xxhash 0.3.0 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: eb65711481aa5bbc9e1b7dc6c8338e0bda70f377
-  data.tar.gz: 5c4c2b5a262bd5831f9658aeccf35457a0556ce8
+  metadata.gz: 7631f1cb865522390841ca6345dadf0247b78404
+  data.tar.gz: 07f7a162f2be6f2df03aad3993e8fa74ded83fcc
 SHA512:
-  metadata.gz: a1160fe558a79aaa74a9fabe74e902890bcb7a7512ad08ff4634839fa40dfb2abe74eef367b689c5d42f178380a77deba6d7d6a8cd15ebf94d52a6600c4a7e85
-  data.tar.gz: 5615f46479f9c86b2af11dc36cea59abd1205ba2b06f0364c572bb5b54aa5a816d5f643619b5c0f43965fc9928763ea3c8b112a8134a391beee09a7d527f4d0a
+  metadata.gz: 278c331aaa693e1489ceaf6c302a8f9f21087457867807a9f043ad3fe673279e09a05906e10e57266ee3980ce51c3d60c13d70b133b7312b75c8cb8962c5cbd2
+  data.tar.gz: e220c7ee0bc95631668cbc981deb5105fdf7f09c1c2acfe4f7a3bf4f532d03e69bdf7950403ade750cd8ceeede2f75df63967df98c478a595e1a6c33bf310ea1

data/.travis.yml

@@ -1,4 +1,9 @@
 language: ruby
+before_install:
+  - gem install bundler
 rvm:
   - 1.9.3
-  - rbx-19mode
+  - 2.0.0
+  - 2.1
+  - 2.2
+script: "bundle exec rake test"

data/CHANGELOG.md

@@ -1,6 +1,13 @@
-### master
+### 0.4.0  (December 11, 2016)
+  * Add xxHash to `Digest` module and make it more compatieble
+    with another hash functions from `Digest`. (by [@justinwsmith](https://github.com/justinwsmith))
+  * Add a `StreamingHash` class that's externally instantiatable. (by [@justinwsmith](https://github.com/justinwsmith))
+  * Fix segfault when nil is passed as param (https://github.com/nashby/xxhash/issues/13) (by [@justinwsmith](https://github.com/justinwsmith))
+  * Update libxxhash to 0.6.2.
+
+### 0.3.0 (January 18, 2015)
   * make seed param optional (by [@weakish](https://github.com/weakish))
-  * add 64-bit xxhash function (by [@justinwsmith](https://github.com/justinwsmith)) 
+  * add 64-bit xxhash function (by [@justinwsmith](https://github.com/justinwsmith))
 
 ### 0.2.0 (September 4, 2013)
   * xxHash updated to [r32](https://code.google.com/p/xxhash/source/detail?r=32)

data/README.md

@@ -1,6 +1,6 @@
 ## xxHash [![Build Status](https://travis-ci.org/nashby/xxhash.png?branch=master)](https://travis-ci.org/nashby/xxhash)
 
-Ruby wrapper for [xxHash](http://code.google.com/p/xxhash/)
+Ruby wrapper for [xxHash](https://github.com/Cyan4973/xxHash)
 
 ### Install
 
@@ -17,6 +17,8 @@ seed = 12345
 XXhash.xxh32(text, seed) # => 3834992036
 ```
 
+If you do not provide a seed, it will use the default value `0`.
+
 You can use it with `IO` objects too:
 
 ```ruby
@@ -25,16 +27,18 @@ XXhash.xxh32_stream(StringIO.new('test'), 123) # => 2758658570
 
 Note that you can also pass a chunk size as third param (it's 32 bytes by default)
 
+XXH64 is also supported: you can use `xxh64` and `xxh64_stream`.
+
 ### Supported Ruby versions
 
-- MRI 1.9.3
+- MRI 1.9.3, 2.0, 2.1, 2.2.
 - rbx-19mode
 
 Note: It doesn't work on JRuby as it uses C extension.
 
 ### Versioning
 
-Version 0.2.0 is equal to [r32](https://code.google.com/p/xxhash/source/detail?r32)
+Version 0.4.0 is equal to [0.6.2](https://github.com/Cyan4973/xxHash/tree/v0.6.2)
 
 ## Contributing
 

data/ext/xxhash/libxxhash.c

@@ -1,154 +1,183 @@
 /*
-xxHash - Fast Hash algorithm
-Copyright (C) 2012-2014, Yann Collet.
-BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
-
-Redistribution and use in source and binary forms, with or without
-modification, are permitted provided that the following conditions are
-met:
-
-* Redistributions of source code must retain the above copyright
-notice, this list of conditions and the following disclaimer.
-* Redistributions in binary form must reproduce the above
-copyright notice, this list of conditions and the following disclaimer
-in the documentation and/or other materials provided with the
-distribution.
-
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
-"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
-LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
-A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
-OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
-SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
-LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
-DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
-THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
-(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
-OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-
-You can contact the author at :
-- xxHash source repository : http://code.google.com/p/xxhash/
-- public discussion board : https://groups.google.com/forum/#!forum/lz4c
+*  xxHash - Fast Hash algorithm
+*  Copyright (C) 2012-2016, Yann Collet
+*
+*  BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
+*
+*  Redistribution and use in source and binary forms, with or without
+*  modification, are permitted provided that the following conditions are
+*  met:
+*
+*  * Redistributions of source code must retain the above copyright
+*  notice, this list of conditions and the following disclaimer.
+*  * Redistributions in binary form must reproduce the above
+*  copyright notice, this list of conditions and the following disclaimer
+*  in the documentation and/or other materials provided with the
+*  distribution.
+*
+*  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+*  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+*  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+*  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+*  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+*  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+*  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+*  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+*  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+*  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+*  OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*
+*  You can contact the author at :
+*  - xxHash homepage: http://www.xxhash.com
+*  - xxHash source repository : https://github.com/Cyan4973/xxHash
 */
 
 
-//**************************************
-// Tuning parameters
-//**************************************
-// Unaligned memory access is automatically enabled for "common" CPU, such as x86.
-// For others CPU, the compiler will be more cautious, and insert extra code to ensure aligned access is respected.
-// If you know your target CPU supports unaligned memory access, you want to force this option manually to improve performance.
-// You can also enable this parameter if you know your input data will always be aligned (boundaries of 4, for U32).
-#if defined(__ARM_FEATURE_UNALIGNED) || defined(__i386) || defined(_M_IX86) || defined(__x86_64__) || defined(_M_X64)
-#  define XXH_USE_UNALIGNED_ACCESS 1
+/* *************************************
+*  Tuning parameters
+***************************************/
+/*!XXH_FORCE_MEMORY_ACCESS :
+ * By default, access to unaligned memory is controlled by `memcpy()`, which is safe and portable.
+ * Unfortunately, on some target/compiler combinations, the generated assembly is sub-optimal.
+ * The below switch allow to select different access method for improved performance.
+ * Method 0 (default) : use `memcpy()`. Safe and portable.
+ * Method 1 : `__packed` statement. It depends on compiler extension (ie, not portable).
+ *            This method is safe if your compiler supports it, and *generally* as fast or faster than `memcpy`.
+ * Method 2 : direct access. This method doesn't depend on compiler but violate C standard.
+ *            It can generate buggy code on targets which do not support unaligned memory accesses.
+ *            But in some circumstances, it's the only known way to get the most performance (ie GCC + ARMv6)
+ * See http://stackoverflow.com/a/32095106/646947 for details.
+ * Prefer these methods in priority order (0 > 1 > 2)
+ */
+#ifndef XXH_FORCE_MEMORY_ACCESS   /* can be defined externally, on command line for example */
+#  if defined(__GNUC__) && ( defined(__ARM_ARCH_6__) || defined(__ARM_ARCH_6J__) || defined(__ARM_ARCH_6K__) || defined(__ARM_ARCH_6Z__) || defined(__ARM_ARCH_6ZK__) || defined(__ARM_ARCH_6T2__) )
+#    define XXH_FORCE_MEMORY_ACCESS 2
+#  elif defined(__INTEL_COMPILER) || \
+  (defined(__GNUC__) && ( defined(__ARM_ARCH_7__) || defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7R__) || defined(__ARM_ARCH_7M__) || defined(__ARM_ARCH_7S__) ))
+#    define XXH_FORCE_MEMORY_ACCESS 1
+#  endif
 #endif
 
-// XXH_ACCEPT_NULL_INPUT_POINTER :
-// If the input pointer is a null pointer, xxHash default behavior is to trigger a memory access error, since it is a bad pointer.
-// When this option is enabled, xxHash output for null input pointers will be the same as a null-length input.
-// This option has a very small performance cost (only measurable on small inputs).
-// By default, this option is disabled. To enable it, uncomment below define :
-// #define XXH_ACCEPT_NULL_INPUT_POINTER 1
-
-// XXH_FORCE_NATIVE_FORMAT :
-// By default, xxHash library provides endian-independant Hash values, based on little-endian convention.
-// Results are therefore identical for little-endian and big-endian CPU.
-// This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format.
-// Should endian-independance be of no importance for your application, you may set the #define below to 1.
-// It will improve speed for Big-endian CPU.
-// This option has no impact on Little_Endian CPU.
-#define XXH_FORCE_NATIVE_FORMAT 0
-
-//**************************************
-// Compiler Specific Options
-//**************************************
-// Disable some Visual warning messages
-#ifdef _MSC_VER  // Visual Studio
-#  pragma warning(disable : 4127)      // disable: C4127: conditional expression is constant
+/*!XXH_ACCEPT_NULL_INPUT_POINTER :
+ * If the input pointer is a null pointer, xxHash default behavior is to trigger a memory access error, since it is a bad pointer.
+ * When this option is enabled, xxHash output for null input pointers will be the same as a null-length input.
+ * By default, this option is disabled. To enable it, uncomment below define :
+ */
+/* #define XXH_ACCEPT_NULL_INPUT_POINTER 1 */
+
+/*!XXH_FORCE_NATIVE_FORMAT :
+ * By default, xxHash library provides endian-independant Hash values, based on little-endian convention.
+ * Results are therefore identical for little-endian and big-endian CPU.
+ * This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format.
+ * Should endian-independance be of no importance for your application, you may set the #define below to 1,
+ * to improve speed for Big-endian CPU.
+ * This option has no impact on Little_Endian CPU.
+ */
+#ifndef XXH_FORCE_NATIVE_FORMAT   /* can be defined externally */
+#  define XXH_FORCE_NATIVE_FORMAT 0
 #endif
 
-#ifdef _MSC_VER    // Visual Studio
-#  define FORCE_INLINE static __forceinline
-#else
-#  ifdef __GNUC__
-#    define FORCE_INLINE static inline __attribute__((always_inline))
+/*!XXH_FORCE_ALIGN_CHECK :
+ * This is a minor performance trick, only useful with lots of very small keys.
+ * It means : check for aligned/unaligned input.
+ * The check costs one initial branch per hash; set to 0 when the input data
+ * is guaranteed to be aligned.
+ */
+#ifndef XXH_FORCE_ALIGN_CHECK /* can be defined externally */
+#  if defined(__i386) || defined(_M_IX86) || defined(__x86_64__) || defined(_M_X64)
+#    define XXH_FORCE_ALIGN_CHECK 0
 #  else
-#    define FORCE_INLINE static inline
+#    define XXH_FORCE_ALIGN_CHECK 1
 #  endif
 #endif
 
-//**************************************
-// Includes & Memory related functions
-//**************************************
-#include "libxxhash.h"
-// Modify the local functions below should you wish to use some other memory routines
-// for malloc(), free()
+
+/* *************************************
+*  Includes & Memory related functions
+***************************************/
+/* Modify the local functions below should you wish to use some other memory routines */
+/* for malloc(), free() */
 #include <stdlib.h>
 static void* XXH_malloc(size_t s) { return malloc(s); }
 static void  XXH_free  (void* p)  { free(p); }
-// for memcpy()
+/* for memcpy() */
 #include <string.h>
-static void* XXH_memcpy(void* dest, const void* src, size_t size)
-{
-    return memcpy(dest,src,size);
-}
+static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); }
 
+#define XXH_STATIC_LINKING_ONLY
+#include "xxhash.h"
 
-//**************************************
-// Basic Types
-//**************************************
-#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L   // C99
-# include <stdint.h>
-typedef uint8_t  BYTE;
-typedef uint16_t U16;
-typedef uint32_t U32;
-typedef  int32_t S32;
-typedef uint64_t U64;
-#else
-typedef unsigned char      BYTE;
-typedef unsigned short     U16;
-typedef unsigned int       U32;
-typedef   signed int       S32;
-typedef unsigned long long U64;
-#endif
 
-#if defined(__GNUC__)  && !defined(XXH_USE_UNALIGNED_ACCESS)
-#  define _PACKED __attribute__ ((packed))
+/* *************************************
+*  Compiler Specific Options
+***************************************/
+#ifdef _MSC_VER    /* Visual Studio */
+#  pragma warning(disable : 4127)      /* disable: C4127: conditional expression is constant */
+#  define FORCE_INLINE static __forceinline
 #else
-#  define _PACKED
-#endif
-
-#if !defined(XXH_USE_UNALIGNED_ACCESS) && !defined(__GNUC__)
-#  ifdef __IBMC__
-#    pragma pack(1)
+#  if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L   /* C99 */
+#    ifdef __GNUC__
+#      define FORCE_INLINE static inline __attribute__((always_inline))
+#    else
+#      define FORCE_INLINE static inline
+#    endif
 #  else
-#    pragma pack(push, 1)
-#  endif
+#    define FORCE_INLINE static
+#  endif /* __STDC_VERSION__ */
 #endif
 
-typedef struct _U32_S
-{
-    U32 v;
-} _PACKED U32_S;
-typedef struct _U64_S
-{
-    U64 v;
-} _PACKED U64_S;
 
-#if !defined(XXH_USE_UNALIGNED_ACCESS) && !defined(__GNUC__)
-#  pragma pack(pop)
+/* *************************************
+*  Basic Types
+***************************************/
+#ifndef MEM_MODULE
+# if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
+#   include <stdint.h>
+    typedef uint8_t  BYTE;
+    typedef uint16_t U16;
+    typedef uint32_t U32;
+    typedef  int32_t S32;
+# else
+    typedef unsigned char      BYTE;
+    typedef unsigned short     U16;
+    typedef unsigned int       U32;
+    typedef   signed int       S32;
+# endif
 #endif
 
-#define A32(x) (((U32_S *)(x))->v)
-#define A64(x) (((U64_S *)(x))->v)
+#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2))
+
+/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */
+static U32 XXH_read32(const void* memPtr) { return *(const U32*) memPtr; }
 
+#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1))
 
-//***************************************
-// Compiler-specific Functions and Macros
-//***************************************
+/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */
+/* currently only defined for gcc and icc */
+typedef union { U32 u32; } __attribute__((packed)) unalign;
+static U32 XXH_read32(const void* ptr) { return ((const unalign*)ptr)->u32; }
+
+#else
+
+/* portable and safe solution. Generally efficient.
+ * see : http://stackoverflow.com/a/32095106/646947
+ */
+static U32 XXH_read32(const void* memPtr)
+{
+    U32 val;
+    memcpy(&val, memPtr, sizeof(val));
+    return val;
+}
+
+#endif   /* XXH_FORCE_DIRECT_MEMORY_ACCESS */
+
+
+/* ****************************************
+*  Compiler-specific Functions and Macros
+******************************************/
 #define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
 
-// Note : although _rotl exists for minGW (GCC under windows), performance seems poor
+/* Note : although _rotl exists for minGW (GCC under windows), performance seems poor */
 #if defined(_MSC_VER)
 #  define XXH_rotl32(x,r) _rotl(x,r)
 #  define XXH_rotl64(x,r) _rotl64(x,r)
@@ -157,76 +186,44 @@ typedef struct _U64_S
 #  define XXH_rotl64(x,r) ((x << r) | (x >> (64 - r)))
 #endif
 
-#if defined(_MSC_VER)     // Visual Studio
+#if defined(_MSC_VER)     /* Visual Studio */
 #  define XXH_swap32 _byteswap_ulong
-#  define XXH_swap64 _byteswap_uint64
 #elif GCC_VERSION >= 403
 #  define XXH_swap32 __builtin_bswap32
-#  define XXH_swap64 __builtin_bswap64
 #else
-static inline U32 XXH_swap32 (U32 x)
+static U32 XXH_swap32 (U32 x)
 {
     return  ((x << 24) & 0xff000000 ) |
             ((x <<  8) & 0x00ff0000 ) |
             ((x >>  8) & 0x0000ff00 ) |
             ((x >> 24) & 0x000000ff );
 }
-static inline U64 XXH_swap64 (U64 x)
-{
-    return  ((x << 56) & 0xff00000000000000ULL) |
-            ((x << 40) & 0x00ff000000000000ULL) |
-            ((x << 24) & 0x0000ff0000000000ULL) |
-            ((x << 8)  & 0x000000ff00000000ULL) |
-            ((x >> 8)  & 0x00000000ff000000ULL) |
-            ((x >> 24) & 0x0000000000ff0000ULL) |
-            ((x >> 40) & 0x000000000000ff00ULL) |
-            ((x >> 56) & 0x00000000000000ffULL);
-}
 #endif
 
 
-//**************************************
-// Constants
-//**************************************
-#define PRIME32_1   2654435761U
-#define PRIME32_2   2246822519U
-#define PRIME32_3   3266489917U
-#define PRIME32_4    668265263U
-#define PRIME32_5    374761393U
-
-#define PRIME64_1 11400714785074694791ULL
-#define PRIME64_2 14029467366897019727ULL
-#define PRIME64_3  1609587929392839161ULL
-#define PRIME64_4  9650029242287828579ULL
-#define PRIME64_5  2870177450012600261ULL
-
-//**************************************
-// Architecture Macros
-//**************************************
+/* *************************************
+*  Architecture Macros
+***************************************/
 typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess;
-#ifndef XXH_CPU_LITTLE_ENDIAN   // It is possible to define XXH_CPU_LITTLE_ENDIAN externally, for example using a compiler switch
-static const int one = 1;
-#   define XXH_CPU_LITTLE_ENDIAN   (*(char*)(&one))
-#endif
-
 
-//**************************************
-// Macros
-//**************************************
-#define XXH_STATIC_ASSERT(c)   { enum { XXH_static_assert = 1/(!!(c)) }; }    // use only *after* variable declarations
+/* XXH_CPU_LITTLE_ENDIAN can be defined externally, for example on the compiler command line */
+#ifndef XXH_CPU_LITTLE_ENDIAN
+    static const int g_one = 1;
+#   define XXH_CPU_LITTLE_ENDIAN   (*(const char*)(&g_one))
+#endif
 
 
-//****************************
-// Memory reads
-//****************************
+/* ***************************
+*  Memory reads
+*****************************/
 typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment;
 
 FORCE_INLINE U32 XXH_readLE32_align(const void* ptr, XXH_endianess endian, XXH_alignment align)
 {
     if (align==XXH_unaligned)
-        return endian==XXH_littleEndian ? A32(ptr) : XXH_swap32(A32(ptr));
+        return endian==XXH_littleEndian ? XXH_read32(ptr) : XXH_swap32(XXH_read32(ptr));
     else
-        return endian==XXH_littleEndian ? *(U32*)ptr : XXH_swap32(*(U32*)ptr);
+        return endian==XXH_littleEndian ? *(const U32*)ptr : XXH_swap32(*(const U32*)ptr);
 }
 
 FORCE_INLINE U32 XXH_readLE32(const void* ptr, XXH_endianess endian)
@@ -234,23 +231,36 @@ FORCE_INLINE U32 XXH_readLE32(const void* ptr, XXH_endianess endian)
     return XXH_readLE32_align(ptr, endian, XXH_unaligned);
 }
 
-FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align)
+static U32 XXH_readBE32(const void* ptr)
 {
-    if (align==XXH_unaligned)
-        return endian==XXH_littleEndian ? A64(ptr) : XXH_swap64(A64(ptr));
-    else
-        return endian==XXH_littleEndian ? *(U64*)ptr : XXH_swap64(*(U64*)ptr);
+    return XXH_CPU_LITTLE_ENDIAN ? XXH_swap32(XXH_read32(ptr)) : XXH_read32(ptr);
 }
 
-FORCE_INLINE U64 XXH_readLE64(const void* ptr, XXH_endianess endian)
+
+/* *************************************
+*  Macros
+***************************************/
+#define XXH_STATIC_ASSERT(c)   { enum { XXH_static_assert = 1/(int)(!!(c)) }; }    /* use only *after* variable declarations */
+XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; }
+
+
+/* *******************************************************************
+*  32-bits hash functions
+*********************************************************************/
+static const U32 PRIME32_1 = 2654435761U;
+static const U32 PRIME32_2 = 2246822519U;
+static const U32 PRIME32_3 = 3266489917U;
+static const U32 PRIME32_4 =  668265263U;
+static const U32 PRIME32_5 =  374761393U;
+
+static U32 XXH32_round(U32 seed, U32 input)
 {
-    return XXH_readLE64_align(ptr, endian, XXH_unaligned);
+    seed += input * PRIME32_2;
+    seed  = XXH_rotl32(seed, 13);
+    seed *= PRIME32_1;
+    return seed;
 }
 
-
-//****************************
-// Simple Hash Functions
-//****************************
 FORCE_INLINE U32 XXH32_endian_align(const void* input, size_t len, U32 seed, XXH_endianess endian, XXH_alignment align)
 {
     const BYTE* p = (const BYTE*)input;
@@ -259,60 +269,40 @@ FORCE_INLINE U32 XXH32_endian_align(const void* input, size_t len, U32 seed, XXH
 #define XXH_get32bits(p) XXH_readLE32_align(p, endian, align)
 
 #ifdef XXH_ACCEPT_NULL_INPUT_POINTER
-    if (p==NULL)
-    {
+    if (p==NULL) {
         len=0;
         bEnd=p=(const BYTE*)(size_t)16;
     }
 #endif
 
-    if (len>=16)
-    {
+    if (len>=16) {
         const BYTE* const limit = bEnd - 16;
         U32 v1 = seed + PRIME32_1 + PRIME32_2;
         U32 v2 = seed + PRIME32_2;
         U32 v3 = seed + 0;
         U32 v4 = seed - PRIME32_1;
 
-        do
-        {
-            v1 += XXH_get32bits(p) * PRIME32_2;
-            v1 = XXH_rotl32(v1, 13);
-            v1 *= PRIME32_1;
-            p+=4;
-            v2 += XXH_get32bits(p) * PRIME32_2;
-            v2 = XXH_rotl32(v2, 13);
-            v2 *= PRIME32_1;
-            p+=4;
-            v3 += XXH_get32bits(p) * PRIME32_2;
-            v3 = XXH_rotl32(v3, 13);
-            v3 *= PRIME32_1;
-            p+=4;
-            v4 += XXH_get32bits(p) * PRIME32_2;
-            v4 = XXH_rotl32(v4, 13);
-            v4 *= PRIME32_1;
-            p+=4;
-        }
-        while (p<=limit);
+        do {
+            v1 = XXH32_round(v1, XXH_get32bits(p)); p+=4;
+            v2 = XXH32_round(v2, XXH_get32bits(p)); p+=4;
+            v3 = XXH32_round(v3, XXH_get32bits(p)); p+=4;
+            v4 = XXH32_round(v4, XXH_get32bits(p)); p+=4;
+        } while (p<=limit);
 
         h32 = XXH_rotl32(v1, 1) + XXH_rotl32(v2, 7) + XXH_rotl32(v3, 12) + XXH_rotl32(v4, 18);
-    }
-    else
-    {
+    } else {
         h32  = seed + PRIME32_5;
     }
 
     h32 += (U32) len;
 
-    while (p+4<=bEnd)
-    {
+    while (p+4<=bEnd) {
         h32 += XXH_get32bits(p) * PRIME32_3;
         h32  = XXH_rotl32(h32, 17) * PRIME32_4 ;
         p+=4;
     }
 
-    while (p<bEnd)
-    {
+    while (p<bEnd) {
         h32 += (*p) * PRIME32_5;
         h32 = XXH_rotl32(h32, 11) * PRIME32_1 ;
         p++;
@@ -328,26 +318,24 @@ FORCE_INLINE U32 XXH32_endian_align(const void* input, size_t len, U32 seed, XXH
 }
 
 
-unsigned int XXH32 (const void* input, size_t len, unsigned seed)
+XXH_PUBLIC_API unsigned int XXH32 (const void* input, size_t len, unsigned int seed)
 {
 #if 0
-    // Simple version, good for code maintenance, but unfortunately slow for small inputs
-    XXH32_state_t state;
-    XXH32_reset(&state, seed);
-    XXH32_update(&state, input, len);
-    return XXH32_digest(&state);
+    /* Simple version, good for code maintenance, but unfortunately slow for small inputs */
+    XXH32_CREATESTATE_STATIC(state);
+    XXH32_reset(state, seed);
+    XXH32_update(state, input, len);
+    return XXH32_digest(state);
 #else
     XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
 
-#  if !defined(XXH_USE_UNALIGNED_ACCESS)
-    if ((((size_t)input) & 3) == 0)   // Input is aligned, let's leverage the speed advantage
-    {
-        if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
-            return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
-        else
-            return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
-    }
-#  endif
+    if (XXH_FORCE_ALIGN_CHECK) {
+        if ((((size_t)input) & 3) == 0) {   /* Input is 4-bytes aligned, leverage the speed benefit */
+            if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
+                return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
+            else
+                return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
+    }   }
 
     if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
         return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned);
@@ -356,231 +344,40 @@ unsigned int XXH32 (const void* input, size_t len, unsigned seed)
 #endif
 }
 
-FORCE_INLINE U64 XXH64_endian_align(const void* input, size_t len, U64 seed, XXH_endianess endian, XXH_alignment align)
-{
-    const BYTE* p = (const BYTE*)input;
-    const BYTE* bEnd = p + len;
-    U64 h64;
-#define XXH_get64bits(p) XXH_readLE64_align(p, endian, align)
 
-#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
-    if (p==NULL)
-    {
-        len=0;
-        bEnd=p=(const BYTE*)(size_t)32;
-    }
-#endif
 
-    if (len>=32)
-    {
-        const BYTE* const limit = bEnd - 32;
-        U64 v1 = seed + PRIME64_1 + PRIME64_2;
-        U64 v2 = seed + PRIME64_2;
-        U64 v3 = seed + 0;
-        U64 v4 = seed - PRIME64_1;
-
-        do
-        {
-            v1 += XXH_get64bits(p) * PRIME64_2;
-            p+=8;
-            v1 = XXH_rotl64(v1, 31);
-            v1 *= PRIME64_1;
-            v2 += XXH_get64bits(p) * PRIME64_2;
-            p+=8;
-            v2 = XXH_rotl64(v2, 31);
-            v2 *= PRIME64_1;
-            v3 += XXH_get64bits(p) * PRIME64_2;
-            p+=8;
-            v3 = XXH_rotl64(v3, 31);
-            v3 *= PRIME64_1;
-            v4 += XXH_get64bits(p) * PRIME64_2;
-            p+=8;
-            v4 = XXH_rotl64(v4, 31);
-            v4 *= PRIME64_1;
-        }
-        while (p<=limit);
-
-        h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18);
-
-        v1 *= PRIME64_2;
-        v1 = XXH_rotl64(v1, 31);
-        v1 *= PRIME64_1;
-        h64 ^= v1;
-        h64 = h64 * PRIME64_1 + PRIME64_4;
-
-        v2 *= PRIME64_2;
-        v2 = XXH_rotl64(v2, 31);
-        v2 *= PRIME64_1;
-        h64 ^= v2;
-        h64 = h64 * PRIME64_1 + PRIME64_4;
-
-        v3 *= PRIME64_2;
-        v3 = XXH_rotl64(v3, 31);
-        v3 *= PRIME64_1;
-        h64 ^= v3;
-        h64 = h64 * PRIME64_1 + PRIME64_4;
-
-        v4 *= PRIME64_2;
-        v4 = XXH_rotl64(v4, 31);
-        v4 *= PRIME64_1;
-        h64 ^= v4;
-        h64 = h64 * PRIME64_1 + PRIME64_4;
-    }
-    else
-    {
-        h64  = seed + PRIME64_5;
-    }
+/*======   Hash streaming   ======*/
 
-    h64 += (U64) len;
-
-    while (p+8<=bEnd)
-    {
-        U64 k1 = XXH_get64bits(p);
-        k1 *= PRIME64_2;
-        k1 = XXH_rotl64(k1,31);
-        k1 *= PRIME64_1;
-        h64 ^= k1;
-        h64 = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4;
-        p+=8;
-    }
-
-    if (p+4<=bEnd)
-    {
-        h64 ^= (U64)(XXH_get32bits(p)) * PRIME64_1;
-        h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3;
-        p+=4;
-    }
-
-    while (p<bEnd)
-    {
-        h64 ^= (*p) * PRIME64_5;
-        h64 = XXH_rotl64(h64, 11) * PRIME64_1;
-        p++;
-    }
-
-    h64 ^= h64 >> 33;
-    h64 *= PRIME64_2;
-    h64 ^= h64 >> 29;
-    h64 *= PRIME64_3;
-    h64 ^= h64 >> 32;
-
-    return h64;
-}
-
-
-unsigned long long XXH64 (const void* input, size_t len, unsigned long long seed)
+XXH_PUBLIC_API XXH32_state_t* XXH32_createState(void)
 {
-#if 0
-    // Simple version, good for code maintenance, but unfortunately slow for small inputs
-    XXH64_state_t state;
-    XXH64_reset(&state, seed);
-    XXH64_update(&state, input, len);
-    return XXH64_digest(&state);
-#else
-    XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
-
-#  if !defined(XXH_USE_UNALIGNED_ACCESS)
-    if ((((size_t)input) & 7)==0)   // Input is aligned, let's leverage the speed advantage
-    {
-        if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
-            return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
-        else
-            return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
-    }
-#  endif
-
-    if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
-        return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned);
-    else
-        return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned);
-#endif
-}
-
-/****************************************************
- *  Advanced Hash Functions
-****************************************************/
-
-/*** Allocation ***/
-typedef struct
-{
-    U64 total_len;
-    U32 seed;
-    U32 v1;
-    U32 v2;
-    U32 v3;
-    U32 v4;
-    U32 mem32[4];   /* defined as U32 for alignment */
-    U32 memsize;
-} XXH_istate32_t;
-
-typedef struct
-{
-    U64 total_len;
-    U64 seed;
-    U64 v1;
-    U64 v2;
-    U64 v3;
-    U64 v4;
-    U64 mem64[4];   /* defined as U64 for alignment */
-    U32 memsize;
-} XXH_istate64_t;
-
-
-XXH32_state_t* XXH32_createState(void)
-{
-    XXH_STATIC_ASSERT(sizeof(XXH32_state_t) >= sizeof(XXH_istate32_t));   // A compilation error here means XXH32_state_t is not large enough
     return (XXH32_state_t*)XXH_malloc(sizeof(XXH32_state_t));
 }
-XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr)
+XXH_PUBLIC_API XXH_errorcode XXH32_freeState(XXH32_state_t* statePtr)
 {
     XXH_free(statePtr);
     return XXH_OK;
 }
 
-XXH64_state_t* XXH64_createState(void)
+XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* restrict dstState, const XXH32_state_t* restrict srcState)
 {
-    XXH_STATIC_ASSERT(sizeof(XXH64_state_t) >= sizeof(XXH_istate64_t));   // A compilation error here means XXH64_state_t is not large enough
-    return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t));
-}
-XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr)
-{
-    XXH_free(statePtr);
-    return XXH_OK;
-}
-
-
-/*** Hash feed ***/
-
-XXH_errorcode XXH32_reset(XXH32_state_t* state_in, U32 seed)
-{
-    XXH_istate32_t* state = (XXH_istate32_t*) state_in;
-    state->seed = seed;
-    state->v1 = seed + PRIME32_1 + PRIME32_2;
-    state->v2 = seed + PRIME32_2;
-    state->v3 = seed + 0;
-    state->v4 = seed - PRIME32_1;
-    state->total_len = 0;
-    state->memsize = 0;
-    return XXH_OK;
+    memcpy(dstState, srcState, sizeof(*dstState));
 }
 
-XXH_errorcode XXH64_reset(XXH64_state_t* state_in, unsigned long long seed)
+XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int seed)
 {
-    XXH_istate64_t* state = (XXH_istate64_t*) state_in;
-    state->seed = seed;
-    state->v1 = seed + PRIME64_1 + PRIME64_2;
-    state->v2 = seed + PRIME64_2;
-    state->v3 = seed + 0;
-    state->v4 = seed - PRIME64_1;
-    state->total_len = 0;
-    state->memsize = 0;
+    XXH32_state_t state;   /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
+    memset(&state, 0, sizeof(state)-4);   /* do not write into reserved, for future removal */
+    state.v1 = seed + PRIME32_1 + PRIME32_2;
+    state.v2 = seed + PRIME32_2;
+    state.v3 = seed + 0;
+    state.v4 = seed - PRIME32_1;
+    memcpy(statePtr, &state, sizeof(state));
     return XXH_OK;
 }
 
 
-FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state_in, const void* input, size_t len, XXH_endianess endian)
+FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state, const void* input, size_t len, XXH_endianess endian)
 {
-    XXH_istate32_t* state = (XXH_istate32_t *) state_in;
     const BYTE* p = (const BYTE*)input;
     const BYTE* const bEnd = p + len;
 
@@ -588,69 +385,40 @@ FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state_in, const v
     if (input==NULL) return XXH_ERROR;
 #endif
 
-    state->total_len += len;
+    state->total_len_32 += (unsigned)len;
+    state->large_len |= (len>=16) | (state->total_len_32>=16);
 
-    if (state->memsize + len < 16)   // fill in tmp buffer
-    {
+    if (state->memsize + len < 16)  {   /* fill in tmp buffer */
         XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, len);
-        state->memsize += (U32)len;
+        state->memsize += (unsigned)len;
         return XXH_OK;
     }
 
-    if (state->memsize)   // some data left from previous update
-    {
+    if (state->memsize) {   /* some data left from previous update */
         XXH_memcpy((BYTE*)(state->mem32) + state->memsize, input, 16-state->memsize);
-        {
-            const U32* p32 = state->mem32;
-            state->v1 += XXH_readLE32(p32, endian) * PRIME32_2;
-            state->v1 = XXH_rotl32(state->v1, 13);
-            state->v1 *= PRIME32_1;
-            p32++;
-            state->v2 += XXH_readLE32(p32, endian) * PRIME32_2;
-            state->v2 = XXH_rotl32(state->v2, 13);
-            state->v2 *= PRIME32_1;
-            p32++;
-            state->v3 += XXH_readLE32(p32, endian) * PRIME32_2;
-            state->v3 = XXH_rotl32(state->v3, 13);
-            state->v3 *= PRIME32_1;
-            p32++;
-            state->v4 += XXH_readLE32(p32, endian) * PRIME32_2;
-            state->v4 = XXH_rotl32(state->v4, 13);
-            state->v4 *= PRIME32_1;
-            p32++;
+        {   const U32* p32 = state->mem32;
+            state->v1 = XXH32_round(state->v1, XXH_readLE32(p32, endian)); p32++;
+            state->v2 = XXH32_round(state->v2, XXH_readLE32(p32, endian)); p32++;
+            state->v3 = XXH32_round(state->v3, XXH_readLE32(p32, endian)); p32++;
+            state->v4 = XXH32_round(state->v4, XXH_readLE32(p32, endian)); p32++;
         }
         p += 16-state->memsize;
         state->memsize = 0;
     }
 
-    if (p <= bEnd-16)
-    {
+    if (p <= bEnd-16) {
         const BYTE* const limit = bEnd - 16;
         U32 v1 = state->v1;
         U32 v2 = state->v2;
         U32 v3 = state->v3;
         U32 v4 = state->v4;
 
-        do
-        {
-            v1 += XXH_readLE32(p, endian) * PRIME32_2;
-            v1 = XXH_rotl32(v1, 13);
-            v1 *= PRIME32_1;
-            p+=4;
-            v2 += XXH_readLE32(p, endian) * PRIME32_2;
-            v2 = XXH_rotl32(v2, 13);
-            v2 *= PRIME32_1;
-            p+=4;
-            v3 += XXH_readLE32(p, endian) * PRIME32_2;
-            v3 = XXH_rotl32(v3, 13);
-            v3 *= PRIME32_1;
-            p+=4;
-            v4 += XXH_readLE32(p, endian) * PRIME32_2;
-            v4 = XXH_rotl32(v4, 13);
-            v4 *= PRIME32_1;
-            p+=4;
-        }
-        while (p<=limit);
+        do {
+            v1 = XXH32_round(v1, XXH_readLE32(p, endian)); p+=4;
+            v2 = XXH32_round(v2, XXH_readLE32(p, endian)); p+=4;
+            v3 = XXH32_round(v3, XXH_readLE32(p, endian)); p+=4;
+            v4 = XXH32_round(v4, XXH_readLE32(p, endian)); p+=4;
+        } while (p<=limit);
 
         state->v1 = v1;
         state->v2 = v2;
@@ -658,16 +426,15 @@ FORCE_INLINE XXH_errorcode XXH32_update_endian (XXH32_state_t* state_in, const v
         state->v4 = v4;
     }
 
-    if (p < bEnd)
-    {
+    if (p < bEnd) {
         XXH_memcpy(state->mem32, p, (size_t)(bEnd-p));
-        state->memsize = (U32)(bEnd-p);
+        state->memsize = (unsigned)(bEnd-p);
     }
 
     return XXH_OK;
 }
 
-XXH_errorcode XXH32_update (XXH32_state_t* state_in, const void* input, size_t len)
+XXH_PUBLIC_API XXH_errorcode XXH32_update (XXH32_state_t* state_in, const void* input, size_t len)
 {
     XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
 
@@ -679,35 +446,29 @@ XXH_errorcode XXH32_update (XXH32_state_t* state_in, const void* input, size_t l
 
 
 
-FORCE_INLINE U32 XXH32_digest_endian (const XXH32_state_t* state_in, XXH_endianess endian)
+FORCE_INLINE U32 XXH32_digest_endian (const XXH32_state_t* state, XXH_endianess endian)
 {
-    XXH_istate32_t* state = (XXH_istate32_t*) state_in;
     const BYTE * p = (const BYTE*)state->mem32;
-    BYTE* bEnd = (BYTE*)(state->mem32) + state->memsize;
+    const BYTE* const bEnd = (const BYTE*)(state->mem32) + state->memsize;
     U32 h32;
 
-    if (state->total_len >= 16)
-    {
+    if (state->large_len) {
         h32 = XXH_rotl32(state->v1, 1) + XXH_rotl32(state->v2, 7) + XXH_rotl32(state->v3, 12) + XXH_rotl32(state->v4, 18);
-    }
-    else
-    {
-        h32  = state->seed + PRIME32_5;
+    } else {
+        h32 = state->v3 /* == seed */ + PRIME32_5;
     }
 
-    h32 += (U32) state->total_len;
+    h32 += state->total_len_32;
 
-    while (p+4<=bEnd)
-    {
+    while (p+4<=bEnd) {
         h32 += XXH_readLE32(p, endian) * PRIME32_3;
         h32  = XXH_rotl32(h32, 17) * PRIME32_4;
         p+=4;
     }
 
-    while (p<bEnd)
-    {
+    while (p<bEnd) {
         h32 += (*p) * PRIME32_5;
-        h32 = XXH_rotl32(h32, 11) * PRIME32_1;
+        h32  = XXH_rotl32(h32, 11) * PRIME32_1;
         p++;
     }
 
@@ -721,7 +482,7 @@ FORCE_INLINE U32 XXH32_digest_endian (const XXH32_state_t* state_in, XXH_endiane
 }
 
 
-U32 XXH32_digest (const XXH32_state_t* state_in)
+XXH_PUBLIC_API unsigned int XXH32_digest (const XXH32_state_t* state_in)
 {
     XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
 
@@ -732,9 +493,261 @@ U32 XXH32_digest (const XXH32_state_t* state_in)
 }
 
 
-FORCE_INLINE XXH_errorcode XXH64_update_endian (XXH64_state_t* state_in, const void* input, size_t len, XXH_endianess endian)
+/*======   Canonical representation   ======*/
+
+/*! Default XXH result types are basic unsigned 32 and 64 bits.
+*   The canonical representation follows human-readable write convention, aka big-endian (large digits first).
+*   These functions allow transformation of hash result into and from its canonical format.
+*   This way, hash values can be written into a file or buffer, and remain comparable across different systems and programs.
+*/
+
+XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t hash)
+{
+    XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t));
+    if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash);
+    memcpy(dst, &hash, sizeof(*dst));
+}
+
+XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src)
+{
+    return XXH_readBE32(src);
+}
+
+
+#ifndef XXH_NO_LONG_LONG
+
+/* *******************************************************************
+*  64-bits hash functions
+*********************************************************************/
+
+/*======   Memory access   ======*/
+
+#ifndef MEM_MODULE
+# define MEM_MODULE
+# if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
+#   include <stdint.h>
+    typedef uint64_t U64;
+# else
+    typedef unsigned long long U64;   /* if your compiler doesn't support unsigned long long, replace by another 64-bit type here. Note that xxhash.h will also need to be updated. */
+# endif
+#endif
+
+
+#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2))
+
+/* Force direct memory access. Only works on CPU which support unaligned memory access in hardware */
+static U64 XXH_read64(const void* memPtr) { return *(const U64*) memPtr; }
+
+#elif (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==1))
+
+/* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */
+/* currently only defined for gcc and icc */
+typedef union { U32 u32; U64 u64; } __attribute__((packed)) unalign64;
+
+static U64 XXH_read64(const void* ptr) { return ((const unalign64*)ptr)->u64; }
+
+#else
+
+/* portable and safe solution. Generally efficient.
+ * see : http://stackoverflow.com/a/32095106/646947
+ */
+
+static U64 XXH_read64(const void* memPtr)
+{
+    U64 val;
+    memcpy(&val, memPtr, sizeof(val));
+    return val;
+}
+
+#endif   /* XXH_FORCE_DIRECT_MEMORY_ACCESS */
+
+#if defined(_MSC_VER)     /* Visual Studio */
+#  define XXH_swap64 _byteswap_uint64
+#elif GCC_VERSION >= 403
+#  define XXH_swap64 __builtin_bswap64
+#else
+static U64 XXH_swap64 (U64 x)
+{
+    return  ((x << 56) & 0xff00000000000000ULL) |
+            ((x << 40) & 0x00ff000000000000ULL) |
+            ((x << 24) & 0x0000ff0000000000ULL) |
+            ((x << 8)  & 0x000000ff00000000ULL) |
+            ((x >> 8)  & 0x00000000ff000000ULL) |
+            ((x >> 24) & 0x0000000000ff0000ULL) |
+            ((x >> 40) & 0x000000000000ff00ULL) |
+            ((x >> 56) & 0x00000000000000ffULL);
+}
+#endif
+
+FORCE_INLINE U64 XXH_readLE64_align(const void* ptr, XXH_endianess endian, XXH_alignment align)
+{
+    if (align==XXH_unaligned)
+        return endian==XXH_littleEndian ? XXH_read64(ptr) : XXH_swap64(XXH_read64(ptr));
+    else
+        return endian==XXH_littleEndian ? *(const U64*)ptr : XXH_swap64(*(const U64*)ptr);
+}
+
+FORCE_INLINE U64 XXH_readLE64(const void* ptr, XXH_endianess endian)
+{
+    return XXH_readLE64_align(ptr, endian, XXH_unaligned);
+}
+
+static U64 XXH_readBE64(const void* ptr)
+{
+    return XXH_CPU_LITTLE_ENDIAN ? XXH_swap64(XXH_read64(ptr)) : XXH_read64(ptr);
+}
+
+
+/*======   xxh64   ======*/
+
+static const U64 PRIME64_1 = 11400714785074694791ULL;
+static const U64 PRIME64_2 = 14029467366897019727ULL;
+static const U64 PRIME64_3 =  1609587929392839161ULL;
+static const U64 PRIME64_4 =  9650029242287828579ULL;
+static const U64 PRIME64_5 =  2870177450012600261ULL;
+
+static U64 XXH64_round(U64 acc, U64 input)
+{
+    acc += input * PRIME64_2;
+    acc  = XXH_rotl64(acc, 31);
+    acc *= PRIME64_1;
+    return acc;
+}
+
+static U64 XXH64_mergeRound(U64 acc, U64 val)
+{
+    val  = XXH64_round(0, val);
+    acc ^= val;
+    acc  = acc * PRIME64_1 + PRIME64_4;
+    return acc;
+}
+
+FORCE_INLINE U64 XXH64_endian_align(const void* input, size_t len, U64 seed, XXH_endianess endian, XXH_alignment align)
+{
+    const BYTE* p = (const BYTE*)input;
+    const BYTE* const bEnd = p + len;
+    U64 h64;
+#define XXH_get64bits(p) XXH_readLE64_align(p, endian, align)
+
+#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
+    if (p==NULL) {
+        len=0;
+        bEnd=p=(const BYTE*)(size_t)32;
+    }
+#endif
+
+    if (len>=32) {
+        const BYTE* const limit = bEnd - 32;
+        U64 v1 = seed + PRIME64_1 + PRIME64_2;
+        U64 v2 = seed + PRIME64_2;
+        U64 v3 = seed + 0;
+        U64 v4 = seed - PRIME64_1;
+
+        do {
+            v1 = XXH64_round(v1, XXH_get64bits(p)); p+=8;
+            v2 = XXH64_round(v2, XXH_get64bits(p)); p+=8;
+            v3 = XXH64_round(v3, XXH_get64bits(p)); p+=8;
+            v4 = XXH64_round(v4, XXH_get64bits(p)); p+=8;
+        } while (p<=limit);
+
+        h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18);
+        h64 = XXH64_mergeRound(h64, v1);
+        h64 = XXH64_mergeRound(h64, v2);
+        h64 = XXH64_mergeRound(h64, v3);
+        h64 = XXH64_mergeRound(h64, v4);
+
+    } else {
+        h64  = seed + PRIME64_5;
+    }
+
+    h64 += (U64) len;
+
+    while (p+8<=bEnd) {
+        U64 const k1 = XXH64_round(0, XXH_get64bits(p));
+        h64 ^= k1;
+        h64  = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4;
+        p+=8;
+    }
+
+    if (p+4<=bEnd) {
+        h64 ^= (U64)(XXH_get32bits(p)) * PRIME64_1;
+        h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3;
+        p+=4;
+    }
+
+    while (p<bEnd) {
+        h64 ^= (*p) * PRIME64_5;
+        h64 = XXH_rotl64(h64, 11) * PRIME64_1;
+        p++;
+    }
+
+    h64 ^= h64 >> 33;
+    h64 *= PRIME64_2;
+    h64 ^= h64 >> 29;
+    h64 *= PRIME64_3;
+    h64 ^= h64 >> 32;
+
+    return h64;
+}
+
+
+XXH_PUBLIC_API unsigned long long XXH64 (const void* input, size_t len, unsigned long long seed)
+{
+#if 0
+    /* Simple version, good for code maintenance, but unfortunately slow for small inputs */
+    XXH64_CREATESTATE_STATIC(state);
+    XXH64_reset(state, seed);
+    XXH64_update(state, input, len);
+    return XXH64_digest(state);
+#else
+    XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
+
+    if (XXH_FORCE_ALIGN_CHECK) {
+        if ((((size_t)input) & 7)==0) {  /* Input is aligned, let's leverage the speed advantage */
+            if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
+                return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
+            else
+                return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
+    }   }
+
+    if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
+        return XXH64_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned);
+    else
+        return XXH64_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned);
+#endif
+}
+
+/*======   Hash Streaming   ======*/
+
+XXH_PUBLIC_API XXH64_state_t* XXH64_createState(void)
+{
+    return (XXH64_state_t*)XXH_malloc(sizeof(XXH64_state_t));
+}
+XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr)
+{
+    XXH_free(statePtr);
+    return XXH_OK;
+}
+
+XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* restrict dstState, const XXH64_state_t* restrict srcState)
+{
+    memcpy(dstState, srcState, sizeof(*dstState));
+}
+
+XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long long seed)
+{
+    XXH64_state_t state;   /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
+    memset(&state, 0, sizeof(state)-8);   /* do not write into reserved, for future removal */
+    state.v1 = seed + PRIME64_1 + PRIME64_2;
+    state.v2 = seed + PRIME64_2;
+    state.v3 = seed + 0;
+    state.v4 = seed - PRIME64_1;
+    memcpy(statePtr, &state, sizeof(state));
+    return XXH_OK;
+}
+
+FORCE_INLINE XXH_errorcode XXH64_update_endian (XXH64_state_t* state, const void* input, size_t len, XXH_endianess endian)
 {
-    XXH_istate64_t * state = (XXH_istate64_t *) state_in;
     const BYTE* p = (const BYTE*)input;
     const BYTE* const bEnd = p + len;
 
@@ -744,67 +757,35 @@ FORCE_INLINE XXH_errorcode XXH64_update_endian (XXH64_state_t* state_in, const v
 
     state->total_len += len;
 
-    if (state->memsize + len < 32)   // fill in tmp buffer
-    {
+    if (state->memsize + len < 32) {  /* fill in tmp buffer */
         XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, len);
         state->memsize += (U32)len;
         return XXH_OK;
     }
 
-    if (state->memsize)   // some data left from previous update
-    {
+    if (state->memsize) {   /* tmp buffer is full */
         XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, 32-state->memsize);
-        {
-            const U64* p64 = state->mem64;
-            state->v1 += XXH_readLE64(p64, endian) * PRIME64_2;
-            state->v1 = XXH_rotl64(state->v1, 31);
-            state->v1 *= PRIME64_1;
-            p64++;
-            state->v2 += XXH_readLE64(p64, endian) * PRIME64_2;
-            state->v2 = XXH_rotl64(state->v2, 31);
-            state->v2 *= PRIME64_1;
-            p64++;
-            state->v3 += XXH_readLE64(p64, endian) * PRIME64_2;
-            state->v3 = XXH_rotl64(state->v3, 31);
-            state->v3 *= PRIME64_1;
-            p64++;
-            state->v4 += XXH_readLE64(p64, endian) * PRIME64_2;
-            state->v4 = XXH_rotl64(state->v4, 31);
-            state->v4 *= PRIME64_1;
-            p64++;
-        }
+        state->v1 = XXH64_round(state->v1, XXH_readLE64(state->mem64+0, endian));
+        state->v2 = XXH64_round(state->v2, XXH_readLE64(state->mem64+1, endian));
+        state->v3 = XXH64_round(state->v3, XXH_readLE64(state->mem64+2, endian));
+        state->v4 = XXH64_round(state->v4, XXH_readLE64(state->mem64+3, endian));
         p += 32-state->memsize;
         state->memsize = 0;
     }
 
-    if (p+32 <= bEnd)
-    {
+    if (p+32 <= bEnd) {
         const BYTE* const limit = bEnd - 32;
         U64 v1 = state->v1;
         U64 v2 = state->v2;
         U64 v3 = state->v3;
         U64 v4 = state->v4;
 
-        do
-        {
-            v1 += XXH_readLE64(p, endian) * PRIME64_2;
-            v1 = XXH_rotl64(v1, 31);
-            v1 *= PRIME64_1;
-            p+=8;
-            v2 += XXH_readLE64(p, endian) * PRIME64_2;
-            v2 = XXH_rotl64(v2, 31);
-            v2 *= PRIME64_1;
-            p+=8;
-            v3 += XXH_readLE64(p, endian) * PRIME64_2;
-            v3 = XXH_rotl64(v3, 31);
-            v3 *= PRIME64_1;
-            p+=8;
-            v4 += XXH_readLE64(p, endian) * PRIME64_2;
-            v4 = XXH_rotl64(v4, 31);
-            v4 *= PRIME64_1;
-            p+=8;
-        }
-        while (p<=limit);
+        do {
+            v1 = XXH64_round(v1, XXH_readLE64(p, endian)); p+=8;
+            v2 = XXH64_round(v2, XXH_readLE64(p, endian)); p+=8;
+            v3 = XXH64_round(v3, XXH_readLE64(p, endian)); p+=8;
+            v4 = XXH64_round(v4, XXH_readLE64(p, endian)); p+=8;
+        } while (p<=limit);
 
         state->v1 = v1;
         state->v2 = v2;
@@ -812,16 +793,15 @@ FORCE_INLINE XXH_errorcode XXH64_update_endian (XXH64_state_t* state_in, const v
         state->v4 = v4;
     }
 
-    if (p < bEnd)
-    {
+    if (p < bEnd) {
         XXH_memcpy(state->mem64, p, (size_t)(bEnd-p));
-        state->memsize = (U32)(bEnd-p);
+        state->memsize = (unsigned)(bEnd-p);
     }
 
     return XXH_OK;
 }
 
-XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void* input, size_t len)
+XXH_PUBLIC_API XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void* input, size_t len)
 {
     XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
 
@@ -831,77 +811,45 @@ XXH_errorcode XXH64_update (XXH64_state_t* state_in, const void* input, size_t l
         return XXH64_update_endian(state_in, input, len, XXH_bigEndian);
 }
 
-
-
-FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state_in, XXH_endianess endian)
+FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state, XXH_endianess endian)
 {
-    XXH_istate64_t * state = (XXH_istate64_t *) state_in;
     const BYTE * p = (const BYTE*)state->mem64;
-    BYTE* bEnd = (BYTE*)state->mem64 + state->memsize;
+    const BYTE* const bEnd = (const BYTE*)state->mem64 + state->memsize;
     U64 h64;
 
-    if (state->total_len >= 32)
-    {
-        U64 v1 = state->v1;
-        U64 v2 = state->v2;
-        U64 v3 = state->v3;
-        U64 v4 = state->v4;
+    if (state->total_len >= 32) {
+        U64 const v1 = state->v1;
+        U64 const v2 = state->v2;
+        U64 const v3 = state->v3;
+        U64 const v4 = state->v4;
 
         h64 = XXH_rotl64(v1, 1) + XXH_rotl64(v2, 7) + XXH_rotl64(v3, 12) + XXH_rotl64(v4, 18);
-
-        v1 *= PRIME64_2;
-        v1 = XXH_rotl64(v1, 31);
-        v1 *= PRIME64_1;
-        h64 ^= v1;
-        h64 = h64*PRIME64_1 + PRIME64_4;
-
-        v2 *= PRIME64_2;
-        v2 = XXH_rotl64(v2, 31);
-        v2 *= PRIME64_1;
-        h64 ^= v2;
-        h64 = h64*PRIME64_1 + PRIME64_4;
-
-        v3 *= PRIME64_2;
-        v3 = XXH_rotl64(v3, 31);
-        v3 *= PRIME64_1;
-        h64 ^= v3;
-        h64 = h64*PRIME64_1 + PRIME64_4;
-
-        v4 *= PRIME64_2;
-        v4 = XXH_rotl64(v4, 31);
-        v4 *= PRIME64_1;
-        h64 ^= v4;
-        h64 = h64*PRIME64_1 + PRIME64_4;
-    }
-    else
-    {
-        h64  = state->seed + PRIME64_5;
+        h64 = XXH64_mergeRound(h64, v1);
+        h64 = XXH64_mergeRound(h64, v2);
+        h64 = XXH64_mergeRound(h64, v3);
+        h64 = XXH64_mergeRound(h64, v4);
+    } else {
+        h64  = state->v3 + PRIME64_5;
     }
 
     h64 += (U64) state->total_len;
 
-    while (p+8<=bEnd)
-    {
-        U64 k1 = XXH_readLE64(p, endian);
-        k1 *= PRIME64_2;
-        k1 = XXH_rotl64(k1,31);
-        k1 *= PRIME64_1;
+    while (p+8<=bEnd) {
+        U64 const k1 = XXH64_round(0, XXH_readLE64(p, endian));
         h64 ^= k1;
-        h64 = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4;
+        h64  = XXH_rotl64(h64,27) * PRIME64_1 + PRIME64_4;
         p+=8;
     }
 
-    if (p+4<=bEnd)
-    {
+    if (p+4<=bEnd) {
         h64 ^= (U64)(XXH_readLE32(p, endian)) * PRIME64_1;
-        h64 = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3;
+        h64  = XXH_rotl64(h64, 23) * PRIME64_2 + PRIME64_3;
         p+=4;
     }
 
-    while (p<bEnd)
-    {
+    while (p<bEnd) {
         h64 ^= (*p) * PRIME64_5;
-        h64 = XXH_rotl64(h64, 11) * PRIME64_1;
+        h64  = XXH_rotl64(h64, 11) * PRIME64_1;
         p++;
     }
 
@@ -914,8 +862,7 @@ FORCE_INLINE U64 XXH64_digest_endian (const XXH64_state_t* state_in, XXH_endiane
     return h64;
 }
 
-
-unsigned long long XXH64_digest (const XXH64_state_t* state_in)
+XXH_PUBLIC_API unsigned long long XXH64_digest (const XXH64_state_t* state_in)
 {
     XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
 
@@ -925,3 +872,19 @@ unsigned long long XXH64_digest (const XXH64_state_t* state_in)
         return XXH64_digest_endian(state_in, XXH_bigEndian);
 }
 
+
+/*====== Canonical representation   ======*/
+
+XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash)
+{
+    XXH_STATIC_ASSERT(sizeof(XXH64_canonical_t) == sizeof(XXH64_hash_t));
+    if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap64(hash);
+    memcpy(dst, &hash, sizeof(*dst));
+}
+
+XXH_PUBLIC_API XXH64_hash_t XXH64_hashFromCanonical(const XXH64_canonical_t* src)
+{
+    return XXH_readBE64(src);
+}
+
+#endif  /* XXH_NO_LONG_LONG */