X-Git-Url: http://plrg.eecs.uci.edu/git/?p=folly.git;a=blobdiff_plain;f=folly%2FHash.h;h=77ba5878f17e6502c0c89a0e2fbe42f8c4dc757f;hp=cbd07b56ba126cd81efceb2ef1684f1944c87f12;hb=314c43c4f95dc51f2c97d00216be062eebbd5454;hpb=b71e31a5fe51724cd9b05fbb4bbf87510f01e45f

diff --git a/folly/Hash.h b/folly/Hash.h
index cbd07b56..77ba5878 100644
--- a/folly/Hash.h
+++ b/folly/Hash.h
@@ -1,5 +1,5 @@
 /*
- * Copyright 2013 Facebook, Inc.
+ * Copyright 2017 Facebook, Inc.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -14,17 +14,20 @@
  * limitations under the License.
  */
 
-#ifndef FOLLY_BASE_HASH_H_
-#define FOLLY_BASE_HASH_H_
+#pragma once
 
+#include <cstdint>
 #include <cstring>
-#include <stdint.h>
+#include <limits>
 #include <string>
-#include <utility>
 #include <tuple>
+#include <type_traits>
+#include <utility>
 
-#include "folly/SpookyHashV1.h"
-#include "folly/SpookyHashV2.h"
+#include <folly/ApplyTuple.h>
+#include <folly/Bits.h>
+#include <folly/SpookyHashV1.h>
+#include <folly/SpookyHashV2.h>
 
 /*
  * Various hashing functions.
@@ -42,12 +45,12 @@ namespace folly { namespace hash {
 // This is the Hash128to64 function from Google's cityhash (available
 // under the MIT License).  We use it to reduce multiple 64 bit hashes
 // into a single hash.
-inline size_t hash_128_to_64(const size_t upper, const size_t lower) {
+inline uint64_t hash_128_to_64(const uint64_t upper, const uint64_t lower) {
   // Murmur-inspired hashing.
-  const size_t kMul = 0x9ddfea08eb382d69ULL;
-  size_t a = (lower ^ upper) * kMul;
+  const uint64_t kMul = 0x9ddfea08eb382d69ULL;
+  uint64_t a = (lower ^ upper) * kMul;
   a ^= (a >> 47);
-  size_t b = (upper ^ a) * kMul;
+  uint64_t b = (upper ^ a) * kMul;
   b ^= (b >> 47);
   b *= kMul;
   return b;
@@ -59,6 +62,21 @@ inline size_t hash_combine_generic() {
   return 0;
 }
 
+template <
+    class Iter,
+    class Hash = std::hash<typename std::iterator_traits<Iter>::value_type>>
+uint64_t hash_range(Iter begin,
+                    Iter end,
+                    uint64_t hash = 0,
+                    Hash hasher = Hash()) {
+  for (; begin != end; ++begin) {
+    hash = hash_128_to_64(hash, hasher(*begin));
+  }
+  return hash;
+}
+
+inline uint32_t twang_32from64(uint64_t key);
+
 template <class Hasher, typename T, typename... Ts>
 size_t hash_combine_generic(const T& t, const Ts&... ts) {
   size_t seed = Hasher::hash(t);
@@ -66,7 +84,11 @@ size_t hash_combine_generic(const T& t, const Ts&... ts) {
     return seed;
   }
   size_t remainder = hash_combine_generic<Hasher>(ts...);
-  return hash_128_to_64(seed, remainder);
+  /* static */ if (sizeof(size_t) == sizeof(uint32_t)) {
+    return twang_32from64((uint64_t(seed) << 32) | remainder);
+  } else {
+    return static_cast<size_t>(hash_128_to_64(seed, remainder));
+  }
 }
 
 // Simply uses std::hash to hash.  Note that std::hash is not guaranteed
@@ -190,11 +212,13 @@ inline uint32_t jenkins_rev_unmix32(uint32_t key) {
  *     http://www.isthe.com/chongo/tech/comp/fnv/
  */
 
-const uint32_t FNV_32_HASH_START = 216613626UL;
+const uint32_t FNV_32_HASH_START = 2166136261UL;
 const uint64_t FNV_64_HASH_START = 14695981039346656037ULL;
 
-inline uint32_t fnv32(const char* s,
-                      uint32_t hash = FNV_32_HASH_START) {
+inline uint32_t fnv32(const char* buf, uint32_t hash = FNV_32_HASH_START) {
+  // forcing signed char, since other platforms can use unsigned
+  const signed char* s = reinterpret_cast<const signed char*>(buf);
+
   for (; *s; ++s) {
     hash += (hash << 1) + (hash << 4) + (hash << 7) +
             (hash << 8) + (hash << 24);
@@ -204,11 +228,12 @@ inline uint32_t fnv32(const char* s,
 }
 
 inline uint32_t fnv32_buf(const void* buf,
-                          int n,
+                          size_t n,
                           uint32_t hash = FNV_32_HASH_START) {
-  const char* char_buf = reinterpret_cast<const char*>(buf);
+  // forcing signed char, since other platforms can use unsigned
+  const signed char* char_buf = reinterpret_cast<const signed char*>(buf);
 
-  for (int i = 0; i < n; ++i) {
+  for (size_t i = 0; i < n; ++i) {
     hash += (hash << 1) + (hash << 4) + (hash << 7) +
             (hash << 8) + (hash << 24);
     hash ^= char_buf[i];
@@ -218,12 +243,14 @@ inline uint32_t fnv32_buf(const void* buf,
 }
 
 inline uint32_t fnv32(const std::string& str,
-                      uint64_t hash = FNV_32_HASH_START) {
+                      uint32_t hash = FNV_32_HASH_START) {
   return fnv32_buf(str.data(), str.size(), hash);
 }
 
-inline uint64_t fnv64(const char* s,
-                      uint64_t hash = FNV_64_HASH_START) {
+inline uint64_t fnv64(const char* buf, uint64_t hash = FNV_64_HASH_START) {
+  // forcing signed char, since other platforms can use unsigned
+  const signed char* s = reinterpret_cast<const signed char*>(buf);
+
   for (; *s; ++s) {
     hash += (hash << 1) + (hash << 4) + (hash << 5) + (hash << 7) +
       (hash << 8) + (hash << 40);
@@ -233,11 +260,12 @@ inline uint64_t fnv64(const char* s,
 }
 
 inline uint64_t fnv64_buf(const void* buf,
-                          int n,
+                          size_t n,
                           uint64_t hash = FNV_64_HASH_START) {
-  const char* char_buf = reinterpret_cast<const char*>(buf);
+  // forcing signed char, since other platforms can use unsigned
+  const signed char* char_buf = reinterpret_cast<const signed char*>(buf);
 
-  for (int i = 0; i < n; ++i) {
+  for (size_t i = 0; i < n; ++i) {
     hash += (hash << 1) + (hash << 4) + (hash << 5) + (hash << 7) +
       (hash << 8) + (hash << 40);
     hash ^= char_buf[i];
@@ -254,13 +282,14 @@ inline uint64_t fnv64(const std::string& str,
  * Paul Hsieh: http://www.azillionmonkeys.com/qed/hash.html
  */
 
-#define get16bits(d) (*((const uint16_t*) (d)))
+#define get16bits(d) folly::loadUnaligned<uint16_t>(d)
 
-inline uint32_t hsieh_hash32_buf(const void* buf, int len) {
-  const char* s = reinterpret_cast<const char*>(buf);
-  uint32_t hash = len;
+inline uint32_t hsieh_hash32_buf(const void* buf, size_t len) {
+  // forcing signed char, since other platforms can use unsigned
+  const unsigned char* s = reinterpret_cast<const unsigned char*>(buf);
+  uint32_t hash = static_cast<uint32_t>(len);
   uint32_t tmp;
-  int rem;
+  size_t rem;
 
   if (len <= 0 || buf == 0) {
     return 0;
@@ -322,9 +351,29 @@ inline uint32_t hsieh_hash32_str(const std::string& str) {
 
 } // namespace hash
 
-template<class Key>
+template<class Key, class Enable = void>
 struct hasher;
 
+struct Hash {
+  template <class T>
+  size_t operator()(const T& v) const {
+    return hasher<T>()(v);
+  }
+
+  template <class T, class... Ts>
+  size_t operator()(const T& t, const Ts&... ts) const {
+    return hash::hash_128_to_64((*this)(t), (*this)(ts...));
+  }
+};
+
+template <>
+struct hasher<bool> {
+  size_t operator()(bool key) const {
+    // Make sure that all the output bits depend on the input.
+    return key ? std::numeric_limits<size_t>::max() : 0;
+  }
+};
+
 template<> struct hasher<int32_t> {
   size_t operator()(int32_t key) const {
     return hash::jenkins_rev_mix32(uint32_t(key));
@@ -337,15 +386,75 @@ template<> struct hasher<uint32_t> {
   }
 };
 
+template<> struct hasher<int16_t> {
+  size_t operator()(int16_t key) const {
+    return hasher<int32_t>()(key); // as impl accident, sign-extends
+  }
+};
+
+template<> struct hasher<uint16_t> {
+  size_t operator()(uint16_t key) const {
+    return hasher<uint32_t>()(key);
+  }
+};
+
+template<> struct hasher<int8_t> {
+  size_t operator()(int8_t key) const {
+    return hasher<int32_t>()(key); // as impl accident, sign-extends
+  }
+};
+
+template<> struct hasher<uint8_t> {
+  size_t operator()(uint8_t key) const {
+    return hasher<uint32_t>()(key);
+  }
+};
+
+template<> struct hasher<char> {
+  using explicit_type =
+      std::conditional<std::is_signed<char>::value, int8_t, uint8_t>::type;
+  size_t operator()(char key) const {
+    return hasher<explicit_type>()(key); // as impl accident, sign-extends
+  }
+};
+
 template<> struct hasher<int64_t> {
   size_t operator()(int64_t key) const {
-    return hash::twang_mix64(uint64_t(key));
+    return static_cast<size_t>(hash::twang_mix64(uint64_t(key)));
   }
 };
 
 template<> struct hasher<uint64_t> {
   size_t operator()(uint64_t key) const {
-    return hash::twang_mix64(key);
+    return static_cast<size_t>(hash::twang_mix64(key));
+  }
+};
+
+template<> struct hasher<std::string> {
+  size_t operator()(const std::string& key) const {
+    return static_cast<size_t>(
+        hash::SpookyHashV2::Hash64(key.data(), key.size(), 0));
+  }
+};
+
+template <class T>
+struct hasher<T, typename std::enable_if<std::is_enum<T>::value, void>::type> {
+  size_t operator()(T key) const {
+    return Hash()(static_cast<typename std::underlying_type<T>::type>(key));
+  }
+};
+
+template <class T1, class T2>
+struct hasher<std::pair<T1, T2>> {
+  size_t operator()(const std::pair<T1, T2>& key) const {
+    return Hash()(key.first, key.second);
+  }
+};
+
+template <typename... Ts>
+struct hasher<std::tuple<Ts...>> {
+  size_t operator() (const std::tuple<Ts...>& key) const {
+    return applyTuple(Hash(), key);
   }
 };
 
@@ -376,7 +485,7 @@ namespace std {
   // Hash function for pairs. Requires default hash functions for both
   // items in the pair.
   template <typename T1, typename T2>
-  class hash<std::pair<T1, T2> > {
+  struct hash<std::pair<T1, T2> > {
   public:
     size_t operator()(const std::pair<T1, T2>& x) const {
       return folly::hash::hash_combine(x.first, x.second);
@@ -395,5 +504,3 @@ namespace std {
     }
   };
 } // namespace std
-
-#endif