hbase/util/JenkinsHash.java

/**
 * Copyright 2007 The Apache Software Foundation
 *
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package org.apache.hadoop.hbase.util;

/**
 * lookup3.c, by Bob Jenkins, May 2006, Public Domain.
 * <a href="http://burtleburtle.net/bob/c/lookup3.c">lookup3.c</a>
 *
 * You can use this free for any purpose.  It's in the public domain.
 * It has no warranty.
 * 
 * Produces 32-bit hash for hash table lookup.
 */
public class JenkinsHash {
  private static long INT_MASK  = 0x00000000ffffffffL;
  private static long BYTE_MASK = 0x00000000000000ffL;

  private static long rot(long val, int pos) {
    return Long.valueOf(Integer.rotateLeft(
        Long.valueOf(val & INT_MASK).intValue(), pos)).longValue() & INT_MASK;
  }

  /**
   * Alternate form for hashing an entire byte array
   * 
   * @param bytes 
   * @param initval
   * @return hash value
   */
  public static int hash(byte[] bytes, int initval) {
    return hash(bytes, bytes.length, initval);
  }
  
  /**
   * taken from  hashlittle() -- hash a variable-length key into a 32-bit value
   * 
   * @param key the key (the unaligned variable-length array of bytes)
   * @param nbytes number of bytes to include in hash
   * @param initval can be any integer value
   * @return a 32-bit value.  Every bit of the key affects every bit of the
   * return value.  Two keys differing by one or two bits will have totally
   * different hash values.
   * 
   * The best hash table sizes are powers of 2.  There is no need to do mod a
   * prime (mod is sooo slow!).  If you need less than 32 bits, use a bitmask.
   * For example, if you need only 10 bits, do h = (h & hashmask(10));
   * In which case, the hash table should have hashsize(10) elements.
   * 
   * If you are hashing n strings byte[][] k, do it like this:
   * for (int i = 0, h = 0; i < n; ++i) h = hash( k[i], h);
   * 
   * By Bob Jenkins, 2006.  bob_jenkins@burtleburtle.net.  You may use this
   * code any way you wish, private, educational, or commercial.  It's free.
   * 
   * Use for hash table lookup, or anything where one collision in 2^^32 is
   * acceptable.  Do NOT use for cryptographic purposes.
  */
  public static int hash(byte[] key, int nbytes, int initval) {
    int length = nbytes;
    long a, b, c;       // We use longs because we don't have unsigned ints
    a = b = c = (0x00000000deadbeefL + length + initval) & INT_MASK;
    int offset = 0;
    for (; length > 12; offset += 12, length -= 12) {
      a = (a + (key[offset + 0]    & BYTE_MASK)) & INT_MASK;
      a = (a + (((key[offset + 1]  & BYTE_MASK) <<  8) & INT_MASK)) & INT_MASK;
      a = (a + (((key[offset + 2]  & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
      a = (a + (((key[offset + 3]  & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
      b = (b + (key[offset + 4]    & BYTE_MASK)) & INT_MASK;
      b = (b + (((key[offset + 5]  & BYTE_MASK) <<  8) & INT_MASK)) & INT_MASK;
      b = (b + (((key[offset + 6]  & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
      b = (b + (((key[offset + 7]  & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
      c = (c + (key[offset + 8]    & BYTE_MASK)) & INT_MASK;
      c = (c + (((key[offset + 9]  & BYTE_MASK) <<  8) & INT_MASK)) & INT_MASK;
      c = (c + (((key[offset + 10] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
      c = (c + (((key[offset + 11] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
      
      /*
       * mix -- mix 3 32-bit values reversibly.
       * This is reversible, so any information in (a,b,c) before mix() is
       * still in (a,b,c) after mix().
       * 
       * If four pairs of (a,b,c) inputs are run through mix(), or through
       * mix() in reverse, there are at least 32 bits of the output that
       * are sometimes the same for one pair and different for another pair.
       * 
       * This was tested for:
       * - pairs that differed by one bit, by two bits, in any combination
       *   of top bits of (a,b,c), or in any combination of bottom bits of
       *   (a,b,c).
       * - "differ" is defined as +, -, ^, or ~^.  For + and -, I transformed
       *   the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
       *    is commonly produced by subtraction) look like a single 1-bit
       *    difference.
       * - the base values were pseudorandom, all zero but one bit set, or
       *   all zero plus a counter that starts at zero.
       * 
       * Some k values for my "a-=c; a^=rot(c,k); c+=b;" arrangement that
       * satisfy this are
       *     4  6  8 16 19  4
       *     9 15  3 18 27 15
       *    14  9  3  7 17  3
       * Well, "9 15 3 18 27 15" didn't quite get 32 bits diffing for 
       * "differ" defined as + with a one-bit base and a two-bit delta.  I
       * used http://burtleburtle.net/bob/hash/avalanche.html to choose
       * the operations, constants, and arrangements of the variables.
       * 
       * This does not achieve avalanche.  There are input bits of (a,b,c)
       * that fail to affect some output bits of (a,b,c), especially of a.
       * The most thoroughly mixed value is c, but it doesn't really even
       * achieve avalanche in c.
       * 
       * This allows some parallelism.  Read-after-writes are good at doubling
       * the number of bits affected, so the goal of mixing pulls in the
       * opposite direction as the goal of parallelism.  I did what I could.
       * Rotates seem to cost as much as shifts on every machine I could lay
       * my hands on, and rotates are much kinder to the top and bottom bits,
       * so I used rotates.
       *
       * #define mix(a,b,c) \
       * { \
       *   a -= c;  a ^= rot(c, 4);  c += b; \
       *   b -= a;  b ^= rot(a, 6);  a += c; \
       *   c -= b;  c ^= rot(b, 8);  b += a; \
       *   a -= c;  a ^= rot(c,16);  c += b; \
       *   b -= a;  b ^= rot(a,19);  a += c; \
       *   c -= b;  c ^= rot(b, 4);  b += a; \
       * }
       * 
       * mix(a,b,c);
       */
      a = (a - c) & INT_MASK;  a ^= rot(c, 4);  c = (c + b) & INT_MASK;
      b = (b - a) & INT_MASK;  b ^= rot(a, 6);  a = (a + c) & INT_MASK;
      c = (c - b) & INT_MASK;  c ^= rot(b, 8);  b = (b + a) & INT_MASK;
      a = (a - c) & INT_MASK;  a ^= rot(c,16);  c = (c + b) & INT_MASK;
      b = (b - a) & INT_MASK;  b ^= rot(a,19);  a = (a + c) & INT_MASK;
      c = (c - b) & INT_MASK;  c ^= rot(b, 4);  b = (b + a) & INT_MASK;
    }

    //-------------------------------- last block: affect all 32 bits of (c)
    switch (length) {                   // all the case statements fall through
    case 12:
      c = (c + (((key[offset + 11] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
    case 11:
      c = (c + (((key[offset + 10] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
    case 10:
      c = (c + (((key[offset + 9]  & BYTE_MASK) <<  8) & INT_MASK)) & INT_MASK;
    case  9:
      c = (c + (key[offset + 8]    & BYTE_MASK)) & INT_MASK;
    case  8:
      b = (b + (((key[offset + 7]  & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
    case  7:
      b = (b + (((key[offset + 6]  & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
    case  6:
      b = (b + (((key[offset + 5]  & BYTE_MASK) <<  8) & INT_MASK)) & INT_MASK;
    case  5:
      b = (b + (key[offset + 4]    & BYTE_MASK)) & INT_MASK;
    case  4:
      a = (a + (((key[offset + 3]  & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
    case  3:
      a = (a + (((key[offset + 2]  & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
    case  2:
      a = (a + (((key[offset + 1]  & BYTE_MASK) <<  8) & INT_MASK)) & INT_MASK;
    case  1:
      a = (a + (key[offset + 0]    & BYTE_MASK)) & INT_MASK;
      break;
    case  0:
      return Long.valueOf(c & INT_MASK).intValue();
    }
    /*
     * final -- final mixing of 3 32-bit values (a,b,c) into c
     * 
     * Pairs of (a,b,c) values differing in only a few bits will usually
     * produce values of c that look totally different.  This was tested for
     * - pairs that differed by one bit, by two bits, in any combination
     *   of top bits of (a,b,c), or in any combination of bottom bits of
     *   (a,b,c).
     * 
     * - "differ" is defined as +, -, ^, or ~^.  For + and -, I transformed
     *   the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
     *   is commonly produced by subtraction) look like a single 1-bit
     *   difference.
     * 
     * - the base values were pseudorandom, all zero but one bit set, or
     *   all zero plus a counter that starts at zero.
     * 
     * These constants passed:
     *   14 11 25 16 4 14 24
     *   12 14 25 16 4 14 24
     * and these came close:
     *    4  8 15 26 3 22 24
     *   10  8 15 26 3 22 24
     *   11  8 15 26 3 22 24
     * 
     * #define final(a,b,c) \
     * { 
     *   c ^= b; c -= rot(b,14); \
     *   a ^= c; a -= rot(c,11); \
     *   b ^= a; b -= rot(a,25); \
     *   c ^= b; c -= rot(b,16); \
     *   a ^= c; a -= rot(c,4);  \
     *   b ^= a; b -= rot(a,14); \
     *   c ^= b; c -= rot(b,24); \
     * }
     * 
     */
    c ^= b; c = (c - rot(b,14)) & INT_MASK;
    a ^= c; a = (a - rot(c,11)) & INT_MASK;
    b ^= a; b = (b - rot(a,25)) & INT_MASK;
    c ^= b; c = (c - rot(b,16)) & INT_MASK;
    a ^= c; a = (a - rot(c,4))  & INT_MASK;
    b ^= a; b = (b - rot(a,14)) & INT_MASK;
    c ^= b; c = (c - rot(b,24)) & INT_MASK;

    return Long.valueOf(c & INT_MASK).intValue();
  }
}
1	/**
2	* Copyright 2007 The Apache Software Foundation
3	*
4	* Licensed to the Apache Software Foundation (ASF) under one
5	* or more contributor license agreements. See the NOTICE file
6	* distributed with this work for additional information
7	* regarding copyright ownership. The ASF licenses this file
8	* to you under the Apache License, Version 2.0 (the
9	* "License"); you may not use this file except in compliance
10	* with the License. You may obtain a copy of the License at
11	*
12	* http://www.apache.org/licenses/LICENSE-2.0
13	*
14	* Unless required by applicable law or agreed to in writing, software
15	* distributed under the License is distributed on an "AS IS" BASIS,
16	* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
17	* See the License for the specific language governing permissions and
18	* limitations under the License.
19	*/
20
21	package org.apache.hadoop.hbase.util;
22
23	/**
24	* lookup3.c, by Bob Jenkins, May 2006, Public Domain.
25	* <a href="http://burtleburtle.net/bob/c/lookup3.c">lookup3.c</a>
26	*
27	* You can use this free for any purpose. It's in the public domain.
28	* It has no warranty.
29	*
30	* Produces 32-bit hash for hash table lookup.
31	*/
32	public class JenkinsHash {
33	private static long INT_MASK = 0x00000000ffffffffL;
34	private static long BYTE_MASK = 0x00000000000000ffL;
35
36	private static long rot(long val, int pos) {
37	return Long.valueOf(Integer.rotateLeft(
38	Long.valueOf(val & INT_MASK).intValue(), pos)).longValue() & INT_MASK;
39	}
40
41	/**
42	* Alternate form for hashing an entire byte array
43	*
44	* @param bytes
45	* @param initval
46	* @return hash value
47	*/
48	public static int hash(byte[] bytes, int initval) {
49	return hash(bytes, bytes.length, initval);
50	}
51
52	/**
53	* taken from hashlittle() -- hash a variable-length key into a 32-bit value
54	*
55	* @param key the key (the unaligned variable-length array of bytes)
56	* @param nbytes number of bytes to include in hash
57	* @param initval can be any integer value
58	* @return a 32-bit value. Every bit of the key affects every bit of the
59	* return value. Two keys differing by one or two bits will have totally
60	* different hash values.
61	*
62	* The best hash table sizes are powers of 2. There is no need to do mod a
63	* prime (mod is sooo slow!). If you need less than 32 bits, use a bitmask.
64	* For example, if you need only 10 bits, do h = (h & hashmask(10));
65	* In which case, the hash table should have hashsize(10) elements.
66	*
67	* If you are hashing n strings byte[][] k, do it like this:
68	* for (int i = 0, h = 0; i < n; ++i) h = hash( k[i], h);
69	*
70	* By Bob Jenkins, 2006. bob_jenkins@burtleburtle.net. You may use this
71	* code any way you wish, private, educational, or commercial. It's free.
72	*
73	* Use for hash table lookup, or anything where one collision in 2^^32 is
74	* acceptable. Do NOT use for cryptographic purposes.
75	*/
76	public static int hash(byte[] key, int nbytes, int initval) {
77	int length = nbytes;
78	long a, b, c; // We use longs because we don't have unsigned ints
79	a = b = c = (0x00000000deadbeefL + length + initval) & INT_MASK;
80	int offset = 0;
81	for (; length > 12; offset += 12, length -= 12) {
82	a = (a + (key[offset + 0] & BYTE_MASK)) & INT_MASK;
83	a = (a + (((key[offset + 1] & BYTE_MASK) << 8) & INT_MASK)) & INT_MASK;
84	a = (a + (((key[offset + 2] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
85	a = (a + (((key[offset + 3] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
86	b = (b + (key[offset + 4] & BYTE_MASK)) & INT_MASK;
87	b = (b + (((key[offset + 5] & BYTE_MASK) << 8) & INT_MASK)) & INT_MASK;
88	b = (b + (((key[offset + 6] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
89	b = (b + (((key[offset + 7] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
90	c = (c + (key[offset + 8] & BYTE_MASK)) & INT_MASK;
91	c = (c + (((key[offset + 9] & BYTE_MASK) << 8) & INT_MASK)) & INT_MASK;
92	c = (c + (((key[offset + 10] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
93	c = (c + (((key[offset + 11] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
94
95	/*
96	* mix -- mix 3 32-bit values reversibly.
97	* This is reversible, so any information in (a,b,c) before mix() is
98	* still in (a,b,c) after mix().
99	*
100	* If four pairs of (a,b,c) inputs are run through mix(), or through
101	* mix() in reverse, there are at least 32 bits of the output that
102	* are sometimes the same for one pair and different for another pair.
103	*
104	* This was tested for:
105	* - pairs that differed by one bit, by two bits, in any combination
106	* of top bits of (a,b,c), or in any combination of bottom bits of
107	* (a,b,c).
108	* - "differ" is defined as +, -, ^, or ~^. For + and -, I transformed
109	* the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
110	* is commonly produced by subtraction) look like a single 1-bit
111	* difference.
112	* - the base values were pseudorandom, all zero but one bit set, or
113	* all zero plus a counter that starts at zero.
114	*
115	* Some k values for my "a-=c; a^=rot(c,k); c+=b;" arrangement that
116	* satisfy this are
117	* 4 6 8 16 19 4
118	* 9 15 3 18 27 15
119	* 14 9 3 7 17 3
120	* Well, "9 15 3 18 27 15" didn't quite get 32 bits diffing for
121	* "differ" defined as + with a one-bit base and a two-bit delta. I
122	* used http://burtleburtle.net/bob/hash/avalanche.html to choose
123	* the operations, constants, and arrangements of the variables.
124	*
125	* This does not achieve avalanche. There are input bits of (a,b,c)
126	* that fail to affect some output bits of (a,b,c), especially of a.
127	* The most thoroughly mixed value is c, but it doesn't really even
128	* achieve avalanche in c.
129	*
130	* This allows some parallelism. Read-after-writes are good at doubling
131	* the number of bits affected, so the goal of mixing pulls in the
132	* opposite direction as the goal of parallelism. I did what I could.
133	* Rotates seem to cost as much as shifts on every machine I could lay
134	* my hands on, and rotates are much kinder to the top and bottom bits,
135	* so I used rotates.
136	*
137	* #define mix(a,b,c) \
138	* { \
139	* a -= c; a ^= rot(c, 4); c += b; \
140	* b -= a; b ^= rot(a, 6); a += c; \
141	* c -= b; c ^= rot(b, 8); b += a; \
142	* a -= c; a ^= rot(c,16); c += b; \
143	* b -= a; b ^= rot(a,19); a += c; \
144	* c -= b; c ^= rot(b, 4); b += a; \
145	* }
146	*
147	* mix(a,b,c);
148	*/
149	a = (a - c) & INT_MASK; a ^= rot(c, 4); c = (c + b) & INT_MASK;
150	b = (b - a) & INT_MASK; b ^= rot(a, 6); a = (a + c) & INT_MASK;
151	c = (c - b) & INT_MASK; c ^= rot(b, 8); b = (b + a) & INT_MASK;
152	a = (a - c) & INT_MASK; a ^= rot(c,16); c = (c + b) & INT_MASK;
153	b = (b - a) & INT_MASK; b ^= rot(a,19); a = (a + c) & INT_MASK;
154	c = (c - b) & INT_MASK; c ^= rot(b, 4); b = (b + a) & INT_MASK;
155	}
156
157	//-------------------------------- last block: affect all 32 bits of (c)
158	switch (length) { // all the case statements fall through
159	case 12:
160	c = (c + (((key[offset + 11] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
161	case 11:
162	c = (c + (((key[offset + 10] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
163	case 10:
164	c = (c + (((key[offset + 9] & BYTE_MASK) << 8) & INT_MASK)) & INT_MASK;
165	case 9:
166	c = (c + (key[offset + 8] & BYTE_MASK)) & INT_MASK;
167	case 8:
168	b = (b + (((key[offset + 7] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
169	case 7:
170	b = (b + (((key[offset + 6] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
171	case 6:
172	b = (b + (((key[offset + 5] & BYTE_MASK) << 8) & INT_MASK)) & INT_MASK;
173	case 5:
174	b = (b + (key[offset + 4] & BYTE_MASK)) & INT_MASK;
175	case 4:
176	a = (a + (((key[offset + 3] & BYTE_MASK) << 24) & INT_MASK)) & INT_MASK;
177	case 3:
178	a = (a + (((key[offset + 2] & BYTE_MASK) << 16) & INT_MASK)) & INT_MASK;
179	case 2:
180	a = (a + (((key[offset + 1] & BYTE_MASK) << 8) & INT_MASK)) & INT_MASK;
181	case 1:
182	a = (a + (key[offset + 0] & BYTE_MASK)) & INT_MASK;
183	break;
184	case 0:
185	return Long.valueOf(c & INT_MASK).intValue();
186	}
187	/*
188	* final -- final mixing of 3 32-bit values (a,b,c) into c
189	*
190	* Pairs of (a,b,c) values differing in only a few bits will usually
191	* produce values of c that look totally different. This was tested for
192	* - pairs that differed by one bit, by two bits, in any combination
193	* of top bits of (a,b,c), or in any combination of bottom bits of
194	* (a,b,c).
195	*
196	* - "differ" is defined as +, -, ^, or ~^. For + and -, I transformed
197	* the output delta to a Gray code (a^(a>>1)) so a string of 1's (as
198	* is commonly produced by subtraction) look like a single 1-bit
199	* difference.
200	*
201	* - the base values were pseudorandom, all zero but one bit set, or
202	* all zero plus a counter that starts at zero.
203	*
204	* These constants passed:
205	* 14 11 25 16 4 14 24
206	* 12 14 25 16 4 14 24
207	* and these came close:
208	* 4 8 15 26 3 22 24
209	* 10 8 15 26 3 22 24
210	* 11 8 15 26 3 22 24
211	*
212	* #define final(a,b,c) \
213	* {
214	* c ^= b; c -= rot(b,14); \
215	* a ^= c; a -= rot(c,11); \
216	* b ^= a; b -= rot(a,25); \
217	* c ^= b; c -= rot(b,16); \
218	* a ^= c; a -= rot(c,4); \
219	* b ^= a; b -= rot(a,14); \
220	* c ^= b; c -= rot(b,24); \
221	* }
222	*
223	*/
224	c ^= b; c = (c - rot(b,14)) & INT_MASK;
225	a ^= c; a = (a - rot(c,11)) & INT_MASK;
226	b ^= a; b = (b - rot(a,25)) & INT_MASK;
227	c ^= b; c = (c - rot(b,16)) & INT_MASK;
228	a ^= c; a = (a - rot(c,4)) & INT_MASK;
229	b ^= a; b = (b - rot(a,14)) & INT_MASK;
230	c ^= b; c = (c - rot(b,24)) & INT_MASK;
231
232	return Long.valueOf(c & INT_MASK).intValue();
233	}
234	}