CTSpooky.cpp

// Spooky Hash
// A 128-bit noncryptographic hash, for checksums and table lookup
// By Bob Jenkins.  Public domain.
//   Oct 31 2010: published framework, disclaimer ShortHash isn't right
//   Nov 7 2010: disabled ShortHash
//   Oct 31 2011: replace End, ShortMix, ShortEnd, enable ShortHash again
//   April 10 2012: buffer overflow on platforms without unaligned reads
//   July 12 2012: was passing out variables in final to in/out in short
//   July 30 2012: I reintroduced the buffer overflow
//   August 5 2012: SpookyV2: d = should be d += in short hash, and remove extra mix from long hash

#include "CTSpooky.h"


// init spooky state
void CTSpookyHash::Init(uint64 seed1, uint64 seed2)
{
    m_length = 0;
    m_remainder = 0;
    m_state[0] = seed1;
    m_state[1] = seed2;
}


// add a message fragment to the state
void CTSpookyHash::Update(const void *message, size_t length)
{
    uint64 h0,h1,h2,h3,h4,h5,h6,h7,h8,h9,h10,h11;
    size_t newLength = length + m_remainder;
    uint8  remainder;
    union
    {
        const uint8 *p8;
        uint64 *p64;
        size_t i;
    } u;
    const uint64 *end;

    // Is this message fragment too short?  If it is, stuff it away.
    if (newLength < sc_bufSize)
    {
        memcpy(&((uint8 *)m_data)[m_remainder], message, length);
        m_length = length + m_length;
        m_remainder = (uint8)newLength;
        return;
    }

    // init the variables
    if (m_length < sc_bufSize)
    {
        h0=h3=h6=h9  = m_state[0];
        h1=h4=h7=h10 = m_state[1];
        h2=h5=h8=h11 = sc_const;
    }
    else
    {
        h0 = m_state[0];
        h1 = m_state[1];
        h2 = m_state[2];
        h3 = m_state[3];
        h4 = m_state[4];
        h5 = m_state[5];
        h6 = m_state[6];
        h7 = m_state[7];
        h8 = m_state[8];
        h9 = m_state[9];
        h10 = m_state[10];
        h11 = m_state[11];
    }
    m_length = length + m_length;

    // if we've got anything stuffed away, use it now
    if (m_remainder)
    {
        uint8 prefix = sc_bufSize-m_remainder;
        memcpy(&(((uint8 *)m_data)[m_remainder]), message, prefix);
        u.p64 = m_data;
        Mix(u.p64, h0,h1,h2,h3,h4,h5,h6,h7,h8,h9,h10,h11);
        Mix(&u.p64[sc_numVars], h0,h1,h2,h3,h4,h5,h6,h7,h8,h9,h10,h11);
        u.p8 = ((const uint8 *)message) + prefix;
        length -= prefix;
    }
    else
    {
        u.p8 = (const uint8 *)message;
    }

    // handle all whole blocks of sc_blockSize bytes
    end = u.p64 + (length/sc_blockSize)*sc_numVars;
    remainder = (uint8)(length-((const uint8 *)end-u.p8));
    if (ALLOW_UNALIGNED_READS || (u.i & 0x7) == 0)
    {
        while (u.p64 < end)
        {
            Mix(u.p64, h0,h1,h2,h3,h4,h5,h6,h7,h8,h9,h10,h11);
            u.p64 += sc_numVars;
        }
    }
    else
    {
        while (u.p64 < end)
        {
            memcpy(m_data, u.p8, sc_blockSize);
            Mix(m_data, h0,h1,h2,h3,h4,h5,h6,h7,h8,h9,h10,h11);
            u.p64 += sc_numVars;
        }
    }

    // stuff away the last few bytes
    m_remainder = remainder;
    memcpy(m_data, end, remainder);

    // stuff away the variables
    m_state[0] = h0;
    m_state[1] = h1;
    m_state[2] = h2;
    m_state[3] = h3;
    m_state[4] = h4;
    m_state[5] = h5;
    m_state[6] = h6;
    m_state[7] = h7;
    m_state[8] = h8;
    m_state[9] = h9;
    m_state[10] = h10;
    m_state[11] = h11;
}


// report the hash for the concatenation of all message fragments so far
void CTSpookyHash::Final(uint64 *hash1, uint64 *hash2)
{
    // init the variables
    if (m_length < sc_bufSize)
    {
        *hash1 = m_state[0];
        *hash2 = m_state[1];
        Short((const char*)m_data, m_length, hash1, hash2);
        return;
    }

    const uint64 *data = (const uint64 *)m_data;
    uint8 remainder = m_remainder;

    uint64 h0 = m_state[0];
    uint64 h1 = m_state[1];
    uint64 h2 = m_state[2];
    uint64 h3 = m_state[3];
    uint64 h4 = m_state[4];
    uint64 h5 = m_state[5];
    uint64 h6 = m_state[6];
    uint64 h7 = m_state[7];
    uint64 h8 = m_state[8];
    uint64 h9 = m_state[9];
    uint64 h10 = m_state[10];
    uint64 h11 = m_state[11];

    if (remainder >= sc_blockSize)
    {
        // m_data can contain two blocks; handle any whole first block
        Mix(data, h0,h1,h2,h3,h4,h5,h6,h7,h8,h9,h10,h11);
        data += sc_numVars;
        remainder -= sc_blockSize;
    }

    // mix in the last partial block, and the length mod sc_blockSize
    memset(&((uint8 *)data)[remainder], 0, (sc_blockSize-remainder));

    ((uint8 *)data)[sc_blockSize-1] = remainder;

    // do some final mixing
    End(data, h0,h1,h2,h3,h4,h5,h6,h7,h8,h9,h10,h11);

    *hash1 = h0;
    *hash2 = h1;
}