001 /*
002 * Copyright (C) 2011 The Guava Authors
003 *
004 * Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except
005 * in compliance with the License. You may obtain a copy of the License at
006 *
007 * http://www.apache.org/licenses/LICENSE-2.0
008 *
009 * Unless required by applicable law or agreed to in writing, software distributed under the License
010 * is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express
011 * or implied. See the License for the specific language governing permissions and limitations under
012 * the License.
013 */
014
015 package com.google.common.hash;
016
017 import com.google.common.annotations.Beta;
018 import com.google.common.primitives.Ints;
019
020 import java.nio.charset.Charset;
021
022 /**
023 * A hash function is a collision-averse pure function that maps an arbitrary block of
024 * data to a number called a <i>hash code</i>.
025 *
026 * <h3>Definition</h3>
027 *
028 * <p>Unpacking this definition:
029 *
030 * <ul>
031 * <li><b>block of data:</b> the input for a hash function is always, in concept, an
032 * ordered byte array. This hashing API accepts an arbitrary sequence of byte and
033 * multibyte values (via {@link Hasher}), but this is merely a convenience; these are
034 * always translated into raw byte sequences under the covers.
035 *
036 * <li><b>hash code:</b> each hash function always yields hash codes of the same fixed bit
037 * length (given by {@link #bits}). For example, {@link Hashing#sha1} produces a
038 * 160-bit number, while {@link Hashing#murmur3_32()} yields only 32 bits. Because a
039 * {@code long} value is clearly insufficient to hold all hash code values, this API
040 * represents a hash code as an instance of {@link HashCode}.
041 *
042 * <li><b>pure function:</b> the value produced must depend only on the input bytes, in
043 * the order they appear. Input data is never modified.
044 *
045 * <li><b>collision-averse:</b> while it can't be helped that a hash function will
046 * sometimes produce the same hash code for distinct inputs (a "collision"), every
047 * hash function strives to <i>some</i> degree to make this unlikely. (Without this
048 * condition, a function that always returns zero could be called a hash function. It
049 * is not.)
050 * </ul>
051 *
052 * <p>Summarizing the last two points: "equal yield equal <i>always</i>; unequal yield
053 * unequal <i>often</i>." This is the most important characteristic of all hash functions.
054 *
055 * <h3>Desirable properties</h3>
056 *
057 * <p>A high-quality hash function strives for some subset of the following virtues:
058 *
059 * <ul>
060 * <li><b>collision-resistant:</b> while the definition above requires making at least
061 * <i>some</i> token attempt, one measure of the quality of a hash function is <i>how
062 * well</i> it succeeds at this goal. Important note: it may be easy to achieve the
063 * theoretical minimum collision rate when using completely <i>random</i> sample
064 * input. The true test of a hash function is how it performs on representative
065 * real-world data, which tends to contain many hidden patterns and clumps. The goal
066 * of a good hash function is to stamp these patterns out as thoroughly as possible.
067 *
068 * <li><b>bit-dispersing:</b> masking out any <i>single bit</i> from a hash code should
069 * yield only the expected <i>twofold</i> increase to all collision rates. Informally,
070 * the "information" in the hash code should be as evenly "spread out" through the
071 * hash code's bits as possible. The result is that, for example, when choosing a
072 * bucket in a hash table of size 2^8, <i>any</i> eight bits could be consistently
073 * used.
074 *
075 * <li><b>cryptographic:</b> certain hash functions such as {@link Hashing#sha512} are
076 * designed to make it as infeasible as possible to reverse-engineer the input that
077 * produced a given hash code, or even to discover <i>any</i> two distinct inputs that
078 * yield the same result. These are called <i>cryptographic hash functions</i>. But,
079 * whenever it is learned that either of these feats has become computationally
080 * feasible, the function is deemed "broken" and should no longer be used for secure
081 * purposes. (This is the likely eventual fate of <i>all</i> cryptographic hashes.)
082 *
083 * <li><b>fast:</b> perhaps self-explanatory, but often the most important consideration.
084 * We have published <a href="#noWeHaventYet">microbenchmark results</a> for many
085 * common hash functions.
086 * </ul>
087 *
088 * <h3>Providing input to a hash function</h3>
089 *
090 * <p>The primary way to provide the data that your hash function should act on is via a
091 * {@link Hasher}. Obtain a new hasher from the hash function using {@link #newHasher},
092 * "push" the relevant data into it using methods like {@link Hasher#putBytes(byte[])},
093 * and finally ask for the {@code HashCode} when finished using {@link Hasher#hash}. (See
094 * an {@linkplain #newHasher example} of this.)
095 *
096 * <p>If all you want to hash is a single byte array, string or {@code long} value, there
097 * are convenient shortcut methods defined directly on {@link HashFunction} to make this
098 * easier.
099 *
100 * <p>Hasher accepts primitive data types, but can also accept any Object of type {@code
101 * T} provided that you implement a {@link Funnel Funnel<T>} to specify how to "feed" data
102 * from that object into the function. (See {@linkplain Hasher#putObject an example} of
103 * this.)
104 *
105 * <p><b>Compatibility note:</b> Throughout this API, multibyte values are always
106 * interpreted in <i>little-endian</i> order. That is, hashing the byte array {@code
107 * {0x01, 0x02, 0x03, 0x04}} is equivalent to hashing the {@code int} value {@code
108 * 0x04030201}. If this isn't what you need, methods such as {@link Integer#reverseBytes}
109 * and {@link Ints#toByteArray} will help.
110 *
111 * <h3>Relationship to {@link Object#hashCode}</h3>
112 *
113 * <p>Java's baked-in concept of hash codes is constrained to 32 bits, and provides no
114 * separation between hash algorithms and the data they act on, so alternate hash
115 * algorithms can't be easily substituted. Also, implementations of {@code hashCode} tend
116 * to be poor-quality, in part because they end up depending on <i>other</i> existing
117 * poor-quality {@code hashCode} implementations, including those in many JDK classes.
118 *
119 * <p>{@code Object.hashCode} implementations tend to be very fast, but have weak
120 * collision prevention and <i>no</i> expectation of bit dispersion. This leaves them
121 * perfectly suitable for use in hash tables, because extra collisions cause only a slight
122 * performance hit, while poor bit dispersion is easily corrected using a secondary hash
123 * function (which all reasonable hash table implementations in Java use). For the many
124 * uses of hash functions beyond data structures, however, {@code Object.hashCode} almost
125 * always falls short -- hence this library.
126 *
127 * @author Kevin Bourrillion
128 * @since 11.0
129 */
130 @Beta
131 public interface HashFunction {
132 /**
133 * Begins a new hash code computation by returning an initialized, stateful {@code
134 * Hasher} instance that is ready to receive data. Example: <pre> {@code
135 *
136 * HashFunction hf = Hashing.md5();
137 * HashCode hc = hf.newHasher()
138 * .putLong(id)
139 * .putString(name)
140 * .hash();}</pre>
141 */
142 Hasher newHasher();
143
144 /**
145 * Begins a new hash code computation as {@link #newHasher()}, but provides a hint of the
146 * expected size of the input (in bytes). This is only important for non-streaming hash
147 * functions (hash functions that need to buffer their whole input before processing any
148 * of it).
149 */
150 Hasher newHasher(int expectedInputSize);
151
152 /**
153 * Shortcut for {@code newHasher().putInt(input).hash()}; returns the hash code for the given
154 * {@code int} value, interpreted in little-endian byte order. The implementation <i>might</i>
155 * perform better than its longhand equivalent, but should not perform worse.
156 *
157 * @since 12.0
158 */
159 HashCode hashInt(int input);
160
161 /**
162 * Shortcut for {@code newHasher().putLong(input).hash()}; returns the hash code for the
163 * given {@code long} value, interpreted in little-endian byte order. The implementation
164 * <i>might</i> perform better than its longhand equivalent, but should not perform worse.
165 */
166 HashCode hashLong(long input);
167
168 /**
169 * Shortcut for {@code newHasher().putBytes(input).hash()}. The implementation
170 * <i>might</i> perform better than its longhand equivalent, but should not perform
171 * worse.
172 */
173 HashCode hashBytes(byte[] input);
174
175 /**
176 * Shortcut for {@code newHasher().putBytes(input, off, len).hash()}. The implementation
177 * <i>might</i> perform better than its longhand equivalent, but should not perform
178 * worse.
179 *
180 * @throws IndexOutOfBoundsException if {@code off < 0} or {@code off + len > bytes.length}
181 * or {@code len < 0}
182 */
183 HashCode hashBytes(byte[] input, int off, int len);
184
185 /**
186 * Shortcut for {@code newHasher().putString(input).hash()}. The implementation <i>might</i>
187 * perform better than its longhand equivalent, but should not perform worse. Note that no
188 * character encoding is performed; the low byte and high byte of each character are hashed
189 * directly (in that order). This is equivalent to using
190 * {@code hashString(input, Charsets.UTF_16LE)}.
191 */
192 HashCode hashString(CharSequence input);
193
194 /**
195 * Shortcut for {@code newHasher().putString(input, charset).hash()}. Characters are encoded
196 * using the given {@link Charset}. The implementation <i>might</i> perform better than its
197 * longhand equivalent, but should not perform worse.
198 */
199 HashCode hashString(CharSequence input, Charset charset);
200
201 /**
202 * Returns the number of bits (a multiple of 32) that each hash code produced by this
203 * hash function has.
204 */
205 int bits();
206 }