Hash instance, using MD5 and elf

Import Java. io. unsupportedencodingexception; <br/> Import Java. security. messagedigest; <br/> Import Java. security. nosuchalgorithmexception; <br/> Import Java. util. random; <br/> public class hashcomputer {<br/> Private Static messagedigest MD5; <br/> static {<br/> try {<br/> MD5 = messagedigest. getinstance ("MD5"); <br/>} catch (nosuchalgorithmexception e) {<br/> // todo auto-generated Catch Block <br/> E. Printstacktrace (); <br/>}< br/>/** <br/> * Get the MD5 of the given key. <br/> */<br/> Public static byte [] computemd5 (string K) {<br/> md5.reset (); <br/> byte [] keybytes = NULL; <br/> try {<br/> keybytes = K. getbytes ("UTF-8"); <br/>} catch (unsupportedencodingexception e) {<br/> throw new runtimeexception ("unknown string:" + k, e ); <br/>}< br/> md5.update (keybytes); <br/> return md5.digest (); <Br/>}< br/> Public static int Hash (string Str) {<br/> byte [] BB = computemd5 (STR ); <br/> long Rv = (long) bb [3] & 0xff) <24) <br/> | (long) bb [2] & 0xff) <16) <br/> | (long) bb [1] & 0xff) <8) <br/> | (long) bb [0] & 0xff) <0); <br/> return (INT) (RV & 0x7fffffffl ); /* truncate to 32-bits */<br/>}< br/> Public static int hash2 (string Str) <br/>{< br/> int hash = 0; <br/> int x = 0; </P> <P> for (INT I = 0; I <Str. length (); I ++) <br/>{< br/> hash = (hash <4) + Str. charat (I); <br/> If (x = (INT) (hash & 0xf0000000l ))! = 0) <br/>{< br/> hash ^ = (x> 24); <br/> hash & = ~ X; <br/>}< br/> return (hash & 0x7fffffff ); <br/>}</P> <p> Public static void main (string [] ARGs) {<br/> int n = 55; <br/> int AA [] = new int [N]; <br/> for (INT I = 0; I <n; I ++) <br/> AA [I] = 0; </P> <p> random ran = new random (); <br/> for (INT I = 0; I <100000; I ++) {<br/> int length = ran. nextint (30); <br/> stringbuffer sb = new stringbuffer (length); <br/> for (INT n = 0; n <length; n ++) {<br/> Sb. append (char) (ran. nextint (95) + 32); <br/>}< br/> int L = hashcomputer. hash (sb. tostring (); <br/> AA [l % N] ++; <br/>}</P> <p> for (INT I = 0; I <n; I ++) <br/> system. out. println (AA [I]); <br/>}< br/> 

 

When the modulo is reached, there will be a very uneven, depressing

 

 

7422

7365

7483

7319

7610

7280

7654

10627

7495

7473

7625

7341

7306

 

The reason is that the length of the string to be tested is randomly generated: int length = ran. nextint (30); nextint may generate 0. The hash results of these strings with the length of 0 are the same, and they are all calculated in one bucket, as a result, the string of this bucket is significantly greater than that of other buckets. This has nothing to do with the hash function used, but with the tested data.

 

The solution is

Int length = ran. nextint (30 );

Change

Int length = ran. nextint (20) + 10;