html数据摘要算法,MD5消息摘要算法

graph LR;

step1([位填充])-->step2([初始化MD缓存]);

step2 --> step3([数据处理]);

step3 --> step4([拼接输出]);

记有消息\(Msg\), 其位长度为\(x_l\), 那么进行MD5计算之前, 需要经过如下位填充步骤:

在\(Msg\)末尾填充一个位\(1\), 然后填充若干个位\(0\), 以使其位长度模除512值为448;

注意: 即使\(x_l\%512=448\), 依然需要先填充一个1, 然后补齐447个0;

将\(x_l\)放入到64位长度缓存中(\(x_l\)超过64位长度则截断低64位存入, 不足64位则高位补0), 然后依次将该缓存中的数据按照从低字节到高字节填充到\(Msg\)的尾部;

记经过以上两步位填充后的按32位长划分的数组为\(M[0, \cdots, N-1]\), 其中\(N=y_{l}/32\), \(y_l\)表示\(Msg\)填充后的长度;

初始化4个32位长的摘要缓存, 记为\(A, B, C, D\):

uint32_t A = 0x67452301;

uint32_t B = 0xEFCDAB89;

uint32_t C = 0x98BADCFE;

uint32_t D = 0x10325476;

数据处理算法如下:

# <<< 表示循环左移

# uint32->x,y,z

F(x, y, z) = (x & y) | ((~x) & z)

G(x, y, z) = (x & z) | (y & (~z))

H(x, y, z) = x ^ y ^ z

I(x, y, z) = y ^ (x | (~z))

# uint32->a,b,c,d,Mi,s,t

DP(a, b, c, d, Mi, s, t, Callback) =

a = b + ((a + Callback(b, c, d) + Mi + t) <<< s)

CALLBACK = [F, G, H, I]

WORDIDX = [lambda x:x, lambda x:(1+5*x)&15, lambda x:(5+3*x)&15, lambda x:(7*x)&15]

T = lambda x: uint32(pow(uint64(2),32) * abs(sin(x)))

S = [7, 12, 17, 22, 5, 9, 14, 20, 4, 11, 16, 23, 6, 10, 15, 21]

for blockIdx in 0...N/16:

for i in 0...16:

X[i] = M[blockIdx*16+j]

AA = A; BB = B; CC = C; DD = D

for roundIdx in 0...4:

callback = CALLBACK[roundIdx]

wordfuc = WORDIDX[roundIdx]

for wordIdx in 0...4:

DP(A, B, C, D, M[wordfuc(wordIdx*4)], S[roundIdx*4 + wordIdx], T(wordIdx*4), callback)

DP(D, A, B, C, M[wordfuc(wordIdx*4+1)], S[roundIdx*4 + wordIdx], T(wordIdx*4+1), callback)

DP(C, D, A, B, M[wordfuc(wordIdx*4+2)], S[roundIdx*4 + wordIdx], T(wordIdx*4+2), callback)

DP(B, C, D, A, M[wordfuc(wordIdx*4+3)], S[roundIdx*4 + wordIdx], T(wordIdx*4+3), callback)

end

end

A = A + AA

B = B + BB

C = C + CC

D = D + DD

end

MD5 = [A, B, C, D]