Optimize MurmurHash, benchmarked 33% faster

Author: tess3ract

fnlp-core/src/main/java/org/fnlp/util/hash/MurmurHash.java

@@ -19,10 +19,7 @@
 
 package org.fnlp.util.hash;
 
-import java.io.UnsupportedEncodingException;
-
-
-/** 
+/**
  *
  * murmur hash 2.0.
  * 
@@ -44,8 +41,21 @@ public final class MurmurHash extends AbstractHashCode {
 // all methods static; private constructor. 
 public MurmurHash() {}
 
+ protected byte[] toBytesWithoutEncoding(String str) {
+ int len = str.length();
+ int pos = 0;
+ byte[] buf = new byte[len << 1];
+ for (int i = 0; i < len; i++) {
+ char c = str.charAt(i);
+ buf[pos++] = (byte) (c & 0xFF);
+ buf[pos++] = (byte) (c >> 8);
+ }
+ return buf;
+ }
+
 public int hashcode(String str) {
-return hash32(str);
+ byte[] bytes = toBytesWithoutEncoding(str);
+ return hash32(bytes, bytes.length);
 }
 
 /** 
@@ -106,31 +116,6 @@ public int hash32( final byte[] data, int length) {
 }
 
 
-/** 
- * Generates 32 bit hash from a string.
- * 
- * @param text string to hash
- * @return 32 bit hash of the given string
- */
-public int hash32( final String text) {
-final byte[] bytes = text.getBytes(); 
-return hash32( bytes, bytes.length);
-}
-
-
-/** 
- * Generates 32 bit hash from a substring.
- * 
- * @param text string to hash
- * @param from starting index
- * @param length length of the substring to hash
- * @return 32 bit hash of the given string
- */
-public int hash32( final String text, int from, int length) {
-return hash32( text.substring( from, from+length));
-}
-
-
 /** 
  * Generates 64 bit hash from byte array of the given length and seed.
  * 
@@ -192,44 +177,4 @@ public long hash64( final byte[] data, int length) {
 return hash64( data, length, 0xe17a1465);
 }
 
-
-/** 
- * Generates 64 bit hash from a string.
- * 
- * @param text string to hash
- * @return 64 bit hash of the given string
- */
-public long hash64( final String text) {
-
-final byte[] bytes;
-try {
-bytes = text.getBytes("utf-8");
-
-return hash32( bytes, bytes.length);
-} catch (UnsupportedEncodingException e) {
-throw new RuntimeException("utf-8 encoding should be available on system.", e);
-}
-}
-
-
-
-/** 
- * Generates 64 bit hash from a substring.
- * 
- * @param text string to hash
- * @param from starting index
- * @param length length of the substring to hash
- * @return 64 bit hash of the given array
- */
-public long hash64( final String text, int from, int length) {
-
-final byte[] bytes;
-try {
-bytes = text.getBytes("utf-8");
-
-return hash64( bytes, bytes.length);
-} catch (UnsupportedEncodingException e) {
-throw new RuntimeException("utf-8 encoding should be available on system.", e);
-}
-}
-}
+}