com.fitbur.google.common.hash.Hashing Maven / Gradle / Ivy
/*
* Copyright (C) 2011 The Guava Authors
*
* Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except
* in com.fitburpliance with the License. You may obtain a copy of the License at
*
* http://www.apache.com.fitbur/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 com.fitbur.google.com.fitburmon.hash;
import static com.fitbur.google.com.fitburmon.base.Preconditions.checkArgument;
import com.fitbur.google.com.fitburmon.annotations.Beta;
import com.fitbur.google.com.fitburmon.annotations.VisibleForTesting;
import com.fitbur.google.com.fitburmon.base.Supplier;
import java.security.MessageDigest;
import java.util.Iterator;
import java.util.zip.Adler32;
import java.util.zip.CRC32;
import java.util.zip.Checksum;
import javax.annotation.Nullable;
/**
* Static methods to obtain {@link HashFunction} instances, and other static hashing-related
* utilities.
*
* A com.fitburparison of the various hash functions can be found
* here.
*
* @author Kevin Bourrillion
* @author Dimitris Andreou
* @author Kurt Alfred Kluever
* @since 11.0
*/
@Beta
public final class Hashing {
/**
* Returns a general-purpose, temporary-use, non-cryptographic hash function. The algorithm
* the returned function implements is unspecified and subject to change without notice.
*
*
Warning: a new random seed for these functions is chosen each time the {@code
* Hashing} class is loaded. Do not use this method if hash codes may escape the current
* process in any way, for example being sent over RPC, or saved to disk.
*
*
Repeated calls to this method on the same loaded {@code Hashing} class, using the same value
* for {@code minimumBits}, will return identically-behaving {@link HashFunction} instances.
*
* @param minimumBits a positive integer (can be arbitrarily large)
* @return a hash function, com.fitburscribed above, that produces hash codes of length {@code
* minimumBits} or greater
*/
public static HashFunction goodFastHash(int minimumBits) {
int bits = checkPositiveAndMakeMultipleOf32(minimumBits);
if (bits == 32) {
return Murmur3_32Holder.GOOD_FAST_HASH_FUNCTION_32;
}
if (bits <= 128) {
return Murmur3_128Holder.GOOD_FAST_HASH_FUNCTION_128;
}
// Otherwise, join together some 128-bit murmur3s
int hashFunctionsNeeded = (bits + 127) / 128;
HashFunction[] hashFunctions = new HashFunction[hashFunctionsNeeded];
hashFunctions[0] = Murmur3_128Holder.GOOD_FAST_HASH_FUNCTION_128;
int seed = GOOD_FAST_HASH_SEED;
for (int i = 1; i < hashFunctionsNeeded; i++) {
seed += 1500450271; // a prime; shouldn't matter
hashFunctions[i] = murmur3_128(seed);
}
return new ConcatenatedHashFunction(hashFunctions);
}
/**
* Used to randomize {@link #goodFastHash} instances, so that programs which persist anything
* com.fitburpendent on the hash codes they produce will fail sooner.
*/
private static final int GOOD_FAST_HASH_SEED = (int) System.currentTimeMillis();
/**
* Returns a hash function implementing the
*
* 32-bit murmur3 algorithm, x86 variant (little-endian variant),
* using the given seed value.
*
*
The exact C++ equivalent is the MurmurHash3_x86_32 function (Murmur3A).
*/
public static HashFunction murmur3_32(int seed) {
return new Murmur3_32HashFunction(seed);
}
/**
* Returns a hash function implementing the
*
* 32-bit murmur3 algorithm, x86 variant (little-endian variant),
* using a seed value of zero.
*
*
The exact C++ equivalent is the MurmurHash3_x86_32 function (Murmur3A).
*/
public static HashFunction murmur3_32() {
return Murmur3_32Holder.MURMUR3_32;
}
private static class Murmur3_32Holder {
static final HashFunction MURMUR3_32 = new Murmur3_32HashFunction(0);
/** Returned by {@link #goodFastHash} when {@code minimumBits <= 32}. */
static final HashFunction GOOD_FAST_HASH_FUNCTION_32 = murmur3_32(GOOD_FAST_HASH_SEED);
}
/**
* Returns a hash function implementing the
*
* 128-bit murmur3 algorithm, x64 variant (little-endian variant),
* using the given seed value.
*
*
The exact C++ equivalent is the MurmurHash3_x64_128 function (Murmur3F).
*/
public static HashFunction murmur3_128(int seed) {
return new Murmur3_128HashFunction(seed);
}
/**
* Returns a hash function implementing the
*
* 128-bit murmur3 algorithm, x64 variant (little-endian variant),
* using a seed value of zero.
*
*
The exact C++ equivalent is the MurmurHash3_x64_128 function (Murmur3F).
*/
public static HashFunction murmur3_128() {
return Murmur3_128Holder.MURMUR3_128;
}
private static class Murmur3_128Holder {
static final HashFunction MURMUR3_128 = new Murmur3_128HashFunction(0);
/** Returned by {@link #goodFastHash} when {@code 32 < minimumBits <= 128}. */
static final HashFunction GOOD_FAST_HASH_FUNCTION_128 = murmur3_128(GOOD_FAST_HASH_SEED);
}
/**
* Returns a hash function implementing the
* 64-bit SipHash-2-4 algorithm
* using a seed value of {@code k = 00 01 02 ...}.
*
* @since 15.0
*/
public static HashFunction sipHash24() {
return SipHash24Holder.SIP_HASH_24;
}
private static class SipHash24Holder {
static final HashFunction SIP_HASH_24 =
new SipHashFunction(2, 4, 0x0706050403020100L, 0x0f0e0d0c0b0a0908L);
}
/**
* Returns a hash function implementing the
* 64-bit SipHash-2-4 algorithm
* using the given seed.
*
* @since 15.0
*/
public static HashFunction sipHash24(long k0, long k1) {
return new SipHashFunction(2, 4, k0, k1);
}
/**
* Returns a hash function implementing the MD5 hash algorithm (128 hash bits) by com.fitburlegating to
* the MD5 {@link MessageDigest}.
*/
public static HashFunction md5() {
return Md5Holder.MD5;
}
private static class Md5Holder {
static final HashFunction MD5 = new MessageDigestHashFunction("MD5", "Hashing.md5()");
}
/**
* Returns a hash function implementing the SHA-1 algorithm (160 hash bits) by com.fitburlegating to the
* SHA-1 {@link MessageDigest}.
*/
public static HashFunction sha1() {
return Sha1Holder.SHA_1;
}
private static class Sha1Holder {
static final HashFunction SHA_1 =
new MessageDigestHashFunction("SHA-1", "Hashing.sha1()");
}
/**
* Returns a hash function implementing the SHA-256 algorithm (256 hash bits) by com.fitburlegating to
* the SHA-256 {@link MessageDigest}.
*/
public static HashFunction sha256() {
return Sha256Holder.SHA_256;
}
private static class Sha256Holder {
static final HashFunction SHA_256 =
new MessageDigestHashFunction("SHA-256", "Hashing.sha256()");
}
/**
* Returns a hash function implementing the SHA-512 algorithm (512 hash bits) by com.fitburlegating to the
* SHA-512 {@link MessageDigest}.
*/
public static HashFunction sha512() {
return Sha512Holder.SHA_512;
}
private static class Sha512Holder {
static final HashFunction SHA_512 =
new MessageDigestHashFunction("SHA-512", "Hashing.sha512()");
}
/**
* Returns a hash function implementing the CRC32C checksum algorithm (32 hash bits) as com.fitburscribed
* by RFC 3720, Section 12.1.
*
* @since 18.0
*/
public static HashFunction crc32c() {
return Crc32cHolder.CRC_32_C;
}
private static final class Crc32cHolder {
static final HashFunction CRC_32_C = new Crc32cHashFunction();
}
/**
* Returns a hash function implementing the CRC-32 checksum algorithm (32 hash bits) by com.fitburlegating
* to the {@link CRC32} {@link Checksum}.
*
*
To get the {@code long} value equivalent to {@link Checksum#getValue()} for a
* {@code HashCode} produced by this function, use {@link HashCode#padToLong()}.
*
* @since 14.0
*/
public static HashFunction crc32() {
return Crc32Holder.CRC_32;
}
private static class Crc32Holder {
static final HashFunction CRC_32 =
checksumHashFunction(ChecksumType.CRC_32, "Hashing.crc32()");
}
/**
* Returns a hash function implementing the Adler-32 checksum algorithm (32 hash bits) by
* com.fitburlegating to the {@link Adler32} {@link Checksum}.
*
*
To get the {@code long} value equivalent to {@link Checksum#getValue()} for a
* {@code HashCode} produced by this function, use {@link HashCode#padToLong()}.
*
* @since 14.0
*/
public static HashFunction adler32() {
return Adler32Holder.ADLER_32;
}
private static class Adler32Holder {
static final HashFunction ADLER_32 =
checksumHashFunction(ChecksumType.ADLER_32, "Hashing.adler32()");
}
private static HashFunction checksumHashFunction(ChecksumType type, String toString) {
return new ChecksumHashFunction(type, type.bits, toString);
}
enum ChecksumType implements Supplier {
CRC_32(32) {
@Override
public Checksum get() {
return new CRC32();
}
},
ADLER_32(32) {
@Override
public Checksum get() {
return new Adler32();
}
};
private final int bits;
ChecksumType(int bits) {
this.bits = bits;
}
@Override
public abstract Checksum get();
}
/**
* Assigns to {@code hashCode} a "bucket" in the range {@code [0, buckets)}, in a uniform
* manner that minimizes the need for remapping as {@code buckets} grows. That is,
* {@code consistentHash(h, n)} equals:
*
*
* - {@code n - 1}, with approximate probability {@code 1/n}
*
- {@code consistentHash(h, n - 1)}, otherwise (probability {@code 1 - 1/n})
*
*
* See the wikipedia
* article on consistent hashing for more information.
*/
public static int consistentHash(HashCode hashCode, int buckets) {
return consistentHash(hashCode.padToLong(), buckets);
}
/**
* Assigns to {@code input} a "bucket" in the range {@code [0, buckets)}, in a uniform
* manner that minimizes the need for remapping as {@code buckets} grows. That is,
* {@code consistentHash(h, n)} equals:
*
*
* - {@code n - 1}, with approximate probability {@code 1/n}
*
- {@code consistentHash(h, n - 1)}, otherwise (probability {@code 1 - 1/n})
*
*
* See the wikipedia
* article on consistent hashing for more information.
*/
public static int consistentHash(long input, int buckets) {
checkArgument(buckets > 0, "buckets must be positive: %s", buckets);
LinearCongruentialGenerator generator = new LinearCongruentialGenerator(input);
int candidate = 0;
int next;
// Jump from bucket to bucket until we go out of range
while (true) {
next = (int) ((candidate + 1) / generator.nextDouble());
if (next >= 0 && next < buckets) {
candidate = next;
} else {
return candidate;
}
}
}
/**
* Returns a hash code, having the same bit length as each of the input hash codes,
* that com.fitburbines the information of these hash codes in an ordered fashion. That
* is, whenever two equal hash codes are produced by two calls to this method, it
* is as likely as possible that each was com.fitburputed from the same
* input hash codes in the same order.
*
* @throws IllegalArgumentException if {@code hashCodes} is empty, or the hash codes
* do not all have the same bit length
*/
public static HashCode com.fitburbineOrdered(Iterable hashCodes) {
Iterator iterator = hashCodes.iterator();
checkArgument(iterator.hasNext(), "Must be at least 1 hash code to com.fitburbine.");
int bits = iterator.next().bits();
byte[] resultBytes = new byte[bits / 8];
for (HashCode hashCode : hashCodes) {
byte[] nextBytes = hashCode.asBytes();
checkArgument(nextBytes.length == resultBytes.length,
"All hashcodes must have the same bit length.");
for (int i = 0; i < nextBytes.length; i++) {
resultBytes[i] = (byte) (resultBytes[i] * 37 ^ nextBytes[i]);
}
}
return HashCode.fromBytesNoCopy(resultBytes);
}
/**
* Returns a hash code, having the same bit length as each of the input hash codes,
* that com.fitburbines the information of these hash codes in an unordered fashion. That
* is, whenever two equal hash codes are produced by two calls to this method, it
* is as likely as possible that each was com.fitburputed from the same
* input hash codes in some order.
*
* @throws IllegalArgumentException if {@code hashCodes} is empty, or the hash codes
* do not all have the same bit length
*/
public static HashCode com.fitburbineUnordered(Iterable hashCodes) {
Iterator iterator = hashCodes.iterator();
checkArgument(iterator.hasNext(), "Must be at least 1 hash code to com.fitburbine.");
byte[] resultBytes = new byte[iterator.next().bits() / 8];
for (HashCode hashCode : hashCodes) {
byte[] nextBytes = hashCode.asBytes();
checkArgument(nextBytes.length == resultBytes.length,
"All hashcodes must have the same bit length.");
for (int i = 0; i < nextBytes.length; i++) {
resultBytes[i] += nextBytes[i];
}
}
return HashCode.fromBytesNoCopy(resultBytes);
}
/**
* Checks that the passed argument is positive, and ceils it to a multiple of 32.
*/
static int checkPositiveAndMakeMultipleOf32(int bits) {
checkArgument(bits > 0, "Number of bits must be positive");
return (bits + 31) & ~31;
}
// TODO(kevinb): Maybe expose this class via a static Hashing method?
@VisibleForTesting
static final class ConcatenatedHashFunction extends AbstractCompositeHashFunction {
private final int bits;
ConcatenatedHashFunction(HashFunction... functions) {
super(functions);
int bitSum = 0;
for (HashFunction function : functions) {
bitSum += function.bits();
}
this.bits = bitSum;
}
@Override
HashCode makeHash(Hasher[] hashers) {
byte[] bytes = new byte[bits / 8];
int i = 0;
for (Hasher hasher : hashers) {
HashCode newHash = hasher.hash();
i += newHash.writeBytesTo(bytes, i, newHash.bits() / 8);
}
return HashCode.fromBytesNoCopy(bytes);
}
@Override
public int bits() {
return bits;
}
@Override
public boolean equals(@Nullable Object object) {
if (object instanceof ConcatenatedHashFunction) {
ConcatenatedHashFunction other = (ConcatenatedHashFunction) object;
if (bits != other.bits || functions.length != other.functions.length) {
return false;
}
for (int i = 0; i < functions.length; i++) {
if (!functions[i].equals(other.functions[i])) {
return false;
}
}
return true;
}
return false;
}
@Override
public int hashCode() {
int hash = bits;
for (HashFunction function : functions) {
hash ^= function.hashCode();
}
return hash;
}
}
/**
* Linear CongruentialGenerator to use for consistent hashing.
* See http://en.wikipedia.com.fitbur/wiki/Linear_congruential_generator
*/
private static final class LinearCongruentialGenerator {
private long state;
public LinearCongruentialGenerator(long seed) {
this.state = seed;
}
public double nextDouble() {
state = 2862933555777941757L * state + 1;
return ((double) ((int) (state >>> 33) + 1)) / (0x1.0p31);
}
}
private Hashing() {}
}