jens@1: /*
jens@1:  *  MurmurHash.c
jens@1:  *  MYUtilities
jens@1:  *
jens@1:  *  This file created by Jens Alfke on 3/17/08.
jens@1:  *  Algorithm & source code by Austin Appleby, released to public domain.
jens@1:  *  <http://murmurhash.googlepages.com/>
jens@1:  *  Downloaded 3/16/2008.
jens@1:  *  Modified slightly by Jens Alfke (use standard uint32_t and size_t types;
jens@1:  *  change 'm' and 'r' to #defines for better C compatibility.)
jens@1:  *
jens@1:  */
jens@1: 
jens@1: #include "MurmurHash.h"
jens@1: 
jens@1: 
jens@1: //-----------------------------------------------------------------------------
jens@1: // MurmurHash2, by Austin Appleby
jens@1: 
jens@1: // Note - This code makes a few assumptions about how your machine behaves -
jens@1: 
jens@1: // 1. We can read a 4-byte value from any address without crashing
jens@1: // 2. sizeof(int) == 4      **Jens: I fixed this by changing 'unsigned int' to 'uint32_t'**
jens@1: 
jens@1: // And it has a few limitations -
jens@1: 
jens@1: // 1. It will not work incrementally.
jens@1: // 2. It will not produce the same results on little-endian and big-endian
jens@1: //    machines.
jens@1: 
jens@1: uint32_t MurmurHash2 ( const void * key, size_t len, uint32_t seed )
jens@1: {
jens@1:     // 'm' and 'r' are mixing constants generated offline.
jens@1:     // They're not really 'magic', they just happen to work well.
jens@1:     
jens@1:     #define m 0x5bd1e995
jens@1:     #define r 24
jens@1:     
jens@1:     // Initialize the hash to a 'random' value
jens@1:     
jens@1:     uint32_t h = seed ^ len;
jens@1:     
jens@1:     // Mix 4 bytes at a time into the hash
jens@1:     
jens@1:     const unsigned char * data = (const unsigned char *)key;
jens@1:     
jens@1:     while(len >= 4)
jens@1:     {
jens@1:         uint32_t k = *(uint32_t *)data;
jens@1:         
jens@1:         k *= m; 
jens@1:         k ^= k >> r; 
jens@1:         k *= m; 
jens@1:         
jens@1:         h *= m; 
jens@1:         h ^= k;
jens@1:         
jens@1:         data += 4;
jens@1:         len -= 4;
jens@1:     }
jens@1:     
jens@1:     // Handle the last few bytes of the input array
jens@1:     
jens@1:     switch(len)
jens@1:     {
jens@1: 	case 3: h ^= data[2] << 16;
jens@1: 	case 2: h ^= data[1] << 8;
jens@1: 	case 1: h ^= data[0];
jens@1:                 h *= m;
jens@1:     };
jens@1:     
jens@1:     // Do a few final mixes of the hash to ensure the last few
jens@1:     // bytes are well-incorporated.
jens@1:     
jens@1:     h ^= h >> 13;
jens@1:     h *= m;
jens@1:     h ^= h >> 15;
jens@1:     
jens@1:     return h;
jens@1: }