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.ByteBuffer;
020import java.nio.charset.Charset;
021import org.checkerframework.checker.nullness.qual.Nullable;
022
023/**
024 * A {@link PrimitiveSink} that can compute a hash code after reading the input. Each hasher should
025 * translate all multibyte values ({@link #putInt(int)}, {@link #putLong(long)}, etc) to bytes in
026 * little-endian order.
027 *
028 * <p><b>Warning:</b> The result of calling any methods after calling {@link #hash} is undefined.
029 *
030 * <p><b>Warning:</b> Using a specific character encoding when hashing a {@link CharSequence} with
031 * {@link #putString(CharSequence, Charset)} is generally only useful for cross-language
032 * compatibility (otherwise prefer {@link #putUnencodedChars}). However, the character encodings
033 * must be identical across languages. Also beware that {@link Charset} definitions may occasionally
034 * change between Java releases.
035 *
036 * <p><b>Warning:</b> Chunks of data that are put into the {@link Hasher} are not delimited. The
037 * resulting {@link HashCode} is dependent only on the bytes inserted, and the order in which they
038 * were inserted, not how those bytes were chunked into discrete put() operations. For example, the
039 * following three expressions all generate colliding hash codes:
040 *
041 * <pre>{@code
042 * newHasher().putByte(b1).putByte(b2).putByte(b3).hash()
043 * newHasher().putByte(b1).putBytes(new byte[] { b2, b3 }).hash()
044 * newHasher().putBytes(new byte[] { b1, b2, b3 }).hash()
045 * }</pre>
046 *
047 * <p>If you wish to avoid this, you should either prepend or append the size of each chunk. Keep in
048 * mind that when dealing with char sequences, the encoded form of two concatenated char sequences
049 * is not equivalent to the concatenation of their encoded form. Therefore, {@link
050 * #putString(CharSequence, Charset)} should only be used consistently with <i>complete</i>
051 * sequences and not broken into chunks.
052 *
053 * @author Kevin Bourrillion
054 * @since 11.0
055 */
056@Beta
057public interface Hasher extends PrimitiveSink {
058  @CanIgnoreReturnValue
059  @Override
060  Hasher putByte(byte b);
061
062  @CanIgnoreReturnValue
063  @Override
064  Hasher putBytes(byte[] bytes);
065
066  @CanIgnoreReturnValue
067  @Override
068  Hasher putBytes(byte[] bytes, int off, int len);
069
070  @CanIgnoreReturnValue
071  @Override
072  Hasher putBytes(ByteBuffer bytes);
073
074  @CanIgnoreReturnValue
075  @Override
076  Hasher putShort(short s);
077
078  @CanIgnoreReturnValue
079  @Override
080  Hasher putInt(int i);
081
082  @CanIgnoreReturnValue
083  @Override
084  Hasher putLong(long l);
085
086  /** Equivalent to {@code putInt(Float.floatToRawIntBits(f))}. */
087  @CanIgnoreReturnValue
088  @Override
089  Hasher putFloat(float f);
090
091  /** Equivalent to {@code putLong(Double.doubleToRawLongBits(d))}. */
092  @CanIgnoreReturnValue
093  @Override
094  Hasher putDouble(double d);
095
096  /** Equivalent to {@code putByte(b ? (byte) 1 : (byte) 0)}. */
097  @CanIgnoreReturnValue
098  @Override
099  Hasher putBoolean(boolean b);
100
101  @CanIgnoreReturnValue
102  @Override
103  Hasher putChar(char c);
104
105  /**
106   * Equivalent to processing each {@code char} value in the {@code CharSequence}, in order. In
107   * other words, no character encoding is performed; the low byte and high byte of each {@code
108   * char} are hashed directly (in that order). The input must not be updated while this method is
109   * in progress.
110   *
111   * <p><b>Warning:</b> This method will produce different output than most other languages do when
112   * running the same hash function on the equivalent input. For cross-language compatibility, use
113   * {@link #putString}, usually with a charset of UTF-8. For other use cases, use {@code
114   * putUnencodedChars}.
115   *
116   * @since 15.0 (since 11.0 as putString(CharSequence)).
117   */
118  @CanIgnoreReturnValue
119  @Override
120  Hasher putUnencodedChars(CharSequence charSequence);
121
122  /**
123   * Equivalent to {@code putBytes(charSequence.toString().getBytes(charset))}.
124   *
125   * <p><b>Warning:</b> This method, which reencodes the input before hashing it, is useful only for
126   * cross-language compatibility. For other use cases, prefer {@link #putUnencodedChars}, which is
127   * faster, produces the same output across Java releases, and hashes every {@code char} in the
128   * input, even if some are invalid.
129   */
130  @CanIgnoreReturnValue
131  @Override
132  Hasher putString(CharSequence charSequence, Charset charset);
133
134  /** A simple convenience for {@code funnel.funnel(object, this)}. */
135  @CanIgnoreReturnValue
136  <T extends @Nullable Object> Hasher putObject(
137      @ParametricNullness T instance, Funnel<? super T> funnel);
138
139  /**
140   * Computes a hash code based on the data that have been provided to this hasher. The result is
141   * unspecified if this method is called more than once on the same instance.
142   */
143  HashCode hash();
144
145  /**
146   * {@inheritDoc}
147   *
148   * @deprecated This returns {@link Object#hashCode()}; you almost certainly mean to call {@code
149   *     hash().asInt()}.
150   */
151  @Override
152  @Deprecated
153  int hashCode();
154}