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. |
47 | #include <stdint.h> |
48 | #include <utility> |
49 | |
50 | /** This is a version of CityHash that predates v1.0.3 algorithm change. |
51 | * Why we need exactly this version? |
52 | * Although hash values of CityHash are not recommended for storing persistently anywhere, |
53 | * it has already been used this way in ClickHouse: |
54 | * - for calculation of checksums of compressed chunks and for data parts; |
55 | * - this version of CityHash is exposed in cityHash64 function in ClickHouse SQL language; |
56 | * - and already used by many users for data ordering, sampling and sharding. |
57 | */ |
58 | namespace CityHash_v1_0_2 |
59 | { |
60 | |
61 | typedef uint8_t uint8; |
62 | typedef uint32_t uint32; |
63 | typedef uint64_t uint64; |
64 | typedef std::pair<uint64, uint64> uint128; |
65 | |
66 | |
67 | inline uint64 Uint128Low64(const uint128& x) { return x.first; } |
68 | inline uint64 Uint128High64(const uint128& x) { return x.second; } |
69 | |
70 | // Hash function for a byte array. |
71 | uint64 CityHash64(const char *buf, size_t len); |
72 | |
73 | // Hash function for a byte array. For convenience, a 64-bit seed is also |
74 | // hashed into the result. |
75 | uint64 CityHash64WithSeed(const char *buf, size_t len, uint64 seed); |
76 | |
77 | // Hash function for a byte array. For convenience, two seeds are also |
78 | // hashed into the result. |
79 | uint64 CityHash64WithSeeds(const char *buf, size_t len, |
80 | uint64 seed0, uint64 seed1); |
81 | |
82 | // Hash function for a byte array. |
83 | uint128 CityHash128(const char *s, size_t len); |
84 | |
85 | // Hash function for a byte array. For convenience, a 128-bit seed is also |
86 | // hashed into the result. |
87 | uint128 CityHash128WithSeed(const char *s, size_t len, uint128 seed); |
88 | |
89 | // Hash 128 input bits down to 64 bits of output. |
90 | // This is intended to be a reasonably good hash function. |
91 | inline uint64 Hash128to64(const uint128& x) { |
92 | // Murmur-inspired hashing. |
93 | const uint64 kMul = 0x9ddfea08eb382d69ULL; |
94 | uint64 a = (Uint128Low64(x) ^ Uint128High64(x)) * kMul; |
95 | a ^= (a >> 47); |
96 | uint64 b = (Uint128High64(x) ^ a) * kMul; |
97 | b ^= (b >> 47); |
98 | b *= kMul; |
99 | return b; |
100 | } |
101 | |
102 | } |
103 | |
104 | #endif // CITY_HASH_H_ |
105 | |