首页 > 代码库 > 【转】各种字符串Hash函数比较

【转】各种字符串Hash函数比较

2024-07-11 05:07:05   220人阅读

常用的字符串Hash函数还有ELFHash，APHash等等，都是十分简单有效的方法。这些函数使用位运算使得每一个字符都对最后的函数值产生影响。另外还有以MD5和SHA1为代表的杂凑函数，这些函数几乎不可能找到碰撞。

常用字符串哈希函数有BKDRHash，APHash，DJBHash，JSHash，RSHash，SDBMHash，PJWHash，ELFHash等等。对于以上几种哈希函数，我对其进行了一个小小的评测。

Hash函数数据1数据2数据3数据4数据1得分数据2得分数据3得分数据4得分平均分
BKDRHash20477448196.5510090.9582.0592.64
APHash23475449396.5588.4610051.2886.28
DJBHash22497547496.5592.31010083.43
JSHash14476150610084.6296.8317.9581.94
RSHash10486150510010051.5820.5175.96
SDBMHash32484950493.192.3157.0123.0872.41
PJWHash302648785130043.89021.95
ELFHash302648785130043.89021.95

其中数据1为100000个字母和数字组成的随机串哈希冲突个数。数据2为100000个有意义的英文句子哈希冲突个数。数据3为数据1的哈希值与1000003(大素数)求模后存储到线性表中冲突的个数。数据4为数据1的哈希值与10000019(更大素数)求模后存储到线性表中冲突的个数。

经过比较，得出以上平均得分。平均数为平方平均数。可以发现，BKDRHash无论是在实际效果还是编码实现中，效果都是最突出的。APHash也是较为优秀的算法。DJBHash,JSHash,RSHash与SDBMHash各有千秋。PJWHash与ELFHash效果最差，但得分相似，其算法本质是相似的。

在信息修竞赛中，要本着易于编码调试的原则，个人认为BKDRHash是最适合记忆和使用的。

BYVoid原创，欢迎建议、交流、批评和指正。

附：各种哈希函数的C语言程序代码

 

  1 unsigned int SDBMHash(char *str)  2 {  3     unsigned int hash = 0;  4    5     while (*str)  6     {  7         // equivalent to: hash = 65599*hash + (*str++);  8         hash = (*str++) + (hash << 6) + (hash << 16) - hash;  9     } 10   11     return (hash & 0x7FFFFFFF); 12 } 13   14 // RS Hash Function 15 unsigned int RSHash(char *str) 16 { 17     unsigned int b = 378551; 18     unsigned int a = 63689; 19     unsigned int hash = 0; 20   21     while (*str) 22     { 23         hash = hash * a + (*str++); 24         a *= b; 25     } 26   27     return (hash & 0x7FFFFFFF); 28 } 29   30 // JS Hash Function 31 unsigned int JSHash(char *str) 32 { 33     unsigned int hash = 1315423911; 34   35     while (*str) 36     { 37         hash ^= ((hash << 5) + (*str++) + (hash >> 2)); 38     } 39   40     return (hash & 0x7FFFFFFF); 41 } 42   43 // P. J. Weinberger Hash Function 44 unsigned int PJWHash(char *str) 45 { 46     unsigned int BitsInUnignedInt = (unsigned int)(sizeof(unsigned int) * 8); 47     unsigned int ThreeQuarters    = (unsigned int)((BitsInUnignedInt  * 3) / 4); 48     unsigned int OneEighth        = (unsigned int)(BitsInUnignedInt / 8); 49     unsigned int HighBits         = (unsigned int)(0xFFFFFFFF) << (BitsInUnignedInt - OneEighth); 50     unsigned int hash             = 0; 51     unsigned int test             = 0; 52   53     while (*str) 54     { 55         hash = (hash << OneEighth) + (*str++); 56         if ((test = hash & HighBits) != 0) 57         { 58             hash = ((hash ^ (test >> ThreeQuarters)) & (~HighBits)); 59         } 60     } 61   62     return (hash & 0x7FFFFFFF); 63 } 64   65 // ELF Hash Function 66 unsigned int ELFHash(char *str) 67 { 68     unsigned int hash = 0; 69     unsigned int x    = 0; 70   71     while (*str) 72     { 73         hash = (hash << 4) + (*str++); 74         if ((x = hash & 0xF0000000L) != 0) 75         { 76             hash ^= (x >> 24); 77             hash &= ~x; 78         } 79     } 80   81     return (hash & 0x7FFFFFFF); 82 } 83   84 // BKDR Hash Function 85 unsigned int BKDRHash(char *str) 86 { 87     unsigned int seed = 131; // 31 131 1313 13131 131313 etc.. 88     unsigned int hash = 0; 89   90     while (*str) 91     { 92         hash = hash * seed + (*str++); 93     } 94   95     return (hash & 0x7FFFFFFF); 96 } 97   98 // DJB Hash Function 99 unsigned int DJBHash(char *str)100 {101     unsigned int hash = 5381;102  103     while (*str)104     {105         hash += (hash << 5) + (*str++);106     }107  108     return (hash & 0x7FFFFFFF);109 }110  111 // AP Hash Function112 unsigned int APHash(char *str)113 {114     unsigned int hash = 0;115     int i;116  117     for (i=0; *str; i++)118     {119         if ((i & 1) == 0)120         {121             hash ^= ((hash << 7) ^ (*str++) ^ (hash >> 3));122         }123         else124         {125             hash ^= (~((hash << 11) ^ (*str++) ^ (hash >> 5)));126         }127     }128  129     return (hash & 0x7FFFFFFF);130 }

转载自：http://www.byvoid.com/blog/string-hash-compare/

 


联系
我们