git.m6w6.name Git - m6w6/libmemcached/blob - libhashkit/murmur.cc

   1 /*
   2   "Murmur" hash provided by Austin, tanjent@gmail.com
   3   http://murmurhash.googlepages.com/
   4
   5   Note - This code makes a few assumptions about how your machine behaves -
   6
   7   1. We can read a 4-byte value from any address without crashing
   8   2. sizeof(int) == 4
   9
  10   And it has a few limitations -
  11   1. It will not work incrementally.
  12   2. It will not produce the same results on little-endian and big-endian
  13   machines.
  14
  15   Updated to murmur2 hash - BP
  16 */
  17
  18 #include <libhashkit/common.h>
  19
  20 uint32_t hashkit_murmur(const char *key, size_t length, void *context)
  21 {
  22   /*
  23     'm' and 'r' are mixing constants generated offline.  They're not
  24     really 'magic', they just happen to work well.
  25   */
  26
  27   const unsigned int m= 0x5bd1e995;
  28   const uint32_t seed= (0xdeadbeef * (uint32_t)length);
  29   const int r= 24;
  30
  31
  32   // Initialize the hash to a 'random' value
  33
  34   uint32_t h= seed ^ (uint32_t)length;
  35
  36   // Mix 4 bytes at a time into the hash
  37
  38   const unsigned char * data= (const unsigned char *)key;
  39   (void)context;
  40
  41   while(length >= 4)
  42   {
  43     unsigned int k = *(unsigned int *)data;
  44
  45     k *= m;
  46     k ^= k >> r;
  47     k *= m;
  48
  49     h *= m;
  50     h ^= k;
  51
  52     data += 4;
  53     length -= 4;
  54   }
  55
  56   // Handle the last few bytes of the input array
  57
  58   switch(length)
  59   {
  60   case 3: h ^= ((uint32_t)data[2]) << 16;
  61   case 2: h ^= ((uint32_t)data[1]) << 8;
  62   case 1: h ^= data[0];
  63           h *= m;
  64   default: break;
  65   };
  66
  67   /*
  68     Do a few final mixes of the hash to ensure the last few bytes are
  69     well-incorporated.
  70   */
  71
  72   h ^= h >> 13;
  73   h *= m;
  74   h ^= h >> 15;
  75
  76   return h;
  77 }