tanjent@gmail.com | f3b7897 | 2012-03-01 03:38:55 +0000 | [diff] [blame] | 1 | // Copyright (c) 2011 Google, Inc. |
| 2 | // |
| 3 | // Permission is hereby granted, free of charge, to any person obtaining a copy |
| 4 | // of this software and associated documentation files (the "Software"), to deal |
| 5 | // in the Software without restriction, including without limitation the rights |
| 6 | // to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
| 7 | // copies of the Software, and to permit persons to whom the Software is |
| 8 | // furnished to do so, subject to the following conditions: |
| 9 | // |
| 10 | // The above copyright notice and this permission notice shall be included in |
| 11 | // all copies or substantial portions of the Software. |
| 12 | // |
| 13 | // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 14 | // IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 15 | // FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| 16 | // AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 17 | // LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 18 | // OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN |
| 19 | // THE SOFTWARE. |
| 20 | // |
| 21 | // CityHash, by Geoff Pike and Jyrki Alakuijala |
| 22 | // |
| 23 | // This file provides a few functions for hashing strings. On x86-64 |
| 24 | // hardware in 2011, CityHash64() is faster than other high-quality |
| 25 | // hash functions, such as Murmur. This is largely due to higher |
| 26 | // instruction-level parallelism. CityHash64() and CityHash128() also perform |
| 27 | // well on hash-quality tests. |
| 28 | // |
| 29 | // CityHash128() is optimized for relatively long strings and returns |
| 30 | // a 128-bit hash. For strings more than about 2000 bytes it can be |
| 31 | // faster than CityHash64(). |
| 32 | // |
| 33 | // Functions in the CityHash family are not suitable for cryptography. |
| 34 | // |
| 35 | // WARNING: This code has not been tested on big-endian platforms! |
| 36 | // It is known to work well on little-endian platforms that have a small penalty |
| 37 | // for unaligned reads, such as current Intel and AMD moderate-to-high-end CPUs. |
| 38 | // |
| 39 | // By the way, for some hash functions, given strings a and b, the hash |
| 40 | // of a+b is easily derived from the hashes of a and b. This property |
| 41 | // doesn't hold for any hash functions in this file. |
| 42 | |
| 43 | #ifndef CITY_HASH_H_ |
| 44 | #define CITY_HASH_H_ |
| 45 | |
| 46 | #include <stdlib.h> // for size_t. |
tanjent@gmail.com | f3b7897 | 2012-03-01 03:38:55 +0000 | [diff] [blame] | 47 | #include <utility> |
| 48 | |
aappleby@google.com | e87738e | 2014-04-22 21:40:50 +0000 | [diff] [blame] | 49 | // Microsoft Visual Studio may not have stdint.h. |
| 50 | #if defined(_MSC_VER) && (_MSC_VER < 1600) |
| 51 | typedef unsigned char uint8_t; |
| 52 | typedef unsigned int uint32_t; |
| 53 | typedef unsigned __int64 uint64_t; |
| 54 | #else // defined(_MSC_VER) |
| 55 | #include <stdint.h> |
| 56 | #endif // !defined(_MSC_VER) |
| 57 | |
tanjent@gmail.com | f3b7897 | 2012-03-01 03:38:55 +0000 | [diff] [blame] | 58 | typedef uint8_t uint8; |
| 59 | typedef uint32_t uint32; |
| 60 | typedef uint64_t uint64; |
| 61 | typedef std::pair<uint64, uint64> uint128; |
| 62 | |
| 63 | inline uint64 Uint128Low64(const uint128& x) { return x.first; } |
| 64 | inline uint64 Uint128High64(const uint128& x) { return x.second; } |
| 65 | |
| 66 | // Hash function for a byte array. |
| 67 | uint64 CityHash64(const char *buf, size_t len); |
| 68 | |
| 69 | // Hash function for a byte array. For convenience, a 64-bit seed is also |
| 70 | // hashed into the result. |
| 71 | uint64 CityHash64WithSeed(const char *buf, size_t len, uint64 seed); |
| 72 | |
| 73 | // Hash function for a byte array. For convenience, two seeds are also |
| 74 | // hashed into the result. |
| 75 | uint64 CityHash64WithSeeds(const char *buf, size_t len, |
| 76 | uint64 seed0, uint64 seed1); |
| 77 | |
| 78 | // Hash function for a byte array. |
| 79 | uint128 CityHash128(const char *s, size_t len); |
| 80 | |
| 81 | // Hash function for a byte array. For convenience, a 128-bit seed is also |
| 82 | // hashed into the result. |
| 83 | uint128 CityHash128WithSeed(const char *s, size_t len, uint128 seed); |
| 84 | |
| 85 | // Hash 128 input bits down to 64 bits of output. |
| 86 | // This is intended to be a reasonably good hash function. |
| 87 | inline uint64 Hash128to64(const uint128& x) { |
| 88 | // Murmur-inspired hashing. |
| 89 | const uint64 kMul = 0x9ddfea08eb382d69ULL; |
| 90 | uint64 a = (Uint128Low64(x) ^ Uint128High64(x)) * kMul; |
| 91 | a ^= (a >> 47); |
| 92 | uint64 b = (Uint128High64(x) ^ a) * kMul; |
| 93 | b ^= (b >> 47); |
| 94 | b *= kMul; |
| 95 | return b; |
| 96 | } |
| 97 | |
| 98 | // Conditionally include declarations for versions of City that require SSE4.2 |
| 99 | // instructions to be available. |
tanjent@gmail.com | 8adb133 | 2013-04-19 19:57:16 +0000 | [diff] [blame] | 100 | #if defined(__SSE4_2__) && defined(__x86_64__) |
tanjent@gmail.com | f3b7897 | 2012-03-01 03:38:55 +0000 | [diff] [blame] | 101 | |
| 102 | // Hash function for a byte array. |
| 103 | uint128 CityHashCrc128(const char *s, size_t len); |
| 104 | |
| 105 | // Hash function for a byte array. For convenience, a 128-bit seed is also |
| 106 | // hashed into the result. |
| 107 | uint128 CityHashCrc128WithSeed(const char *s, size_t len, uint128 seed); |
| 108 | |
| 109 | // Hash function for a byte array. Sets result[0] ... result[3]. |
| 110 | void CityHashCrc256(const char *s, size_t len, uint64 *result); |
| 111 | |
| 112 | #endif // __SSE4_2__ |
| 113 | |
| 114 | #endif // CITY_HASH_H_ |