【轉載】字符串Hash函數比較
原址鏈接:http://www.cmykrgb123.cn/blog/string-hash-compare/
常用的字符串Hash函數還有ELFHash,APHash等等,都是十分簡單有效的方法。這些函數使用位運算使得每一個字符都對最後的函數值產生影響。另外還有以MD5和SHA1爲代表的雜湊函數,這些函數幾乎不可能找到碰撞。
常用字符串哈希函數有BKDRHash,APHash,DJBHash,JSHash,RSHash,SDBMHash,PJWHash,ELFHash等等。對於以上幾種哈希函數,我對其進行了一個小小的評測。
Hash函數 | 數據1 | 數據2 | 數據3 | 數據4 | 數據1得分 | 數據2得分 | 數據3得分 | 數據4得分 | 平均分 |
BKDRHash | 2 | 0 | 4774 | 481 | 96.55 | 100 | 90.95 | 82.05 | 92.64 |
APHash | 2 | 3 | 4754 | 493 | 96.55 | 88.46 | 100 | 51.28 | 86.28 |
DJBHash | 2 | 2 | 4975 | 474 | 96.55 | 92.31 | 0 | 100 | 83.43 |
JSHash | 1 | 4 | 4761 | 506 | 100 | 84.62 | 96.83 | 17.95 | 81.94 |
RSHash | 1 | 0 | 4861 | 505 | 100 | 100 | 51.58 | 20.51 | 75.96 |
SDBMHash | 3 | 2 | 4849 | 504 | 93.1 | 92.31 | 57.01 | 23.08 | 72.41 |
PJWHash | 30 | 26 | 4878 | 513 | 0 | 0 | 43.89 | 0 | 21.95 |
ELFHash | 30 | 26 | 4878 | 513 | 0 | 0 | 43.89 | 0 | 21.95 |
其中數據1爲100000個字母和數字組成的隨機串哈希衝突個數。數據2爲100000個有意義的英文句子哈希衝突個數。數據3爲數據1的哈希值與 1000003(大素數)求模後存儲到線性表中衝突的個數。數據4爲數據1的哈希值與10000019(更大素數)求模後存儲到線性表中衝突的個數。
經過比較,得出以上平均得分。平均數爲平方平均數。可以發現,BKDRHash無論是在實際效果還是編碼實現中,效果都是最突出的。APHash也 是較爲優秀的算法。DJBHash,JSHash,RSHash與SDBMHash各有千秋。PJWHash與ELFHash效果最差,但得分相似,其算 法本質是相似的。
在信息修競賽中,要本着易於編碼調試的原則,個人認爲BKDRHash是最適合記憶和使用的。
CmYkRgB123原創,歡迎建議、交流、批評和指正。
附:各種哈希函數的C語言程序代碼
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 Function
112 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 else
124 {
125 hash ^= (~((hash << 11) ^ (*str++) ^ (hash >> 5)));
126 }
127 }
128
129 return (hash & 0x7FFFFFFF);
130 }