)
i = i + 1
def test5(self):
"Test updating cloned objects."
cases = (
"123",
"1234",
"12345",
"123456",
"1234567",
"12345678",
"123456789 123456789 123456789 ",
"123456789 123456789 ",
"123456789 123456789 1",
"123456789 123456789 12",
"123456789 123456789 123",
"123456789 123456789 1234",
"123456789 123456789 123456789 1",
"123456789 123456789 123456789 12",
"123456789 123456789 123456789 123",
"123456789 123456789 123456789 1234",
"123456789 123456789 123456789 12345",
"123456789 123456789 123456789 123456",
"123456789 123456789 123456789 1234567",
"123456789 123456789 123456789 12345678",
)
# Load both with same prefix.
prefix1 = 2**10 * 'a'
m1 = md5.md5()
m1.update(prefix1)
m1c = m1.copy()
m2 = md5py.md5()
m2.update(prefix1)
m2c = m2.copy()
# Update and compare...
for i in xrange(len(cases)):
message = cases[i][0]
m1c.update(message)
d1 = m1c.hexdigest()
m2c.update(message)
d2 = m2c.hexdigest()
assert d1 == d2
def makeSuite():
suite = unittest.TestSuite()
suite.addTest(MD5CompareTestCase('test1'))
suite.addTest(MD5CompareTestCase('test2'))
suite.addTest(MD5CompareTestCase('test3'))
suite.addTest(MD5CompareTestCase('test4'))
suite.addTest(MD5CompareTestCase('test5'))
return suite
if __name__ == "__main__":
unittest.TextTestRunner().run(makeSuite())