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
015package com.google.common.hash;
016
017import com.google.common.annotations.Beta;
018import com.google.errorprone.annotations.CanIgnoreReturnValue;
019import java.nio.charset.Charset;
020
021/**
022 * A {@link PrimitiveSink} that can compute a hash code after reading the input. Each hasher should
023 * translate all multibyte values ({@link #putInt(int)}, {@link #putLong(long)}, etc) to bytes in
024 * little-endian order.
025 *
026 * <p><b>Warning:</b> The result of calling any methods after calling {@link #hash} is undefined.
027 *
028 * <p><b>Warning:</b> Using a specific character encoding when hashing a {@link CharSequence} with
029 * {@link #putString(CharSequence, Charset)} is generally only useful for cross-language
030 * compatibility (otherwise prefer {@link #putUnencodedChars}). However, the character encodings
031 * must be identical across languages. Also beware that {@link Charset} definitions may occasionally
032 * change between Java releases.
033 *
034 * <p><b>Warning:</b> Chunks of data that are put into the {@link Hasher} are not delimited. The
035 * resulting {@link HashCode} is dependent only on the bytes inserted, and the order in which they
036 * were inserted, not how those bytes were chunked into discrete put() operations. For example, the
037 * following three expressions all generate colliding hash codes: <pre>   {@code
038 *
039 *   newHasher().putByte(b1).putByte(b2).putByte(b3).hash()
040 *   newHasher().putByte(b1).putBytes(new byte[] { b2, b3 }).hash()
041 *   newHasher().putBytes(new byte[] { b1, b2, b3 }).hash()}</pre>
042 *
043 * <p>If you wish to avoid this, you should either prepend or append the size of each chunk. Keep in
044 * mind that when dealing with char sequences, the encoded form of two concatenated char sequences
045 * is not equivalent to the concatenation of their encoded form. Therefore,
046 * {@link #putString(CharSequence, Charset)} should only be used consistently with <i>complete</i>
047 * sequences and not broken into chunks.
048 *
049 * @author Kevin Bourrillion
050 * @since 11.0
051 */
052@Beta
053@CanIgnoreReturnValue
054public interface Hasher extends PrimitiveSink {
055  @Override
056  Hasher putByte(byte b);
057
058  @Override
059  Hasher putBytes(byte[] bytes);
060
061  @Override
062  Hasher putBytes(byte[] bytes, int off, int len);
063
064  @Override
065  Hasher putShort(short s);
066
067  @Override
068  Hasher putInt(int i);
069
070  @Override
071  Hasher putLong(long l);
072
073  /**
074   * Equivalent to {@code putInt(Float.floatToRawIntBits(f))}.
075   */
076  @Override
077  Hasher putFloat(float f);
078
079  /**
080   * Equivalent to {@code putLong(Double.doubleToRawLongBits(d))}.
081   */
082  @Override
083  Hasher putDouble(double d);
084
085  /**
086   * Equivalent to {@code putByte(b ? (byte) 1 : (byte) 0)}.
087   */
088  @Override
089  Hasher putBoolean(boolean b);
090
091  @Override
092  Hasher putChar(char c);
093
094  /**
095   * Equivalent to processing each {@code char} value in the {@code CharSequence}, in order. In
096   * other words, no character encoding is performed; the low byte and high byte of each {@code
097   * char} are hashed directly (in that order). The input must not be updated while this method is
098   * in progress.
099   *
100   * <p><b>Warning:</b> This method will produce different output than most other languages do when
101   * running the same hash function on the equivalent input. For cross-language compatibility, use
102   * {@link #putString}, usually with a charset of UTF-8. For other use cases, use {@code
103   * putUnencodedChars}.
104   *
105   * @since 15.0 (since 11.0 as putString(CharSequence)).
106   */
107  @Override
108  Hasher putUnencodedChars(CharSequence charSequence);
109
110  /**
111   * Equivalent to {@code putBytes(charSequence.toString().getBytes(charset))}.
112   *
113   * <p><b>Warning:</b> This method, which reencodes the input before hashing it, is useful only for
114   * cross-language compatibility. For other use cases, prefer {@link #putUnencodedChars}, which is
115   * faster, produces the same output across Java releases, and hashes every {@code char} in the
116   * input, even if some are invalid.
117   */
118  @Override
119  Hasher putString(CharSequence charSequence, Charset charset);
120
121  /**
122   * A simple convenience for {@code funnel.funnel(object, this)}.
123   */
124  <T> Hasher putObject(T instance, Funnel<? super T> funnel);
125
126  /**
127   * Computes a hash code based on the data that have been provided to this hasher. The result is
128   * unspecified if this method is called more than once on the same instance.
129   */
130  HashCode hash();
131
132  /**
133   * {@inheritDoc}
134   *
135   * @deprecated This returns {@link Object#hashCode()}; you almost certainly mean to call
136   *     {@code hash().asInt()}.
137   */
138  @Override
139  @Deprecated
140  int hashCode();
141}