Upgrade LZ4 to latest version.

ext/lz4/lz4.c

@@ -1,822 +1,1249 @@
-/*
-   LZ4 - Fast LZ compression algorithm
-   Copyright (C) 2011-2013, 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 :
-   - LZ4 source repository : http://code.google.com/p/lz4/
-   - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
-*/
-
-//**************************************
-// Tuning parameters
-//**************************************
-// MEMORY_USAGE :
-// Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.)
-// Increasing memory usage improves compression ratio
-// Reduced memory usage can improve speed, due to cache effect
-// Default value is 14, for 16KB, which nicely fits into Intel x86 L1 cache
-#define MEMORY_USAGE 14
-
-// HEAPMODE :
-// Select how default compression functions will allocate memory for their hash table,
-// in memory stack (0:default, fastest), or in memory heap (1:requires memory allocation (malloc)).
-#define HEAPMODE 0
-
-
-//**************************************
-// CPU Feature Detection
-//**************************************
-// 32 or 64 bits ?
-#if (defined(__x86_64__) || defined(_M_X64) || defined(_WIN64) \
-  || defined(__powerpc64__) || defined(__ppc64__) || defined(__PPC64__) \
-  || defined(__64BIT__) || defined(_LP64) || defined(__LP64__) \
-  || defined(__ia64) || defined(__itanium__) || defined(_M_IA64) )   // Detects 64 bits mode
-#  define LZ4_ARCH64 1
-#else
-#  define LZ4_ARCH64 0
-#endif
-
-// Little Endian or Big Endian ?
-// Overwrite the #define below if you know your architecture endianess
-#if defined (__GLIBC__)
-#  include <endian.h>
-#  if (__BYTE_ORDER == __BIG_ENDIAN)
-#     define LZ4_BIG_ENDIAN 1
-#  endif
-#elif (defined(__BIG_ENDIAN__) || defined(__BIG_ENDIAN) || defined(_BIG_ENDIAN)) && !(defined(__LITTLE_ENDIAN__) || defined(__LITTLE_ENDIAN) || defined(_LITTLE_ENDIAN))
-#  define LZ4_BIG_ENDIAN 1
-#elif defined(__sparc) || defined(__sparc__) \
-   || defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) \
-   || defined(__hpux)  || defined(__hppa) \
-   || defined(_MIPSEB) || defined(__s390__)
-#  define LZ4_BIG_ENDIAN 1
-#else
-// Little Endian assumed. PDP Endian and other very rare endian format are unsupported.
-#endif
-
-// Unaligned memory access is automatically enabled for "common" CPU, such as x86.
-// For others CPU, such as ARM, the compiler may be more cautious, inserting unnecessary extra code to ensure aligned access property
-// If you know your target CPU supports unaligned memory access, you want to force this option manually to improve performance
-#if defined(__ARM_FEATURE_UNALIGNED)
-#  define LZ4_FORCE_UNALIGNED_ACCESS 1
-#endif
-
-// Define this parameter if your target system or compiler does not support hardware bit count
-#if defined(_MSC_VER) && defined(_WIN32_WCE)            // Visual Studio for Windows CE does not support Hardware bit count
-#  define LZ4_FORCE_SW_BITCOUNT
-#endif
-
-// BIG_ENDIAN_NATIVE_BUT_INCOMPATIBLE :
-// This option may provide a small boost to performance for some big endian cpu, although probably modest.
-// You may set this option to 1 if data will remain within closed environment.
-// This option is useless on Little_Endian CPU (such as x86)
-//#define BIG_ENDIAN_NATIVE_BUT_INCOMPATIBLE 1
-
-
-//**************************************
-// Compiler Options
-//**************************************
-#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L)   // C99
-/* "restrict" is a known keyword */
-#else
-#  define restrict // Disable restrict
-#endif
-
-#ifdef _MSC_VER    // Visual Studio
-#  define FORCE_INLINE static __forceinline
-#  include <intrin.h>                    // For Visual 2005
-#  if LZ4_ARCH64   // 64-bits
-#    pragma intrinsic(_BitScanForward64) // For Visual 2005
-#    pragma intrinsic(_BitScanReverse64) // For Visual 2005
-#  else            // 32-bits
-#    pragma intrinsic(_BitScanForward)   // For Visual 2005
-#    pragma intrinsic(_BitScanReverse)   // For Visual 2005
-#  endif
-#  pragma warning(disable : 4127)        // disable: C4127: conditional expression is constant
-#else 
-#  ifdef __GNUC__
-#    define FORCE_INLINE static inline __attribute__((always_inline))
-#  else
-#    define FORCE_INLINE static inline
-#  endif
-#endif
-
-#ifdef _MSC_VER
-#  define lz4_bswap16(x) _byteswap_ushort(x)
-#else
-#  define lz4_bswap16(x) ((unsigned short int) ((((x) >> 8) & 0xffu) | (((x) & 0xffu) << 8)))
-#endif
-
-#define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
-
-#if (GCC_VERSION >= 302) || (__INTEL_COMPILER >= 800) || defined(__clang__)
-#  define expect(expr,value)    (__builtin_expect ((expr),(value)) )
-#else
-#  define expect(expr,value)    (expr)
-#endif
-
-#define likely(expr)     expect((expr) != 0, 1)
-#define unlikely(expr)   expect((expr) != 0, 0)
-
-
-//**************************************
-// Memory routines
-//**************************************
-#include <stdlib.h>   // malloc, calloc, free
-#define ALLOCATOR(n,s) calloc(n,s)
-#define FREEMEM        free
-#include <string.h>   // memset, memcpy
-#define MEM_INIT       memset
-
-
-//**************************************
-// Includes
-//**************************************
-#include "lz4.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(LZ4_FORCE_UNALIGNED_ACCESS)
-#  define _PACKED __attribute__ ((packed))
-#else
-#  define _PACKED
-#endif
-
-#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
-#  if defined(__IBMC__) || defined(__SUNPRO_C) || defined(__SUNPRO_CC)
-#    pragma pack(1)
-#  else
-#    pragma pack(push, 1)
-#  endif
-#endif
-
-typedef struct { U16 v; }  _PACKED U16_S;
-typedef struct { U32 v; }  _PACKED U32_S;
-typedef struct { U64 v; }  _PACKED U64_S;
-typedef struct {size_t v;} _PACKED size_t_S;
-
-#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
-#  if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
-#    pragma pack(0)
-#  else
-#    pragma pack(pop)
-#  endif
-#endif
-
-#define A16(x)   (((U16_S *)(x))->v)
-#define A32(x)   (((U32_S *)(x))->v)
-#define A64(x)   (((U64_S *)(x))->v)
-#define AARCH(x) (((size_t_S *)(x))->v)
-
-
-//**************************************
-// Constants
-//**************************************
-#define LZ4_HASHLOG   (MEMORY_USAGE-2)
-#define HASHTABLESIZE (1 << MEMORY_USAGE)
-#define HASHNBCELLS4  (1 << LZ4_HASHLOG)
-
-#define MINMATCH 4
-
-#define COPYLENGTH 8
-#define LASTLITERALS 5
-#define MFLIMIT (COPYLENGTH+MINMATCH)
-const int LZ4_minLength = (MFLIMIT+1);
-
-#define LZ4_64KLIMIT ((1<<16) + (MFLIMIT-1))
-#define SKIPSTRENGTH 6     // Increasing this value will make the compression run slower on incompressible data
-
-#define MAXD_LOG 16
-#define MAX_DISTANCE ((1 << MAXD_LOG) - 1)
-
-#define ML_BITS  4
-#define ML_MASK  ((1U<<ML_BITS)-1)
-#define RUN_BITS (8-ML_BITS)
-#define RUN_MASK ((1U<<RUN_BITS)-1)
-
-#define KB *(1U<<10)
-#define MB *(1U<<20)
-#define GB *(1U<<30)
-
-
-//**************************************
-// Structures and local types
-//**************************************
-
-typedef struct {
-    U32 hashTable[HASHNBCELLS4];
-    const BYTE* bufferStart;
-    const BYTE* base;
-    const BYTE* nextBlock;
-} LZ4_Data_Structure;
-
-typedef enum { notLimited = 0, limited = 1 } limitedOutput_directive;
-typedef enum { byPtr, byU32, byU16 } tableType_t;
-
-typedef enum { noPrefix = 0, withPrefix = 1 } prefix64k_directive;
-
-typedef enum { endOnOutputSize = 0, endOnInputSize = 1 } endCondition_directive;
-typedef enum { full = 0, partial = 1 } earlyEnd_directive;
-
-
-//**************************************
-// Architecture-specific macros
-//**************************************
-#define STEPSIZE                  sizeof(size_t)
-#define LZ4_COPYSTEP(d,s)         { AARCH(d) = AARCH(s); d+=STEPSIZE; s+=STEPSIZE; }
-#define LZ4_COPY8(d,s)            { LZ4_COPYSTEP(d,s); if (STEPSIZE<8) LZ4_COPYSTEP(d,s); }
-#define LZ4_SECURECOPY(d,s,e)     { if ((STEPSIZE==4)||(d<e)) LZ4_WILDCOPY(d,s,e); }
-
-#if LZ4_ARCH64   // 64-bit
-#  define HTYPE                   U32
-#  define INITBASE(base)          const BYTE* const base = ip
-#else            // 32-bit
-#  define HTYPE                   const BYTE*
-#  define INITBASE(base)          const int base = 0
-#endif
-
-#if (defined(LZ4_BIG_ENDIAN) && !defined(BIG_ENDIAN_NATIVE_BUT_INCOMPATIBLE))
-#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { U16 v = A16(p); v = lz4_bswap16(v); d = (s) - v; }
-#  define LZ4_WRITE_LITTLEENDIAN_16(p,i)  { U16 v = (U16)(i); v = lz4_bswap16(v); A16(p) = v; p+=2; }
-#else      // Little Endian
-#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { d = (s) - A16(p); }
-#  define LZ4_WRITE_LITTLEENDIAN_16(p,v)  { A16(p) = v; p+=2; }
-#endif
-
-
-//**************************************
-// Macros
-//**************************************
-#define LZ4_WILDCOPY(d,s,e)     { do { LZ4_COPY8(d,s) } while (d<e); }           // at the end, d>=e;
-
-
-//****************************
-// Private functions
-//****************************
-#if LZ4_ARCH64
-
-FORCE_INLINE int LZ4_NbCommonBytes (register U64 val)
-{
-# if defined(LZ4_BIG_ENDIAN)
-#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r = 0;
-    _BitScanReverse64( &r, val );
-    return (int)(r>>3);
-#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_clzll(val) >> 3);
-#   else
-    int r;
-    if (!(val>>32)) { r=4; } else { r=0; val>>=32; }
-    if (!(val>>16)) { r+=2; val>>=8; } else { val>>=24; }
-    r += (!val);
-    return r;
-#   endif
-# else
-#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r = 0;
-    _BitScanForward64( &r, val );
-    return (int)(r>>3);
-#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_ctzll(val) >> 3);
-#   else
-    static const int DeBruijnBytePos[64] = { 0, 0, 0, 0, 0, 1, 1, 2, 0, 3, 1, 3, 1, 4, 2, 7, 0, 2, 3, 6, 1, 5, 3, 5, 1, 3, 4, 4, 2, 5, 6, 7, 7, 0, 1, 2, 3, 3, 4, 6, 2, 6, 5, 5, 3, 4, 5, 6, 7, 1, 2, 4, 6, 4, 4, 5, 7, 2, 6, 5, 7, 6, 7, 7 };
-    return DeBruijnBytePos[((U64)((val & -(long long)val) * 0x0218A392CDABBD3FULL)) >> 58];
-#   endif
-# endif
-}
-
-#else
-
-FORCE_INLINE int LZ4_NbCommonBytes (register U32 val)
-{
-# if defined(LZ4_BIG_ENDIAN)
-#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r = 0;
-    _BitScanReverse( &r, val );
-    return (int)(r>>3);
-#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_clz(val) >> 3);
-#   else
-    int r;
-    if (!(val>>16)) { r=2; val>>=8; } else { r=0; val>>=24; }
-    r += (!val);
-    return r;
-#   endif
-# else
-#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r;
-    _BitScanForward( &r, val );
-    return (int)(r>>3);
-#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_ctz(val) >> 3);
-#   else
-    static const int DeBruijnBytePos[32] = { 0, 0, 3, 0, 3, 1, 3, 0, 3, 2, 2, 1, 3, 2, 0, 1, 3, 3, 1, 2, 2, 2, 2, 0, 3, 1, 2, 0, 1, 0, 1, 1 };
-    return DeBruijnBytePos[((U32)((val & -(S32)val) * 0x077CB531U)) >> 27];
-#   endif
-# endif
-}
-
-#endif
-
-
-//****************************
-// Compression functions
-//****************************
-FORCE_INLINE int LZ4_hashSequence(U32 sequence, tableType_t tableType)
-{
-    if (tableType == byU16)
-        return (((sequence) * 2654435761U) >> ((MINMATCH*8)-(LZ4_HASHLOG+1)));
-    else
-        return (((sequence) * 2654435761U) >> ((MINMATCH*8)-LZ4_HASHLOG));
-}
-
-FORCE_INLINE int LZ4_hashPosition(const BYTE* p, tableType_t tableType) { return LZ4_hashSequence(A32(p), tableType); }
-
-FORCE_INLINE void LZ4_putPositionOnHash(const BYTE* p, U32 h, void* tableBase, tableType_t tableType, const BYTE* srcBase)
-{
-    switch (tableType)
-    {
-    case byPtr: { const BYTE** hashTable = (const BYTE**) tableBase; hashTable[h] = p; break; }
-    case byU32: { U32* hashTable = (U32*) tableBase; hashTable[h] = (U32)(p-srcBase); break; }
-    case byU16: { U16* hashTable = (U16*) tableBase; hashTable[h] = (U16)(p-srcBase); break; }
-    }
-}
-
-FORCE_INLINE void LZ4_putPosition(const BYTE* p, void* tableBase, tableType_t tableType, const BYTE* srcBase)
-{
-    U32 h = LZ4_hashPosition(p, tableType);
-    LZ4_putPositionOnHash(p, h, tableBase, tableType, srcBase);
-}
-
-FORCE_INLINE const BYTE* LZ4_getPositionOnHash(U32 h, void* tableBase, tableType_t tableType, const BYTE* srcBase)
-{
-    if (tableType == byPtr) { const BYTE** hashTable = (const BYTE**) tableBase; return hashTable[h]; }
-    if (tableType == byU32) { U32* hashTable = (U32*) tableBase; return hashTable[h] + srcBase; }
-    { U16* hashTable = (U16*) tableBase; return hashTable[h] + srcBase; }   // default, to ensure a return
-}
-
-FORCE_INLINE const BYTE* LZ4_getPosition(const BYTE* p, void* tableBase, tableType_t tableType, const BYTE* srcBase)
-{
-    U32 h = LZ4_hashPosition(p, tableType);
-    return LZ4_getPositionOnHash(h, tableBase, tableType, srcBase);
-}
-
-
-FORCE_INLINE int LZ4_compress_generic(
-                 void* ctx,
-                 const char* source,
-                 char* dest,
-                 int inputSize,
-                 int maxOutputSize,
-
-                 limitedOutput_directive limitedOutput,
-                 tableType_t tableType,
-                 prefix64k_directive prefix)
-{
-    const BYTE* ip = (const BYTE*) source;
-    const BYTE* const base = (prefix==withPrefix) ? ((LZ4_Data_Structure*)ctx)->base : (const BYTE*) source;
-    const BYTE* const lowLimit = ((prefix==withPrefix) ? ((LZ4_Data_Structure*)ctx)->bufferStart : (const BYTE*)source);
-    const BYTE* anchor = (const BYTE*) source;
-    const BYTE* const iend = ip + inputSize;
-    const BYTE* const mflimit = iend - MFLIMIT;
-    const BYTE* const matchlimit = iend - LASTLITERALS;
-
-    BYTE* op = (BYTE*) dest;
-    BYTE* const oend = op + maxOutputSize;
-
-    int length;
-    const int skipStrength = SKIPSTRENGTH;
-    U32 forwardH;
-
-    // Init conditions
-    if ((U32)inputSize > (U32)LZ4_MAX_INPUT_SIZE) return 0;                                // Unsupported input size, too large (or negative)
-    if ((prefix==withPrefix) && (ip != ((LZ4_Data_Structure*)ctx)->nextBlock)) return 0;   // must continue from end of previous block
-    if (prefix==withPrefix) ((LZ4_Data_Structure*)ctx)->nextBlock=iend;                    // do it now, due to potential early exit
-    if ((tableType == byU16) && (inputSize>=LZ4_64KLIMIT)) return 0;                       // Size too large (not within 64K limit)
-    if (inputSize<LZ4_minLength) goto _last_literals;                                      // Input too small, no compression (all literals)
-
-    // First Byte
-    LZ4_putPosition(ip, ctx, tableType, base);
-    ip++; forwardH = LZ4_hashPosition(ip, tableType);
-
-    // Main Loop
-    for ( ; ; )
-    {
-        int findMatchAttempts = (1U << skipStrength) + 3;
-        const BYTE* forwardIp = ip;
-        const BYTE* ref;
-        BYTE* token;
-
-        // Find a match
-        do {
-            U32 h = forwardH;
-            int step = findMatchAttempts++ >> skipStrength;
-            ip = forwardIp;
-            forwardIp = ip + step;
-
-            if unlikely(forwardIp > mflimit) { goto _last_literals; }
-
-            forwardH = LZ4_hashPosition(forwardIp, tableType);
-            ref = LZ4_getPositionOnHash(h, ctx, tableType, base);
-            LZ4_putPositionOnHash(ip, h, ctx, tableType, base);
-
-        } while ((ref + MAX_DISTANCE < ip) || (A32(ref) != A32(ip)));
-
-        // Catch up
-        while ((ip>anchor) && (ref > lowLimit) && unlikely(ip[-1]==ref[-1])) { ip--; ref--; }
-
-        // Encode Literal length
-        length = (int)(ip - anchor);
-        token = op++;
-        if ((limitedOutput) && unlikely(op + length + (2 + 1 + LASTLITERALS) + (length>>8) > oend)) return 0;   // Check output limit
-        if (length>=(int)RUN_MASK) 
-        { 
-            int len = length-RUN_MASK; 
-            *token=(RUN_MASK<<ML_BITS); 
-            for(; len >= 255 ; len-=255) *op++ = 255; 
-            *op++ = (BYTE)len; 
-        }
-        else *token = (BYTE)(length<<ML_BITS);
-
-        // Copy Literals
-        { BYTE* end=(op)+(length); LZ4_WILDCOPY(op,anchor,end); op=end; }
-
-_next_match:
-        // Encode Offset
-        LZ4_WRITE_LITTLEENDIAN_16(op,(U16)(ip-ref));
-
-        // Start Counting
-        ip+=MINMATCH; ref+=MINMATCH;    // MinMatch already verified
-        anchor = ip;
-        while likely(ip<matchlimit-(STEPSIZE-1))
-        {
-            size_t diff = AARCH(ref) ^ AARCH(ip);
-            if (!diff) { ip+=STEPSIZE; ref+=STEPSIZE; continue; }
-            ip += LZ4_NbCommonBytes(diff);
-            goto _endCount;
-        }
-        if (LZ4_ARCH64) if ((ip<(matchlimit-3)) && (A32(ref) == A32(ip))) { ip+=4; ref+=4; }
-        if ((ip<(matchlimit-1)) && (A16(ref) == A16(ip))) { ip+=2; ref+=2; }
-        if ((ip<matchlimit) && (*ref == *ip)) ip++;
-_endCount:
-
-        // Encode MatchLength
-        length = (int)(ip - anchor);
-        if ((limitedOutput) && unlikely(op + (1 + LASTLITERALS) + (length>>8) > oend)) return 0;    // Check output limit
-        if (length>=(int)ML_MASK) 
-        { 
-            *token += ML_MASK; 
-            length -= ML_MASK; 
-            for (; length > 509 ; length-=510) { *op++ = 255; *op++ = 255; } 
-            if (length >= 255) { length-=255; *op++ = 255; } 
-            *op++ = (BYTE)length; 
-        }
-        else *token += (BYTE)(length);
-
-        // Test end of chunk
-        if (ip > mflimit) { anchor = ip;  break; }
-
-        // Fill table
-        LZ4_putPosition(ip-2, ctx, tableType, base);
-
-        // Test next position
-        ref = LZ4_getPosition(ip, ctx, tableType, base); 
-        LZ4_putPosition(ip, ctx, tableType, base);
-        if ((ref + MAX_DISTANCE >= ip) && (A32(ref) == A32(ip))) { token = op++; *token=0; goto _next_match; }
-
-        // Prepare next loop
-        anchor = ip++;
-        forwardH = LZ4_hashPosition(ip, tableType);
-    }
-
-_last_literals:
-    // Encode Last Literals
-    {
-        int lastRun = (int)(iend - anchor);
-        if ((limitedOutput) && (((char*)op - dest) + lastRun + 1 + ((lastRun+255-RUN_MASK)/255) > (U32)maxOutputSize)) return 0;  // Check output limit
-        if (lastRun>=(int)RUN_MASK) { *op++=(RUN_MASK<<ML_BITS); lastRun-=RUN_MASK; for(; lastRun >= 255 ; lastRun-=255) *op++ = 255; *op++ = (BYTE) lastRun; }
-        else *op++ = (BYTE)(lastRun<<ML_BITS);
-        memcpy(op, anchor, iend - anchor);
-        op += iend-anchor;
-    }
-
-    // End
-    return (int) (((char*)op)-dest);
-}
-
-
-int LZ4_compress(const char* source, char* dest, int inputSize)
-{
-#if (HEAPMODE)
-    void* ctx = ALLOCATOR(HASHNBCELLS4, 4);   // Aligned on 4-bytes boundaries
-#else
-    U32 ctx[1U<<(MEMORY_USAGE-2)] = {0};           // Ensure data is aligned on 4-bytes boundaries
-#endif
-    int result;
-
-    if (inputSize < (int)LZ4_64KLIMIT)
-        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, 0, notLimited, byU16, noPrefix);
-    else
-        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, 0, notLimited, (sizeof(void*)==8) ? byU32 : byPtr, noPrefix);
-
-#if (HEAPMODE)
-    FREEMEM(ctx);
-#endif
-    return result;
-}
-
-int LZ4_compress_continue (void* LZ4_Data, const char* source, char* dest, int inputSize)
-{
-    return LZ4_compress_generic(LZ4_Data, source, dest, inputSize, 0, notLimited, byU32, withPrefix);
-}
-
-
-int LZ4_compress_limitedOutput(const char* source, char* dest, int inputSize, int maxOutputSize)
-{
-#if (HEAPMODE)
-    void* ctx = ALLOCATOR(HASHNBCELLS4, 4);   // Aligned on 4-bytes boundaries
-#else
-    U32 ctx[1U<<(MEMORY_USAGE-2)] = {0};           // Ensure data is aligned on 4-bytes boundaries
-#endif
-    int result;
-
-    if (inputSize < (int)LZ4_64KLIMIT)
-        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, maxOutputSize, limited, byU16, noPrefix);
-    else
-        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, maxOutputSize, limited, (sizeof(void*)==8) ? byU32 : byPtr, noPrefix);
-
-#if (HEAPMODE)
-    FREEMEM(ctx);
-#endif
-    return result;
-}
-
-int LZ4_compress_limitedOutput_continue (void* LZ4_Data, const char* source, char* dest, int inputSize, int maxOutputSize)
-{
-    return LZ4_compress_generic(LZ4_Data, source, dest, inputSize, maxOutputSize, limited, byU32, withPrefix);
-}
-
-
-//****************************
-// Stream functions
-//****************************
-
-FORCE_INLINE void LZ4_init(LZ4_Data_Structure* lz4ds, const BYTE* base)
-{
-    MEM_INIT(lz4ds->hashTable, 0, sizeof(lz4ds->hashTable));
-    lz4ds->bufferStart = base;
-    lz4ds->base = base;
-    lz4ds->nextBlock = base;
-}
-
-
-void* LZ4_create (const char* inputBuffer)
-{
-    void* lz4ds = ALLOCATOR(1, sizeof(LZ4_Data_Structure));
-    LZ4_init ((LZ4_Data_Structure*)lz4ds, (const BYTE*)inputBuffer);
-    return lz4ds;
-}
-
-
-int LZ4_free (void* LZ4_Data)
-{
-    FREEMEM(LZ4_Data);
-    return (0);
-}
-
-
-char* LZ4_slideInputBuffer (void* LZ4_Data)
-{
-    LZ4_Data_Structure* lz4ds = (LZ4_Data_Structure*)LZ4_Data;
-    size_t delta = lz4ds->nextBlock - (lz4ds->bufferStart + 64 KB);
-
-    if ( (lz4ds->base - delta > lz4ds->base)                          // underflow control
-       || ((size_t)(lz4ds->nextBlock - lz4ds->base) > 0xE0000000) )   // close to 32-bits limit
-    {
-        size_t deltaLimit = (lz4ds->nextBlock - 64 KB) - lz4ds->base;
-        int nH;
-
-        for (nH=0; nH < HASHNBCELLS4; nH++)
-        {
-            if ((size_t)(lz4ds->hashTable[nH]) < deltaLimit) lz4ds->hashTable[nH] = 0;
-            else lz4ds->hashTable[nH] -= (U32)deltaLimit;
-        }
-        memcpy((void*)(lz4ds->bufferStart), (const void*)(lz4ds->nextBlock - 64 KB), 64 KB);
-        lz4ds->base = lz4ds->bufferStart;
-        lz4ds->nextBlock = lz4ds->base + 64 KB;
-    }
-    else
-    {
-		memcpy((void*)(lz4ds->bufferStart), (const void*)(lz4ds->nextBlock - 64 KB), 64 KB);
-		lz4ds->nextBlock -= delta;
-		lz4ds->base -= delta;
-	}
-
-    return (char*)(lz4ds->nextBlock);
-}
-
-
-//****************************
-// Decompression functions
-//****************************
-
-// This generic decompression function cover all use cases.
-// It shall be instanciated several times, using different sets of directives
-// Note that it is essential this generic function is really inlined, 
-// in order to remove useless branches during compilation optimisation.
-FORCE_INLINE int LZ4_decompress_generic(
-                 const char* source,
-                 char* dest,
-                 int inputSize,          //
-                 int outputSize,         // If endOnInput==endOnInputSize, this value is the max size of Output Buffer.
-
-                 int endOnInput,         // endOnOutputSize, endOnInputSize
-                 int prefix64k,          // noPrefix, withPrefix
-                 int partialDecoding,    // full, partial
-                 int targetOutputSize    // only used if partialDecoding==partial
-                 )
-{
-    // Local Variables
-    const BYTE* restrict ip = (const BYTE*) source;
-    const BYTE* ref;
-    const BYTE* const iend = ip + inputSize;
-
-    BYTE* op = (BYTE*) dest;
-    BYTE* const oend = op + outputSize;
-    BYTE* cpy;
-    BYTE* oexit = op + targetOutputSize;
-
-    const size_t dec32table[] = {0, 3, 2, 3, 0, 0, 0, 0};   // static reduces speed for LZ4_decompress_safe() on GCC64
-    static const size_t dec64table[] = {0, 0, 0, (size_t)-1, 0, 1, 2, 3};
-
-
-    // Special cases
-    if ((partialDecoding) && (oexit> oend-MFLIMIT)) oexit = oend-MFLIMIT;                        // targetOutputSize too high => decode everything
-    if ((endOnInput) && unlikely(outputSize==0)) return ((inputSize==1) && (*ip==0)) ? 0 : -1;   // Empty output buffer
-    if ((!endOnInput) && unlikely(outputSize==0)) return (*ip==0?1:-1);
-
-
-    // Main Loop
-    while (1)
-    {
-        unsigned token;
-        size_t length;
-
-        // get runlength
-        token = *ip++;
-        if ((length=(token>>ML_BITS)) == RUN_MASK)
-        { 
-            unsigned s=255; 
-            while (((endOnInput)?ip<iend:1) && (s==255))
-            { 
-                s = *ip++; 
-                length += s; 
-            } 
-        }
-
-        // copy literals
-        cpy = op+length;
-        if (((endOnInput) && ((cpy>(partialDecoding?oexit:oend-MFLIMIT)) || (ip+length>iend-(2+1+LASTLITERALS))) )
-            || ((!endOnInput) && (cpy>oend-COPYLENGTH)))
-        {
-            if (partialDecoding)
-            {
-                if (cpy > oend) goto _output_error;                           // Error : write attempt beyond end of output buffer
-                if ((endOnInput) && (ip+length > iend)) goto _output_error;   // Error : read attempt beyond end of input buffer
-            }
-            else
-            {
-                if ((!endOnInput) && (cpy != oend)) goto _output_error;       // Error : block decoding must stop exactly there
-                if ((endOnInput) && ((ip+length != iend) || (cpy > oend))) goto _output_error;   // Error : input must be consumed
-            }
-            memcpy(op, ip, length);
-            ip += length;
-            op += length;
-            break;                                       // Necessarily EOF, due to parsing restrictions
-        }
-        LZ4_WILDCOPY(op, ip, cpy); ip -= (op-cpy); op = cpy;
-
-        // get offset
-        LZ4_READ_LITTLEENDIAN_16(ref,cpy,ip); ip+=2;
-        if ((prefix64k==noPrefix) && unlikely(ref < (BYTE* const)dest)) goto _output_error;   // Error : offset outside destination buffer
-
-        // get matchlength
-        if ((length=(token&ML_MASK)) == ML_MASK) 
-        { 
-            while ((!endOnInput) || (ip<iend-(LASTLITERALS+1)))   // Ensure enough bytes remain for LASTLITERALS + token
-            {
-                unsigned s = *ip++; 
-                length += s; 
-                if (s==255) continue; 
-                break; 
-            }
-        }
-
-        // copy repeated sequence
-        if unlikely((op-ref)<(int)STEPSIZE)
-        {
-            const size_t dec64 = dec64table[(sizeof(void*)==4) ? 0 : op-ref];
-            op[0] = ref[0];
-            op[1] = ref[1];
-            op[2] = ref[2];
-            op[3] = ref[3];
-            op += 4, ref += 4; ref -= dec32table[op-ref];
-            A32(op) = A32(ref); 
-            op += STEPSIZE-4; ref -= dec64;
-        } else { LZ4_COPYSTEP(op,ref); }
-        cpy = op + length - (STEPSIZE-4);
-
-        if unlikely(cpy>oend-COPYLENGTH-(STEPSIZE-4))
-        {
-            if (cpy > oend-LASTLITERALS) goto _output_error;    // Error : last 5 bytes must be literals
-            LZ4_SECURECOPY(op, ref, (oend-COPYLENGTH));
-            while(op<cpy) *op++=*ref++;
-            op=cpy;
-            continue;
-        }
-        LZ4_WILDCOPY(op, ref, cpy);
-        op=cpy;   // correction
-    }
-
-    // end of decoding
-    if (endOnInput)
-       return (int) (((char*)op)-dest);     // Nb of output bytes decoded
-    else
-       return (int) (((char*)ip)-source);   // Nb of input bytes read
-
-    // Overflow error detected
-_output_error:
-    return (int) (-(((char*)ip)-source))-1;
-}
-
-
-int LZ4_decompress_safe(const char* source, char* dest, int inputSize, int maxOutputSize)
-{
-    return LZ4_decompress_generic(source, dest, inputSize, maxOutputSize, endOnInputSize, noPrefix, full, 0);
-}
-
-int LZ4_decompress_safe_withPrefix64k(const char* source, char* dest, int inputSize, int maxOutputSize)
-{
-    return LZ4_decompress_generic(source, dest, inputSize, maxOutputSize, endOnInputSize, withPrefix, full, 0);
-}
-
-int LZ4_decompress_safe_partial(const char* source, char* dest, int inputSize, int targetOutputSize, int maxOutputSize)
-{
-    return LZ4_decompress_generic(source, dest, inputSize, maxOutputSize, endOnInputSize, noPrefix, partial, targetOutputSize);
-}
-
-int LZ4_decompress_fast_withPrefix64k(const char* source, char* dest, int outputSize)
-{
-    return LZ4_decompress_generic(source, dest, 0, outputSize, endOnOutputSize, withPrefix, full, 0);
-}
-
-int LZ4_decompress_fast(const char* source, char* dest, int outputSize)
-{
-#ifdef _MSC_VER   // This version is faster with Visual
-    return LZ4_decompress_generic(source, dest, 0, outputSize, endOnOutputSize, noPrefix, full, 0);
-#else
-    return LZ4_decompress_generic(source, dest, 0, outputSize, endOnOutputSize, withPrefix, full, 0);
-#endif
-}
-
+/*
+   LZ4 - Fast LZ compression algorithm
+   Copyright (C) 2011-2014, 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 :
+   - LZ4 source repository : http://code.google.com/p/lz4/
+   - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
+*/
+
+/**************************************
+   Tuning parameters
+**************************************/
+/*
+ * HEAPMODE :
+ * Select how default compression functions will allocate memory for their hash table,
+ * in memory stack (0:default, fastest), or in memory heap (1:requires memory allocation (malloc)).
+ */
+#define HEAPMODE 0
+
+
+/**************************************
+   CPU Feature Detection
+**************************************/
+/* 32 or 64 bits ? */
+#if (defined(__x86_64__) || defined(_M_X64) || defined(_WIN64) \
+  || defined(__powerpc64__) || defined(__powerpc64le__) \
+  || defined(__ppc64__) || defined(__ppc64le__) \
+  || defined(__PPC64__) || defined(__PPC64LE__) \
+  || defined(__ia64) || defined(__itanium__) || defined(_M_IA64) )   /* Detects 64 bits mode */
+#  define LZ4_ARCH64 1
+#else
+#  define LZ4_ARCH64 0
+#endif
+#define LZ4_32BITS (sizeof(void*)==4)
+#define LZ4_64BITS (sizeof(void*)==8)
+
+/*
+ * Little Endian or Big Endian ?
+ * Overwrite the #define below if you know your architecture endianess
+ */
+#include <stdlib.h>   /* Apparently required to detect endianess */
+#if defined (__GLIBC__)
+#  include <endian.h>
+#  if (__BYTE_ORDER == __BIG_ENDIAN)
+#     define LZ4_BIG_ENDIAN 1
+#  endif
+#elif (defined(__BIG_ENDIAN__) || defined(__BIG_ENDIAN) || defined(_BIG_ENDIAN)) && !(defined(__LITTLE_ENDIAN__) || defined(__LITTLE_ENDIAN) || defined(_LITTLE_ENDIAN))
+#  define LZ4_BIG_ENDIAN 1
+#elif defined(__sparc) || defined(__sparc__) \
+   || defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) \
+   || defined(__hpux)  || defined(__hppa) \
+   || defined(_MIPSEB) || defined(__s390__)
+#  define LZ4_BIG_ENDIAN 1
+#else
+/* Little Endian assumed. PDP Endian and other very rare endian format are unsupported. */
+#endif
+
+/*
+ * Unaligned memory access is automatically enabled for "common" CPU, such as x86.
+ * For others CPU, such as ARM, the compiler may be more cautious, inserting unnecessary extra code to ensure aligned access property
+ * If you know your target CPU supports unaligned memory access, you want to force this option manually to improve performance
+ */
+#if defined(__ARM_FEATURE_UNALIGNED)
+#  define LZ4_FORCE_UNALIGNED_ACCESS 1
+#endif
+
+/* Define this parameter if your target system or compiler does not support hardware bit count */
+#if defined(_MSC_VER) && defined(_WIN32_WCE)   /* Visual Studio for Windows CE does not support Hardware bit count */
+#  define LZ4_FORCE_SW_BITCOUNT
+#endif
+
+/*
+ * BIG_ENDIAN_NATIVE_BUT_INCOMPATIBLE :
+ * This option may provide a small boost to performance for some big endian cpu, although probably modest.
+ * You may set this option to 1 if data will remain within closed environment.
+ * This option is useless on Little_Endian CPU (such as x86)
+ */
+
+/* #define BIG_ENDIAN_NATIVE_BUT_INCOMPATIBLE 1 */
+
+
+/**************************************
+ Compiler Options
+**************************************/
+#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L)   /* C99 */
+/* "restrict" is a known keyword */
+#else
+#  define restrict /* Disable restrict */
+#endif
+
+#ifdef _MSC_VER    /* Visual Studio */
+#  define FORCE_INLINE static __forceinline
+#  include <intrin.h>                    /* For Visual 2005 */
+#  if LZ4_ARCH64   /* 64-bits */
+#    pragma intrinsic(_BitScanForward64) /* For Visual 2005 */
+#    pragma intrinsic(_BitScanReverse64) /* For Visual 2005 */
+#  else            /* 32-bits */
+#    pragma intrinsic(_BitScanForward)   /* For Visual 2005 */
+#    pragma intrinsic(_BitScanReverse)   /* For Visual 2005 */
+#  endif
+#  pragma warning(disable : 4127)        /* disable: C4127: conditional expression is constant */
+#else
+#  ifdef __GNUC__
+#    define FORCE_INLINE static inline __attribute__((always_inline))
+#  else
+#    define FORCE_INLINE static inline
+#  endif
+#endif
+
+#ifdef _MSC_VER  /* Visual Studio */
+#  define lz4_bswap16(x) _byteswap_ushort(x)
+#else
+#  define lz4_bswap16(x) ((unsigned short int) ((((x) >> 8) & 0xffu) | (((x) & 0xffu) << 8)))
+#endif
+
+#define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
+
+#if (GCC_VERSION >= 302) || (__INTEL_COMPILER >= 800) || defined(__clang__)
+#  define expect(expr,value)    (__builtin_expect ((expr),(value)) )
+#else
+#  define expect(expr,value)    (expr)
+#endif
+
+#define likely(expr)     expect((expr) != 0, 1)
+#define unlikely(expr)   expect((expr) != 0, 0)
+
+
+/**************************************
+   Memory routines
+**************************************/
+#include <stdlib.h>   /* malloc, calloc, free */
+#define ALLOCATOR(n,s) calloc(n,s)
+#define FREEMEM        free
+#include <string.h>   /* memset, memcpy */
+#define MEM_INIT       memset
+
+
+/**************************************
+   Includes
+**************************************/
+#include "lz4.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(LZ4_FORCE_UNALIGNED_ACCESS)
+#  define _PACKED __attribute__ ((packed))
+#else
+#  define _PACKED
+#endif
+
+#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
+#  if defined(__IBMC__) || defined(__SUNPRO_C) || defined(__SUNPRO_CC)
+#    pragma pack(1)
+#  else
+#    pragma pack(push, 1)
+#  endif
+#endif
+
+typedef struct { U16 v; }  _PACKED U16_S;
+typedef struct { U32 v; }  _PACKED U32_S;
+typedef struct { U64 v; }  _PACKED U64_S;
+typedef struct {size_t v;} _PACKED size_t_S;
+
+#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
+#  if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
+#    pragma pack(0)
+#  else
+#    pragma pack(pop)
+#  endif
+#endif
+
+#define A16(x)   (((U16_S *)(x))->v)
+#define A32(x)   (((U32_S *)(x))->v)
+#define A64(x)   (((U64_S *)(x))->v)
+#define AARCH(x) (((size_t_S *)(x))->v)
+
+
+/**************************************
+   Constants
+**************************************/
+#define LZ4_HASHLOG   (LZ4_MEMORY_USAGE-2)
+#define HASHTABLESIZE (1 << LZ4_MEMORY_USAGE)
+#define HASH_SIZE_U32 (1 << LZ4_HASHLOG)
+
+#define MINMATCH 4
+
+#define COPYLENGTH 8
+#define LASTLITERALS 5
+#define MFLIMIT (COPYLENGTH+MINMATCH)
+static const int LZ4_minLength = (MFLIMIT+1);
+
+#define KB *(1U<<10)
+#define MB *(1U<<20)
+#define GB *(1U<<30)
+
+#define LZ4_64KLIMIT ((64 KB) + (MFLIMIT-1))
+#define SKIPSTRENGTH 6   /* Increasing this value will make the compression run slower on incompressible data */
+
+#define MAXD_LOG 16
+#define MAX_DISTANCE ((1 << MAXD_LOG) - 1)
+
+#define ML_BITS  4
+#define ML_MASK  ((1U<<ML_BITS)-1)
+#define RUN_BITS (8-ML_BITS)
+#define RUN_MASK ((1U<<RUN_BITS)-1)
+
+
+/**************************************
+   Structures and local types
+**************************************/
+typedef struct {
+    U32  hashTable[HASH_SIZE_U32];
+    U32  currentOffset;
+    U32  initCheck;
+    const BYTE* dictionary;
+    const BYTE* bufferStart;
+    U32  dictSize;
+} LZ4_stream_t_internal;
+
+typedef enum { notLimited = 0, limitedOutput = 1 } limitedOutput_directive;
+typedef enum { byPtr, byU32, byU16 } tableType_t;
+
+typedef enum { noDict = 0, withPrefix64k, usingExtDict } dict_directive;
+typedef enum { noDictIssue = 0, dictSmall } dictIssue_directive;
+
+typedef enum { endOnOutputSize = 0, endOnInputSize = 1 } endCondition_directive;
+typedef enum { full = 0, partial = 1 } earlyEnd_directive;
+
+
+/**************************************
+   Architecture-specific macros
+**************************************/
+#define STEPSIZE                  sizeof(size_t)
+#define LZ4_COPYSTEP(d,s)         { AARCH(d) = AARCH(s); d+=STEPSIZE; s+=STEPSIZE; }
+#define LZ4_COPY8(d,s)            { LZ4_COPYSTEP(d,s); if (STEPSIZE<8) LZ4_COPYSTEP(d,s); }
+
+#if (defined(LZ4_BIG_ENDIAN) && !defined(BIG_ENDIAN_NATIVE_BUT_INCOMPATIBLE))
+#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { U16 v = A16(p); v = lz4_bswap16(v); d = (s) - v; }
+#  define LZ4_WRITE_LITTLEENDIAN_16(p,i)  { U16 v = (U16)(i); v = lz4_bswap16(v); A16(p) = v; p+=2; }
+#else      /* Little Endian */
+#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { d = (s) - A16(p); }
+#  define LZ4_WRITE_LITTLEENDIAN_16(p,v)  { A16(p) = v; p+=2; }
+#endif
+
+
+/**************************************
+   Macros
+**************************************/
+#define LZ4_STATIC_ASSERT(c)    { enum { LZ4_static_assert = 1/(!!(c)) }; }   /* use only *after* variable declarations */
+#if LZ4_ARCH64 || !defined(__GNUC__)
+#  define LZ4_WILDCOPY(d,s,e)   { do { LZ4_COPY8(d,s) } while (d<e); }        /* at the end, d>=e; */
+#else
+#  define LZ4_WILDCOPY(d,s,e)   { if (likely(e-d <= 8)) LZ4_COPY8(d,s) else do { LZ4_COPY8(d,s) } while (d<e); }
+#endif
+
+
+/****************************
+   Private local functions
+****************************/
+#if LZ4_ARCH64
+
+static int LZ4_NbCommonBytes (register U64 val)
+{
+# if defined(LZ4_BIG_ENDIAN)
+#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r = 0;
+    _BitScanReverse64( &r, val );
+    return (int)(r>>3);
+#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_clzll(val) >> 3);
+#   else
+    int r;
+    if (!(val>>32)) { r=4; } else { r=0; val>>=32; }
+    if (!(val>>16)) { r+=2; val>>=8; } else { val>>=24; }
+    r += (!val);
+    return r;
+#   endif
+# else
+#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r = 0;
+    _BitScanForward64( &r, val );
+    return (int)(r>>3);
+#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_ctzll(val) >> 3);
+#   else
+    static const int DeBruijnBytePos[64] = { 0, 0, 0, 0, 0, 1, 1, 2, 0, 3, 1, 3, 1, 4, 2, 7, 0, 2, 3, 6, 1, 5, 3, 5, 1, 3, 4, 4, 2, 5, 6, 7, 7, 0, 1, 2, 3, 3, 4, 6, 2, 6, 5, 5, 3, 4, 5, 6, 7, 1, 2, 4, 6, 4, 4, 5, 7, 2, 6, 5, 7, 6, 7, 7 };
+    return DeBruijnBytePos[((U64)((val & -(long long)val) * 0x0218A392CDABBD3FULL)) >> 58];
+#   endif
+# endif
+}
+
+#else
+
+static int LZ4_NbCommonBytes (register U32 val)
+{
+# if defined(LZ4_BIG_ENDIAN)
+#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r = 0;
+    _BitScanReverse( &r, val );
+    return (int)(r>>3);
+#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_clz(val) >> 3);
+#   else
+    int r;
+    if (!(val>>16)) { r=2; val>>=8; } else { r=0; val>>=24; }
+    r += (!val);
+    return r;
+#   endif
+# else
+#   if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r;
+    _BitScanForward( &r, val );
+    return (int)(r>>3);
+#   elif defined(__GNUC__) && (GCC_VERSION >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_ctz(val) >> 3);
+#   else
+    static const int DeBruijnBytePos[32] = { 0, 0, 3, 0, 3, 1, 3, 0, 3, 2, 2, 1, 3, 2, 0, 1, 3, 3, 1, 2, 2, 2, 2, 0, 3, 1, 2, 0, 1, 0, 1, 1 };
+    return DeBruijnBytePos[((U32)((val & -(S32)val) * 0x077CB531U)) >> 27];
+#   endif
+# endif
+}
+
+#endif
+
+
+/********************************
+   Compression functions
+********************************/
+int LZ4_versionNumber (void) { return LZ4_VERSION_NUMBER; }
+int LZ4_compressBound(int isize)  { return LZ4_COMPRESSBOUND(isize); }
+
+static int LZ4_hashSequence(U32 sequence, tableType_t tableType)
+{
+    if (tableType == byU16)
+        return (((sequence) * 2654435761U) >> ((MINMATCH*8)-(LZ4_HASHLOG+1)));
+    else
+        return (((sequence) * 2654435761U) >> ((MINMATCH*8)-LZ4_HASHLOG));
+}
+
+static int LZ4_hashPosition(const BYTE* p, tableType_t tableType) { return LZ4_hashSequence(A32(p), tableType); }
+
+static void LZ4_putPositionOnHash(const BYTE* p, U32 h, void* tableBase, tableType_t tableType, const BYTE* srcBase)
+{
+    switch (tableType)
+    {
+    case byPtr: { const BYTE** hashTable = (const BYTE**) tableBase; hashTable[h] = p; break; }
+    case byU32: { U32* hashTable = (U32*) tableBase; hashTable[h] = (U32)(p-srcBase); break; }
+    case byU16: { U16* hashTable = (U16*) tableBase; hashTable[h] = (U16)(p-srcBase); break; }
+    }
+}
+
+static void LZ4_putPosition(const BYTE* p, void* tableBase, tableType_t tableType, const BYTE* srcBase)
+{
+    U32 h = LZ4_hashPosition(p, tableType);
+    LZ4_putPositionOnHash(p, h, tableBase, tableType, srcBase);
+}
+
+static const BYTE* LZ4_getPositionOnHash(U32 h, void* tableBase, tableType_t tableType, const BYTE* srcBase)
+{
+    if (tableType == byPtr) { const BYTE** hashTable = (const BYTE**) tableBase; return hashTable[h]; }
+    if (tableType == byU32) { U32* hashTable = (U32*) tableBase; return hashTable[h] + srcBase; }
+    { U16* hashTable = (U16*) tableBase; return hashTable[h] + srcBase; }   /* default, to ensure a return */
+}
+
+static const BYTE* LZ4_getPosition(const BYTE* p, void* tableBase, tableType_t tableType, const BYTE* srcBase)
+{
+    U32 h = LZ4_hashPosition(p, tableType);
+    return LZ4_getPositionOnHash(h, tableBase, tableType, srcBase);
+}
+
+static unsigned LZ4_count(const BYTE* pIn, const BYTE* pRef, const BYTE* pInLimit)
+{
+    const BYTE* const pStart = pIn;
+
+    while (likely(pIn<pInLimit-(STEPSIZE-1)))
+    {
+        size_t diff = AARCH(pRef) ^ AARCH(pIn);
+        if (!diff) { pIn+=STEPSIZE; pRef+=STEPSIZE; continue; }
+        pIn += LZ4_NbCommonBytes(diff);
+        return (unsigned)(pIn - pStart);
+    }
+    if (LZ4_64BITS) if ((pIn<(pInLimit-3)) && (A32(pRef) == A32(pIn))) { pIn+=4; pRef+=4; }
+    if ((pIn<(pInLimit-1)) && (A16(pRef) == A16(pIn))) { pIn+=2; pRef+=2; }
+    if ((pIn<pInLimit) && (*pRef == *pIn)) pIn++;
+
+    return (unsigned)(pIn - pStart);
+}
+
+
+static int LZ4_compress_generic(
+                 void* ctx,
+                 const char* source,
+                 char* dest,
+                 int inputSize,
+                 int maxOutputSize,
+
+                 limitedOutput_directive outputLimited,
+                 tableType_t tableType,
+                 dict_directive dict,
+                 dictIssue_directive dictIssue)
+{
+    LZ4_stream_t_internal* const dictPtr = (LZ4_stream_t_internal*)ctx;
+
+    const BYTE* ip = (const BYTE*) source;
+    const BYTE* base;
+    const BYTE* lowLimit;
+    const BYTE* const lowRefLimit = ip - dictPtr->dictSize;
+    const BYTE* const dictionary = dictPtr->dictionary;
+    const BYTE* const dictEnd = dictionary + dictPtr->dictSize;
+    const size_t dictDelta = dictEnd - (const BYTE*)source;
+    const BYTE* anchor = (const BYTE*) source;
+    const BYTE* const iend = ip + inputSize;
+    const BYTE* const mflimit = iend - MFLIMIT;
+    const BYTE* const matchlimit = iend - LASTLITERALS;
+
+    BYTE* op = (BYTE*) dest;
+    BYTE* const olimit = op + maxOutputSize;
+
+    const int skipStrength = SKIPSTRENGTH;
+    U32 forwardH;
+    size_t refDelta=0;
+
+    /* Init conditions */
+    if ((U32)inputSize > (U32)LZ4_MAX_INPUT_SIZE) return 0;          /* Unsupported input size, too large (or negative) */
+    switch(dict)
+    {
+    case noDict:
+    default:
+        base = (const BYTE*)source;
+        lowLimit = (const BYTE*)source;
+        break;
+    case withPrefix64k:
+        base = (const BYTE*)source - dictPtr->currentOffset;
+        lowLimit = (const BYTE*)source - dictPtr->dictSize;
+        break;
+    case usingExtDict:
+        base = (const BYTE*)source - dictPtr->currentOffset;
+        lowLimit = (const BYTE*)source;
+        break;
+    }
+    if ((tableType == byU16) && (inputSize>=(int)LZ4_64KLIMIT)) return 0;   /* Size too large (not within 64K limit) */
+    if (inputSize<LZ4_minLength) goto _last_literals;                       /* Input too small, no compression (all literals) */
+
+    /* First Byte */
+    LZ4_putPosition(ip, ctx, tableType, base);
+    ip++; forwardH = LZ4_hashPosition(ip, tableType);
+
+    /* Main Loop */
+    for ( ; ; )
+    {
+        const BYTE* ref;
+        BYTE* token;
+        {
+            const BYTE* forwardIp = ip;
+            unsigned step=1;
+            unsigned searchMatchNb = (1U << skipStrength);
+
+            /* Find a match */
+            do {
+                U32 h = forwardH;
+                ip = forwardIp;
+                forwardIp += step;
+                step = searchMatchNb++ >> skipStrength;
+
+                if (unlikely(forwardIp > mflimit)) goto _last_literals;
+
+                ref = LZ4_getPositionOnHash(h, ctx, tableType, base);
+                if (dict==usingExtDict)
+                {
+                    if (ref<(const BYTE*)source)
+                    {
+                        refDelta = dictDelta;
+                        lowLimit = dictionary;
+                    }
+                    else
+                    {
+                        refDelta = 0;
+                        lowLimit = (const BYTE*)source;
+                    }
+                }
+                forwardH = LZ4_hashPosition(forwardIp, tableType);
+                LZ4_putPositionOnHash(ip, h, ctx, tableType, base);
+
+            } while ( ((dictIssue==dictSmall) ? (ref < lowRefLimit) : 0)
+                || ((tableType==byU16) ? 0 : (ref + MAX_DISTANCE < ip))
+                || (A32(ref+refDelta) != A32(ip)) );
+        }
+
+        /* Catch up */
+        while ((ip>anchor) && (ref+refDelta > lowLimit) && (unlikely(ip[-1]==ref[refDelta-1]))) { ip--; ref--; }
+
+        {
+            /* Encode Literal length */
+            unsigned litLength = (unsigned)(ip - anchor);
+            token = op++;
+            if ((outputLimited) && (unlikely(op + litLength + (2 + 1 + LASTLITERALS) + (litLength/255) > olimit)))
+                return 0;   /* Check output limit */
+            if (litLength>=RUN_MASK)
+            {
+                int len = (int)litLength-RUN_MASK;
+                *token=(RUN_MASK<<ML_BITS);
+                for(; len >= 255 ; len-=255) *op++ = 255;
+                *op++ = (BYTE)len;
+            }
+            else *token = (BYTE)(litLength<<ML_BITS);
+
+            /* Copy Literals */
+            { BYTE* end = op+litLength; LZ4_WILDCOPY(op,anchor,end); op=end; }
+        }
+
+_next_match:
+        /* Encode Offset */
+        LZ4_WRITE_LITTLEENDIAN_16(op, (U16)(ip-ref));
+
+        /* Encode MatchLength */
+        {
+            unsigned matchLength;
+
+            if ((dict==usingExtDict) && (lowLimit==dictionary))
+            {
+                const BYTE* limit;
+                ref += refDelta;
+                limit = ip + (dictEnd-ref);
+                if (limit > matchlimit) limit = matchlimit;
+                matchLength = LZ4_count(ip+MINMATCH, ref+MINMATCH, limit);
+                ip += MINMATCH + matchLength;
+                if (ip==limit)
+                {
+                    unsigned more = LZ4_count(ip, (const BYTE*)source, matchlimit);
+                    matchLength += more;
+                    ip += more;
+                }
+            }
+            else
+            {
+                matchLength = LZ4_count(ip+MINMATCH, ref+MINMATCH, matchlimit);
+                ip += MINMATCH + matchLength;
+            }
+
+            if (matchLength>=ML_MASK)
+            {
+                if ((outputLimited) && (unlikely(op + (1 + LASTLITERALS) + (matchLength>>8) > olimit)))
+                    return 0;    /* Check output limit */
+                *token += ML_MASK;
+                matchLength -= ML_MASK;
+                for (; matchLength >= 510 ; matchLength-=510) { *op++ = 255; *op++ = 255; }
+                if (matchLength >= 255) { matchLength-=255; *op++ = 255; }
+                *op++ = (BYTE)matchLength;
+            }
+            else *token += (BYTE)(matchLength);
+        }
+
+        anchor = ip;
+
+        /* Test end of chunk */
+        if (ip > mflimit) break;
+
+        /* Fill table */
+        LZ4_putPosition(ip-2, ctx, tableType, base);
+
+        /* Test next position */
+        ref = LZ4_getPosition(ip, ctx, tableType, base);
+        if (dict==usingExtDict)
+        {
+            if (ref<(const BYTE*)source)
+            {
+                refDelta = dictDelta;
+                lowLimit = dictionary;
+            }
+            else
+            {
+                refDelta = 0;
+                lowLimit = (const BYTE*)source;
+            }
+        }
+        LZ4_putPosition(ip, ctx, tableType, base);
+        if ( ((dictIssue==dictSmall) ? (ref>=lowRefLimit) : 1)
+            && (ref+MAX_DISTANCE>=ip)
+            && (A32(ref+refDelta)==A32(ip)) )
+        { token=op++; *token=0; goto _next_match; }
+
+        /* Prepare next loop */
+        forwardH = LZ4_hashPosition(++ip, tableType);
+    }
+
+_last_literals:
+    /* Encode Last Literals */
+    {
+        int lastRun = (int)(iend - anchor);
+        if ((outputLimited) && (((char*)op - dest) + lastRun + 1 + ((lastRun+255-RUN_MASK)/255) > (U32)maxOutputSize))
+            return 0;   /* Check output limit */
+        if (lastRun>=(int)RUN_MASK) { *op++=(RUN_MASK<<ML_BITS); lastRun-=RUN_MASK; for(; lastRun >= 255 ; lastRun-=255) *op++ = 255; *op++ = (BYTE) lastRun; }
+        else *op++ = (BYTE)(lastRun<<ML_BITS);
+        memcpy(op, anchor, iend - anchor);
+        op += iend-anchor;
+    }
+
+    /* End */
+    return (int) (((char*)op)-dest);
+}
+
+
+int LZ4_compress(const char* source, char* dest, int inputSize)
+{
+#if (HEAPMODE)
+    void* ctx = ALLOCATOR(LZ4_STREAMSIZE_U32, 4);   /* Aligned on 4-bytes boundaries */
+#else
+    U32 ctx[LZ4_STREAMSIZE_U32] = {0};      /* Ensure data is aligned on 4-bytes boundaries */
+#endif
+    int result;
+
+    if (inputSize < (int)LZ4_64KLIMIT)
+        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, 0, notLimited, byU16, noDict, noDictIssue);
+    else
+        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, 0, notLimited, LZ4_64BITS ? byU32 : byPtr, noDict, noDictIssue);
+
+#if (HEAPMODE)
+    FREEMEM(ctx);
+#endif
+    return result;
+}
+
+int LZ4_compress_limitedOutput(const char* source, char* dest, int inputSize, int maxOutputSize)
+{
+#if (HEAPMODE)
+    void* ctx = ALLOCATOR(LZ4_STREAMSIZE_U32, 4);   /* Aligned on 4-bytes boundaries */
+#else
+    U32 ctx[LZ4_STREAMSIZE_U32] = {0};      /* Ensure data is aligned on 4-bytes boundaries */
+#endif
+    int result;
+
+    if (inputSize < (int)LZ4_64KLIMIT)
+        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, maxOutputSize, limitedOutput, byU16, noDict, noDictIssue);
+    else
+        result = LZ4_compress_generic((void*)ctx, source, dest, inputSize, maxOutputSize, limitedOutput, LZ4_64BITS ? byU32 : byPtr, noDict, noDictIssue);
+
+#if (HEAPMODE)
+    FREEMEM(ctx);
+#endif
+    return result;
+}
+
+
+/*****************************************
+   Experimental : Streaming functions
+*****************************************/
+
+/*
+ * LZ4_initStream
+ * Use this function once, to init a newly allocated LZ4_stream_t structure
+ * Return : 1 if OK, 0 if error
+ */
+void LZ4_resetStream (LZ4_stream_t* LZ4_stream)
+{
+    MEM_INIT(LZ4_stream, 0, sizeof(LZ4_stream_t));
+}
+
+LZ4_stream_t* LZ4_createStream(void)
+{
+    LZ4_stream_t* lz4s = (LZ4_stream_t*)ALLOCATOR(4, LZ4_STREAMSIZE_U32);
+    LZ4_resetStream(lz4s);
+    return lz4s;
+}
+
+int LZ4_freeStream (LZ4_stream_t* LZ4_stream)
+{
+    FREEMEM(LZ4_stream);
+    return (0);
+}
+
+
+int LZ4_loadDict (LZ4_stream_t* LZ4_dict, const char* dictionary, int dictSize)
+{
+    LZ4_stream_t_internal* dict = (LZ4_stream_t_internal*) LZ4_dict;
+    const BYTE* p = (const BYTE*)dictionary;
+    const BYTE* const dictEnd = p + dictSize;
+    const BYTE* base;
+
+    LZ4_STATIC_ASSERT(LZ4_STREAMSIZE >= sizeof(LZ4_stream_t_internal));    /* A compilation error here means LZ4_STREAMSIZE is not large enough */
+    if (dict->initCheck) LZ4_resetStream(LZ4_dict);                         /* Uninitialized structure detected */
+
+    if (dictSize < MINMATCH)
+    {
+        dict->dictionary = NULL;
+        dict->dictSize = 0;
+        return 1;
+    }
+
+    if (p <= dictEnd - 64 KB) p = dictEnd - 64 KB;
+    base = p - dict->currentOffset;
+    dict->dictionary = p;
+    dict->dictSize = (U32)(dictEnd - p);
+    dict->currentOffset += dict->dictSize;
+
+    while (p <= dictEnd-MINMATCH)
+    {
+        LZ4_putPosition(p, dict, byU32, base);
+        p+=3;
+    }
+
+    return 1;
+}
+
+
+static void LZ4_renormDictT(LZ4_stream_t_internal* LZ4_dict, const BYTE* src)
+{
+    if ((LZ4_dict->currentOffset > 0x80000000) ||
+        ((size_t)LZ4_dict->currentOffset > (size_t)src))   /* address space overflow */
+    {
+        /* rescale hash table */
+        U32 delta = LZ4_dict->currentOffset - 64 KB;
+        const BYTE* dictEnd = LZ4_dict->dictionary + LZ4_dict->dictSize;
+        int i;
+        for (i=0; i<HASH_SIZE_U32; i++)
+        {
+            if (LZ4_dict->hashTable[i] < delta) LZ4_dict->hashTable[i]=0;
+            else LZ4_dict->hashTable[i] -= delta;
+        }
+        LZ4_dict->currentOffset = 64 KB;
+        if (LZ4_dict->dictSize > 64 KB) LZ4_dict->dictSize = 64 KB;
+        LZ4_dict->dictionary = dictEnd - LZ4_dict->dictSize;
+    }
+}
+
+
+FORCE_INLINE int LZ4_compress_continue_generic (void* LZ4_stream, const char* source, char* dest, int inputSize,
+                                                int maxOutputSize, limitedOutput_directive limit)
+{
+    LZ4_stream_t_internal* streamPtr = (LZ4_stream_t_internal*)LZ4_stream;
+    const BYTE* const dictEnd = streamPtr->dictionary + streamPtr->dictSize;
+
+    const BYTE* smallest = (const BYTE*) source;
+    if (streamPtr->initCheck) return 0;   /* Uninitialized structure detected */
+    if ((streamPtr->dictSize>0) && (smallest>dictEnd)) smallest = dictEnd;
+    LZ4_renormDictT(streamPtr, smallest);
+
+    /* Check overlapping input/dictionary space */
+    {
+        const BYTE* sourceEnd = (const BYTE*) source + inputSize;
+        if ((sourceEnd > streamPtr->dictionary) && (sourceEnd < dictEnd))
+        {
+            streamPtr->dictSize = (U32)(dictEnd - sourceEnd);
+            if (streamPtr->dictSize > 64 KB) streamPtr->dictSize = 64 KB;
+            if (streamPtr->dictSize < 4) streamPtr->dictSize = 0;
+            streamPtr->dictionary = dictEnd - streamPtr->dictSize;
+        }
+    }
+
+    /* prefix mode : source data follows dictionary */
+    if (dictEnd == (const BYTE*)source)
+    {
+        int result;
+        if ((streamPtr->dictSize < 64 KB) && (streamPtr->dictSize < streamPtr->currentOffset))
+            result = LZ4_compress_generic(LZ4_stream, source, dest, inputSize, maxOutputSize, limit, byU32, withPrefix64k, dictSmall);
+        else
+            result = LZ4_compress_generic(LZ4_stream, source, dest, inputSize, maxOutputSize, limit, byU32, withPrefix64k, noDictIssue);
+        streamPtr->dictSize += (U32)inputSize;
+        streamPtr->currentOffset += (U32)inputSize;
+        return result;
+    }
+
+    /* external dictionary mode */
+    {
+        int result;
+        if ((streamPtr->dictSize < 64 KB) && (streamPtr->dictSize < streamPtr->currentOffset))
+            result = LZ4_compress_generic(LZ4_stream, source, dest, inputSize, maxOutputSize, limit, byU32, usingExtDict, dictSmall);
+        else
+            result = LZ4_compress_generic(LZ4_stream, source, dest, inputSize, maxOutputSize, limit, byU32, usingExtDict, noDictIssue);
+        streamPtr->dictionary = (const BYTE*)source;
+        streamPtr->dictSize = (U32)inputSize;
+        streamPtr->currentOffset += (U32)inputSize;
+        return result;
+    }
+}
+
+int LZ4_compress_continue (LZ4_stream_t* LZ4_stream, const char* source, char* dest, int inputSize)
+{
+    return LZ4_compress_continue_generic(LZ4_stream, source, dest, inputSize, 0, notLimited);
+}
+
+int LZ4_compress_limitedOutput_continue (LZ4_stream_t* LZ4_stream, const char* source, char* dest, int inputSize, int maxOutputSize)
+{
+    return LZ4_compress_continue_generic(LZ4_stream, source, dest, inputSize, maxOutputSize, limitedOutput);
+}
+
+
+/* Hidden debug function, to force separate dictionary mode */
+int LZ4_compress_forceExtDict (LZ4_stream_t* LZ4_dict, const char* source, char* dest, int inputSize)
+{
+    LZ4_stream_t_internal* streamPtr = (LZ4_stream_t_internal*)LZ4_dict;
+    int result;
+    const BYTE* const dictEnd = streamPtr->dictionary + streamPtr->dictSize;
+
+    const BYTE* smallest = dictEnd;
+    if (smallest > (const BYTE*) source) smallest = (const BYTE*) source;
+    LZ4_renormDictT((LZ4_stream_t_internal*)LZ4_dict, smallest);
+
+    result = LZ4_compress_generic(LZ4_dict, source, dest, inputSize, 0, notLimited, byU32, usingExtDict, noDictIssue);
+
+    streamPtr->dictionary = (const BYTE*)source;
+    streamPtr->dictSize = (U32)inputSize;
+    streamPtr->currentOffset += (U32)inputSize;
+
+    return result;
+}
+
+
+int LZ4_saveDict (LZ4_stream_t* LZ4_dict, char* safeBuffer, int dictSize)
+{
+    LZ4_stream_t_internal* dict = (LZ4_stream_t_internal*) LZ4_dict;
+    const BYTE* previousDictEnd = dict->dictionary + dict->dictSize;
+
+    if ((U32)dictSize > 64 KB) dictSize = 64 KB;   /* useless to define a dictionary > 64 KB */
+    if ((U32)dictSize > dict->dictSize) dictSize = dict->dictSize;
+
+    memcpy(safeBuffer, previousDictEnd - dictSize, dictSize);
+
+    dict->dictionary = (const BYTE*)safeBuffer;
+    dict->dictSize = (U32)dictSize;
+
+    return 1;
+}
+
+
+
+/****************************
+   Decompression functions
+****************************/
+/*
+ * This generic decompression function cover all use cases.
+ * It shall be instanciated several times, using different sets of directives
+ * Note that it is essential this generic function is really inlined,
+ * in order to remove useless branches during compilation optimisation.
+ */
+FORCE_INLINE int LZ4_decompress_generic(
+                 const char* source,
+                 char* dest,
+                 int inputSize,
+                 int outputSize,         /* If endOnInput==endOnInputSize, this value is the max size of Output Buffer. */
+
+                 int endOnInput,         /* endOnOutputSize, endOnInputSize */
+                 int partialDecoding,    /* full, partial */
+                 int targetOutputSize,   /* only used if partialDecoding==partial */
+                 int dict,               /* noDict, withPrefix64k, usingExtDict */
+                 const char* dictStart,  /* only if dict==usingExtDict */
+                 int dictSize            /* note : = 0 if noDict */
+                 )
+{
+    /* Local Variables */
+    const BYTE* restrict ip = (const BYTE*) source;
+    const BYTE* ref;
+    const BYTE* const iend = ip + inputSize;
+
+    BYTE* op = (BYTE*) dest;
+    BYTE* const oend = op + outputSize;
+    BYTE* cpy;
+    BYTE* oexit = op + targetOutputSize;
+    const BYTE* const lowLimit = (const BYTE*)dest - dictSize;
+
+    const BYTE* const dictEnd = (const BYTE*)dictStart + dictSize;
+    const size_t dec32table[] = {4-0, 4-3, 4-2, 4-3, 4-0, 4-0, 4-0, 4-0};   /* note : static reduces speed for LZ4_decompress_safe() on GCC64 */
+    static const size_t dec64table[] = {0, 0, 0, (size_t)-1, 0, 1, 2, 3};
+
+    const int safeDecode = (endOnInput==endOnInputSize);
+    const int checkOffset = ((safeDecode) && (dictSize < (int)(64 KB)));
+
+
+    /* Special cases */
+    if ((partialDecoding) && (oexit> oend-MFLIMIT)) oexit = oend-MFLIMIT;                         /* targetOutputSize too high => decode everything */
+    if ((endOnInput) && (unlikely(outputSize==0))) return ((inputSize==1) && (*ip==0)) ? 0 : -1;  /* Empty output buffer */
+    if ((!endOnInput) && (unlikely(outputSize==0))) return (*ip==0?1:-1);
+
+
+    /* Main Loop */
+    while (1)
+    {
+        unsigned token;
+        size_t length;
+
+        /* get runlength */
+        token = *ip++;
+        if ((length=(token>>ML_BITS)) == RUN_MASK)
+        {
+            unsigned s;
+            do
+            {
+                s = *ip++;
+                length += s;
+            }
+            while (likely((endOnInput)?ip<iend-RUN_MASK:1) && (s==255));
+            if ((safeDecode) && LZ4_32BITS && unlikely((size_t)(op+length)<(size_t)(op))) goto _output_error;   /* overflow detection */
+            if ((safeDecode) && LZ4_32BITS && unlikely((size_t)(ip+length)<(size_t)(ip))) goto _output_error;   /* overflow detection */
+        }
+
+        /* copy literals */
+        cpy = op+length;
+        if (((endOnInput) && ((cpy>(partialDecoding?oexit:oend-MFLIMIT)) || (ip+length>iend-(2+1+LASTLITERALS))) )
+            || ((!endOnInput) && (cpy>oend-COPYLENGTH)))
+        {
+            if (partialDecoding)
+            {
+                if (cpy > oend) goto _output_error;                           /* Error : write attempt beyond end of output buffer */
+                if ((endOnInput) && (ip+length > iend)) goto _output_error;   /* Error : read attempt beyond end of input buffer */
+            }
+            else
+            {
+                if ((!endOnInput) && (cpy != oend)) goto _output_error;       /* Error : block decoding must stop exactly there */
+                if ((endOnInput) && ((ip+length != iend) || (cpy > oend))) goto _output_error;   /* Error : input must be consumed */
+            }
+            memcpy(op, ip, length);
+            ip += length;
+            op += length;
+            break;     /* Necessarily EOF, due to parsing restrictions */
+        }
+        LZ4_WILDCOPY(op, ip, cpy); ip -= (op-cpy); op = cpy;
+
+        /* get offset */
+        LZ4_READ_LITTLEENDIAN_16(ref,cpy,ip); ip+=2;
+        if ((checkOffset) && (unlikely(ref < lowLimit))) goto _output_error;   /* Error : offset outside destination buffer */
+
+        /* get matchlength */
+        if ((length=(token&ML_MASK)) == ML_MASK)
+        {
+            unsigned s;
+            do
+            {
+                if ((endOnInput) && (ip > iend-LASTLITERALS)) goto _output_error;
+                s = *ip++;
+                length += s;
+            } while (s==255);
+            if ((safeDecode) && LZ4_32BITS && unlikely((size_t)(op+length)<(size_t)op)) goto _output_error;   /* overflow detection */
+        }
+
+        /* check external dictionary */
+        if ((dict==usingExtDict) && (ref < (BYTE* const)dest))
+        {
+            if (unlikely(op+length+MINMATCH > oend-LASTLITERALS)) goto _output_error;
+
+            if (length+MINMATCH <= (size_t)(dest-(char*)ref))
+            {
+                ref = dictEnd - (dest-(char*)ref);
+                memcpy(op, ref, length+MINMATCH);
+                op += length+MINMATCH;
+            }
+            else
+            {
+                size_t copySize = (size_t)(dest-(char*)ref);
+                memcpy(op, dictEnd - copySize, copySize);
+                op += copySize;
+                copySize = length+MINMATCH - copySize;
+                if (copySize > (size_t)((char*)op-dest))   /* overlap */
+                {
+                    BYTE* const endOfMatch = op + copySize;
+                    const BYTE* copyFrom = (BYTE*)dest;
+                    while (op < endOfMatch) *op++ = *copyFrom++;
+                }
+                else
+                {
+                    memcpy(op, dest, copySize);
+                    op += copySize;
+                }
+            }
+            continue;
+        }
+
+        /* copy repeated sequence */
+        if (unlikely((op-ref)<(int)STEPSIZE))
+        {
+            const size_t dec64 = dec64table[LZ4_32BITS ? 0 : op-ref];
+            op[0] = ref[0];
+            op[1] = ref[1];
+            op[2] = ref[2];
+            op[3] = ref[3];
+            ref += dec32table[op-ref];
+            A32(op+4) = A32(ref);
+            op += STEPSIZE; ref -= dec64;
+        } else { LZ4_COPYSTEP(op,ref); }
+        cpy = op + length - (STEPSIZE-4);
+
+        if (unlikely(cpy>oend-COPYLENGTH-(STEPSIZE-4)))
+        {
+            if (cpy > oend-LASTLITERALS) goto _output_error;    /* Error : last 5 bytes must be literals */
+            if (op<oend-COPYLENGTH) LZ4_WILDCOPY(op, ref, (oend-COPYLENGTH));
+            while(op<cpy) *op++=*ref++;
+            op=cpy;
+            continue;
+        }
+        LZ4_WILDCOPY(op, ref, cpy);
+        op=cpy;   /* correction */
+    }
+
+    /* end of decoding */
+    if (endOnInput)
+       return (int) (((char*)op)-dest);     /* Nb of output bytes decoded */
+    else
+       return (int) (((char*)ip)-source);   /* Nb of input bytes read */
+
+    /* Overflow error detected */
+_output_error:
+    return (int) (-(((char*)ip)-source))-1;
+}
+
+
+int LZ4_decompress_safe(const char* source, char* dest, int compressedSize, int maxDecompressedSize)
+{
+    return LZ4_decompress_generic(source, dest, compressedSize, maxDecompressedSize, endOnInputSize, full, 0, noDict, NULL, 0);
+}
+
+int LZ4_decompress_safe_partial(const char* source, char* dest, int compressedSize, int targetOutputSize, int maxDecompressedSize)
+{
+    return LZ4_decompress_generic(source, dest, compressedSize, maxDecompressedSize, endOnInputSize, partial, targetOutputSize, noDict, NULL, 0);
+}
+
+int LZ4_decompress_fast(const char* source, char* dest, int originalSize)
+{
+    return LZ4_decompress_generic(source, dest, 0, originalSize, endOnOutputSize, full, 0, withPrefix64k, NULL, 64 KB);
+}
+
+/* streaming decompression functions */
+
+typedef struct
+{
+    const char* dictionary;
+    int dictSize;
+} LZ4_streamDecode_t_internal;
+
+/*
+ * If you prefer dynamic allocation methods,
+ * LZ4_createStreamDecode()
+ * provides a pointer (void*) towards an initialized LZ4_streamDecode_t structure.
+ */
+LZ4_streamDecode_t* LZ4_createStreamDecode(void)
+{
+    LZ4_streamDecode_t* lz4s = (LZ4_streamDecode_t*) ALLOCATOR(sizeof(U32), LZ4_STREAMDECODESIZE_U32);
+    MEM_INIT(lz4s, 0, LZ4_STREAMDECODESIZE);
+    return lz4s;
+}
+
+int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream)
+{
+    FREEMEM(LZ4_stream);
+    return 0;
+}
+
+/*
+ * LZ4_setStreamDecode
+ * Use this function to instruct where to find the dictionary
+ * This function is not necessary if previous data is still available where it was decoded.
+ * Loading a size of 0 is allowed (same effect as no dictionary).
+ * Return : 1 if OK, 0 if error
+ */
+int LZ4_setStreamDecode (LZ4_streamDecode_t* LZ4_streamDecode, const char* dictionary, int dictSize)
+{
+    LZ4_streamDecode_t_internal* lz4sd = (LZ4_streamDecode_t_internal*) LZ4_streamDecode;
+    lz4sd->dictionary = dictionary;
+    lz4sd->dictSize = dictSize;
+    return 1;
+}
+
+/*
+*_continue() :
+    These decoding functions allow decompression of multiple blocks in "streaming" mode.
+    Previously decoded blocks must still be available at the memory position where they were decoded.
+    If it's not possible, save the relevant part of decoded data into a safe buffer,
+    and indicate where it stands using LZ4_setDictDecode()
+*/
+int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int compressedSize, int maxOutputSize)
+{
+    LZ4_streamDecode_t_internal* lz4sd = (LZ4_streamDecode_t_internal*) LZ4_streamDecode;
+    int result;
+
+    result = LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, endOnInputSize, full, 0, usingExtDict, lz4sd->dictionary, lz4sd->dictSize);
+    if (result <= 0) return result;
+    if (lz4sd->dictionary + lz4sd->dictSize == dest)
+    {
+        lz4sd->dictSize += result;
+    }
+    else
+    {
+        lz4sd->dictionary = dest;
+        lz4sd->dictSize = result;
+    }
+
+    return result;
+}
+
+int LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int originalSize)
+{
+    LZ4_streamDecode_t_internal* lz4sd = (LZ4_streamDecode_t_internal*) LZ4_streamDecode;
+    int result;
+
+    result = LZ4_decompress_generic(source, dest, 0, originalSize, endOnOutputSize, full, 0, usingExtDict, lz4sd->dictionary, lz4sd->dictSize);
+    if (result <= 0) return result;
+    if (lz4sd->dictionary + lz4sd->dictSize == dest)
+    {
+        lz4sd->dictSize += result;
+    }
+    else
+    {
+        lz4sd->dictionary = dest;
+        lz4sd->dictSize = result;
+    }
+
+    return result;
+}
+
+
+/*
+Advanced decoding functions :
+*_usingDict() :
+    These decoding functions work the same as "_continue" ones,
+    the dictionary must be explicitly provided within parameters
+*/
+
+int LZ4_decompress_safe_usingDict(const char* source, char* dest, int compressedSize, int maxOutputSize, const char* dictStart, int dictSize)
+{
+    return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, endOnInputSize, full, 0, usingExtDict, dictStart, dictSize);
+}
+
+int LZ4_decompress_fast_usingDict(const char* source, char* dest, int originalSize, const char* dictStart, int dictSize)
+{
+    return LZ4_decompress_generic(source, dest, 0, originalSize, endOnOutputSize, full, 0, usingExtDict, dictStart, dictSize);
+}
+
+
+/***************************************************
+    Obsolete Functions
+***************************************************/
+/*
+These function names are deprecated and should no longer be used.
+They are only provided here for compatibility with older user programs.
+- LZ4_uncompress is totally equivalent to LZ4_decompress_fast
+- LZ4_uncompress_unknownOutputSize is totally equivalent to LZ4_decompress_safe
+*/
+int LZ4_uncompress (const char* source, char* dest, int outputSize) { return LZ4_decompress_fast(source, dest, outputSize); }
+int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize) { return LZ4_decompress_safe(source, dest, isize, maxOutputSize); }
+
+
+/* Obsolete Streaming functions */
+
+int LZ4_sizeofStreamState() { return LZ4_STREAMSIZE; }
+
+static void LZ4_init(LZ4_stream_t_internal* lz4ds, const BYTE* base)
+{
+    MEM_INIT(lz4ds, 0, LZ4_STREAMSIZE);
+    lz4ds->bufferStart = base;
+}
+
+int LZ4_resetStreamState(void* state, const char* inputBuffer)
+{
+    if ((((size_t)state) & 3) != 0) return 1;   /* Error : pointer is not aligned on 4-bytes boundary */
+    LZ4_init((LZ4_stream_t_internal*)state, (const BYTE*)inputBuffer);
+    return 0;
+}
+
+void* LZ4_create (const char* inputBuffer)
+{
+    void* lz4ds = ALLOCATOR(4, LZ4_STREAMSIZE_U32);
+    LZ4_init ((LZ4_stream_t_internal*)lz4ds, (const BYTE*)inputBuffer);
+    return lz4ds;
+}
+
+char* LZ4_slideInputBuffer (void* LZ4_Data)
+{
+    LZ4_stream_t_internal* lz4ds = (LZ4_stream_t_internal*)LZ4_Data;
+
+    LZ4_saveDict((LZ4_stream_t*)LZ4_Data, (char*)lz4ds->bufferStart, 64 KB);
+
+    return (char*)(lz4ds->bufferStart + 64 KB);
+}
+
+/*  Obsolete compresson functions using User-allocated state */
+
+int LZ4_sizeofState() { return LZ4_STREAMSIZE; }
+
+int LZ4_compress_withState (void* state, const char* source, char* dest, int inputSize)
+{
+    if (((size_t)(state)&3) != 0) return 0;   /* Error : state is not aligned on 4-bytes boundary */
+    MEM_INIT(state, 0, LZ4_STREAMSIZE);
+
+    if (inputSize < (int)LZ4_64KLIMIT)
+        return LZ4_compress_generic(state, source, dest, inputSize, 0, notLimited, byU16, noDict, noDictIssue);
+    else
+        return LZ4_compress_generic(state, source, dest, inputSize, 0, notLimited, LZ4_64BITS ? byU32 : byPtr, noDict, noDictIssue);
+}
+
+int LZ4_compress_limitedOutput_withState (void* state, const char* source, char* dest, int inputSize, int maxOutputSize)
+{
+    if (((size_t)(state)&3) != 0) return 0;   /* Error : state is not aligned on 4-bytes boundary */
+    MEM_INIT(state, 0, LZ4_STREAMSIZE);
+
+    if (inputSize < (int)LZ4_64KLIMIT)
+        return LZ4_compress_generic(state, source, dest, inputSize, maxOutputSize, limitedOutput, byU16, noDict, noDictIssue);
+    else
+        return LZ4_compress_generic(state, source, dest, inputSize, maxOutputSize, limitedOutput, LZ4_64BITS ? byU32 : byPtr, noDict, noDictIssue);
+}
+
+/* Obsolete streaming decompression functions */
+
+int LZ4_decompress_safe_withPrefix64k(const char* source, char* dest, int compressedSize, int maxOutputSize)
+{
+    return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, endOnInputSize, full, 0, withPrefix64k, NULL, 64 KB);
+}
+
+int LZ4_decompress_fast_withPrefix64k(const char* source, char* dest, int originalSize)
+{
+    return LZ4_decompress_generic(source, dest, 0, originalSize, endOnOutputSize, full, 0, withPrefix64k, NULL, 64 KB);
+}

ext/lz4/lz4.h

@@ -1,205 +1,326 @@
-/*
-   LZ4 - Fast LZ compression algorithm
-   Header File
-   Copyright (C) 2011-2013, 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 :
-   - LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
-   - LZ4 source repository : http://code.google.com/p/lz4/
-*/
-#pragma once
-
-#if defined (__cplusplus)
-extern "C" {
-#endif
-
-
-//**************************************
-// Compiler Options
-//**************************************
-#if defined(_MSC_VER) && !defined(__cplusplus)   // Visual Studio
-#  define inline __inline           // Visual C is not C99, but supports some kind of inline
-#endif
-
-
-//****************************
-// Simple Functions
-//****************************
-
-int LZ4_compress        (const char* source, char* dest, int inputSize);
-int LZ4_decompress_safe (const char* source, char* dest, int inputSize, int maxOutputSize);
-
-/*
-LZ4_compress() :
-    Compresses 'inputSize' bytes from 'source' into 'dest'.
-    Destination buffer must be already allocated,
-    and must be sized to handle worst cases situations (input data not compressible)
-    Worst case size evaluation is provided by function LZ4_compressBound()
-    inputSize : Max supported value is LZ4_MAX_INPUT_VALUE
-    return : the number of bytes written in buffer dest
-             or 0 if the compression fails
-
-LZ4_decompress_safe() :
-    maxOutputSize : is the size of the destination buffer (which must be already allocated)
-    return : the number of bytes decoded in the destination buffer (necessarily <= maxOutputSize)
-             If the source stream is detected malformed, the function will stop decoding and return a negative result.
-             This function is protected against buffer overflow exploits (never writes outside of output buffer, and never reads outside of input buffer). Therefore, it is protected against malicious data packets
-*/
-
-
-//****************************
-// Advanced Functions
-//****************************
-#define LZ4_MAX_INPUT_SIZE        0x7E000000   // 2 113 929 216 bytes
-#define LZ4_COMPRESSBOUND(isize)  ((unsigned int)(isize) > (unsigned int)LZ4_MAX_INPUT_SIZE ? 0 : (isize) + ((isize)/255) + 16)
-static inline int LZ4_compressBound(int isize)  { return LZ4_COMPRESSBOUND(isize); }
-
-/*
-LZ4_compressBound() :
-    Provides the maximum size that LZ4 may output in a "worst case" scenario (input data not compressible)
-    primarily useful for memory allocation of output buffer.
-    inline function is recommended for the general case,
-    macro is also provided when result needs to be evaluated at compilation (such as stack memory allocation).
-
-    isize  : is the input size. Max supported value is LZ4_MAX_INPUT_SIZE
-    return : maximum output size in a "worst case" scenario
-             or 0, if input size is too large ( > LZ4_MAX_INPUT_SIZE)
-*/
-
-
-int LZ4_compress_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
-
-/*
-LZ4_compress_limitedOutput() :
-    Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
-    If it cannot achieve it, compression will stop, and result of the function will be zero.
-    This function never writes outside of provided output buffer.
-
-    inputSize  : Max supported value is LZ4_MAX_INPUT_VALUE
-    maxOutputSize : is the size of the destination buffer (which must be already allocated)
-    return : the number of bytes written in buffer 'dest'
-             or 0 if the compression fails
-*/
-
-
-int LZ4_decompress_fast (const char* source, char* dest, int outputSize);
-
-/*
-LZ4_decompress_fast() :
-    outputSize : is the original (uncompressed) size
-    return : the number of bytes read from the source buffer (in other words, the compressed size)
-             If the source stream is malformed, the function will stop decoding and return a negative result.
-    note : This function is a bit faster than LZ4_decompress_safe()
-           This function never writes outside of output buffers, but may read beyond input buffer in case of malicious data packet.
-           Use this function preferably into a trusted environment (data to decode comes from a trusted source).
-           Destination buffer must be already allocated. Its size must be a minimum of 'outputSize' bytes.
-*/
-
-int LZ4_decompress_safe_partial (const char* source, char* dest, int inputSize, int targetOutputSize, int maxOutputSize);
-
-/*
-LZ4_decompress_safe_partial() :
-    This function decompress a compressed block of size 'inputSize' at position 'source'
-    into output buffer 'dest' of size 'maxOutputSize'.
-    The function tries to stop decompressing operation as soon as 'targetOutputSize' has been reached,
-    reducing decompression time.
-    return : the number of bytes decoded in the destination buffer (necessarily <= maxOutputSize)
-       Note : this number can be < 'targetOutputSize' should the compressed block to decode be smaller.
-             Always control how many bytes were decoded.
-             If the source stream is detected malformed, the function will stop decoding and return a negative result.
-             This function never writes outside of output buffer, and never reads outside of input buffer. It is therefore protected against malicious data packets
-*/
-
-
-//****************************
-// Stream Functions
-//****************************
-
-void* LZ4_create (const char* inputBuffer);
-int   LZ4_compress_continue (void* LZ4_Data, const char* source, char* dest, int inputSize);
-int   LZ4_compress_limitedOutput_continue (void* LZ4_Data, const char* source, char* dest, int inputSize, int maxOutputSize);
-char* LZ4_slideInputBuffer (void* LZ4_Data);
-int   LZ4_free (void* LZ4_Data);
-
-/* 
-These functions allow the compression of dependent blocks, where each block benefits from prior 64 KB within preceding blocks.
-In order to achieve this, it is necessary to start creating the LZ4 Data Structure, thanks to the function :
-
-void* LZ4_create (const char* inputBuffer);
-The result of the function is the (void*) pointer on the LZ4 Data Structure.
-This pointer will be needed in all other functions.
-If the pointer returned is NULL, then the allocation has failed, and compression must be aborted.
-The only parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
-The input buffer must be already allocated, and size at least 192KB.
-'inputBuffer' will also be the 'const char* source' of the first block.
-
-All blocks are expected to lay next to each other within the input buffer, starting from 'inputBuffer'.
-To compress each block, use either LZ4_compress_continue() or LZ4_compress_limitedOutput_continue().
-Their behavior are identical to LZ4_compress() or LZ4_compress_limitedOutput(), 
-but require the LZ4 Data Structure as their first argument, and check that each block starts right after the previous one.
-If next block does not begin immediately after the previous one, the compression will fail (return 0).
-
-When it's no longer possible to lay the next block after the previous one (not enough space left into input buffer), a call to : 
-char* LZ4_slideInputBuffer(void* LZ4_Data);
-must be performed. It will typically copy the latest 64KB of input at the beginning of input buffer.
-Note that, for this function to work properly, minimum size of an input buffer must be 192KB.
-==> The memory position where the next input data block must start is provided as the result of the function.
-
-Compression can then resume, using LZ4_compress_continue() or LZ4_compress_limitedOutput_continue(), as usual.
-
-When compression is completed, a call to LZ4_free() will release the memory used by the LZ4 Data Structure.
-*/
-
-
-int LZ4_decompress_safe_withPrefix64k (const char* source, char* dest, int inputSize, int maxOutputSize);
-int LZ4_decompress_fast_withPrefix64k (const char* source, char* dest, int outputSize);
-
-/*
-*_withPrefix64k() :
-    These decoding functions work the same as their "normal name" versions,
-    but can use up to 64KB of data in front of 'char* dest'.
-    These functions are necessary to decode inter-dependant blocks.
-*/
-
-
-//****************************
-// Obsolete Functions
-//****************************
-
-static inline int LZ4_uncompress (const char* source, char* dest, int outputSize) { return LZ4_decompress_fast(source, dest, outputSize); }
-static inline int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize) { return LZ4_decompress_safe(source, dest, isize, maxOutputSize); }
-
-/*
-These functions are deprecated and should no longer be used.
-They are provided here for compatibility with existing user programs.
-*/
-
-
-
-#if defined (__cplusplus)
-}
-#endif
+/*
+   LZ4 - Fast LZ compression algorithm
+   Header File
+   Copyright (C) 2011-2014, 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 :
+   - LZ4 source repository : http://code.google.com/p/lz4/
+   - LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
+*/
+#pragma once
+
+#if defined (__cplusplus)
+extern "C" {
+#endif
+
+
+/**************************************
+   Version
+**************************************/
+#define LZ4_VERSION_MAJOR    1    /* for major interface/format changes  */
+#define LZ4_VERSION_MINOR    3    /* for minor interface/format changes  */
+#define LZ4_VERSION_RELEASE  0    /* for tweaks, bug-fixes, or development */
+#define LZ4_VERSION_NUMBER (LZ4_VERSION_MAJOR *100*100 + LZ4_VERSION_MINOR *100 + LZ4_VERSION_RELEASE)
+int LZ4_versionNumber (void);
+
+/**************************************
+   Tuning parameter
+**************************************/
+/*
+ * LZ4_MEMORY_USAGE :
+ * Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.)
+ * Increasing memory usage improves compression ratio
+ * Reduced memory usage can improve speed, due to cache effect
+ * Default value is 14, for 16KB, which nicely fits into Intel x86 L1 cache
+ */
+#define LZ4_MEMORY_USAGE 14
+
+
+/**************************************
+   Simple Functions
+**************************************/
+
+int LZ4_compress        (const char* source, char* dest, int inputSize);
+int LZ4_decompress_safe (const char* source, char* dest, int compressedSize, int maxDecompressedSize);
+
+/*
+LZ4_compress() :
+    Compresses 'inputSize' bytes from 'source' into 'dest'.
+    Destination buffer must be already allocated,
+    and must be sized to handle worst cases situations (input data not compressible)
+    Worst case size evaluation is provided by function LZ4_compressBound()
+    inputSize : Max supported value is LZ4_MAX_INPUT_SIZE
+    return : the number of bytes written in buffer dest
+             or 0 if the compression fails
+
+LZ4_decompress_safe() :
+    compressedSize : is obviously the source size
+    maxDecompressedSize : is the size of the destination buffer, which must be already allocated.
+    return : the number of bytes decompressed into the destination buffer (necessarily <= maxDecompressedSize)
+             If the destination buffer is not large enough, decoding will stop and output an error code (<0).
+             If the source stream is detected malformed, the function will stop decoding and return a negative result.
+             This function is protected against buffer overflow exploits :
+             it never writes outside of output buffer, and never reads outside of input buffer.
+             Therefore, it is protected against malicious data packets.
+*/
+
+
+/*
+Note :
+    Should you prefer to explicitly allocate compression-table memory using your own allocation method,
+    use the streaming functions provided below, simply reset the memory area between each call to LZ4_compress_continue()
+*/
+
+
+/**************************************
+   Advanced Functions
+**************************************/
+#define LZ4_MAX_INPUT_SIZE        0x7E000000   /* 2 113 929 216 bytes */
+#define LZ4_COMPRESSBOUND(isize)  ((unsigned int)(isize) > (unsigned int)LZ4_MAX_INPUT_SIZE ? 0 : (isize) + ((isize)/255) + 16)
+
+/*
+LZ4_compressBound() :
+    Provides the maximum size that LZ4 may output in a "worst case" scenario (input data not compressible)
+    primarily useful for memory allocation of output buffer.
+    macro is also provided when result needs to be evaluated at compilation (such as stack memory allocation).
+
+    isize  : is the input size. Max supported value is LZ4_MAX_INPUT_SIZE
+    return : maximum output size in a "worst case" scenario
+             or 0, if input size is too large ( > LZ4_MAX_INPUT_SIZE)
+*/
+int LZ4_compressBound(int isize);
+
+
+/*
+LZ4_compress_limitedOutput() :
+    Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
+    If it cannot achieve it, compression will stop, and result of the function will be zero.
+    This function never writes outside of provided output buffer.
+
+    inputSize  : Max supported value is LZ4_MAX_INPUT_VALUE
+    maxOutputSize : is the size of the destination buffer (which must be already allocated)
+    return : the number of bytes written in buffer 'dest'
+             or 0 if the compression fails
+*/
+int LZ4_compress_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
+
+
+/*
+LZ4_compress_withState() :
+    Same compression functions, but using an externally allocated memory space to store compression state.
+    Use LZ4_sizeofState() to know how much memory must be allocated,
+    and then, provide it as 'void* state' to compression functions.
+*/
+int LZ4_sizeofState(void);
+int LZ4_compress_withState               (void* state, const char* source, char* dest, int inputSize);
+int LZ4_compress_limitedOutput_withState (void* state, const char* source, char* dest, int inputSize, int maxOutputSize);
+
+
+/*
+LZ4_decompress_fast() :
+    originalSize : is the original and therefore uncompressed size
+    return : the number of bytes read from the source buffer (in other words, the compressed size)
+             If the source stream is detected malformed, the function will stop decoding and return a negative result.
+             Destination buffer must be already allocated. Its size must be a minimum of 'originalSize' bytes.
+    note : This function fully respect memory boundaries for properly formed compressed data.
+           It is a bit faster than LZ4_decompress_safe().
+           However, it does not provide any protection against intentionnally modified data stream (malicious input).
+           Use this function in trusted environment only (data to decode comes from a trusted source).
+*/
+int LZ4_decompress_fast (const char* source, char* dest, int originalSize);
+
+
+/*
+LZ4_decompress_safe_partial() :
+    This function decompress a compressed block of size 'compressedSize' at position 'source'
+    into destination buffer 'dest' of size 'maxDecompressedSize'.
+    The function tries to stop decompressing operation as soon as 'targetOutputSize' has been reached,
+    reducing decompression time.
+    return : the number of bytes decoded in the destination buffer (necessarily <= maxDecompressedSize)
+       Note : this number can be < 'targetOutputSize' should the compressed block to decode be smaller.
+             Always control how many bytes were decoded.
+             If the source stream is detected malformed, the function will stop decoding and return a negative result.
+             This function never writes outside of output buffer, and never reads outside of input buffer. It is therefore protected against malicious data packets
+*/
+int LZ4_decompress_safe_partial (const char* source, char* dest, int compressedSize, int targetOutputSize, int maxDecompressedSize);
+
+
+/***********************************************
+   Experimental Streaming Compression Functions
+***********************************************/
+
+#define LZ4_STREAMSIZE_U32 ((1 << (LZ4_MEMORY_USAGE-2)) + 8)
+#define LZ4_STREAMSIZE     (LZ4_STREAMSIZE_U32 * sizeof(unsigned int))
+/*
+ * LZ4_stream_t
+ * information structure to track an LZ4 stream.
+ * important : init this structure content before first use !
+ */
+typedef struct { unsigned int table[LZ4_STREAMSIZE_U32]; } LZ4_stream_t;
+
+/*
+ * LZ4_resetStream
+ * Use this function to init a newly allocated LZ4_stream_t structure
+ * You can also reset an existing LZ4_stream_t structure
+ */
+void LZ4_resetStream (LZ4_stream_t* LZ4_stream);
+
+/*
+ * If you prefer dynamic allocation methods,
+ * LZ4_createStream will allocate and initialize an LZ4_stream_t structure
+ * LZ4_freeStream releases its memory.
+ */
+LZ4_stream_t* LZ4_createStream(void);
+int           LZ4_freeStream (LZ4_stream_t* LZ4_stream);
+
+/*
+ * LZ4_loadDict
+ * Use this function to load a static dictionary into LZ4_stream.
+ * Any previous data will be forgotten, only 'dictionary' will remain in memory.
+ * Loading a size of 0 is allowed.
+ * Return : 1 if OK, 0 if error
+ */
+int LZ4_loadDict (LZ4_stream_t* LZ4_stream, const char* dictionary, int dictSize);
+
+/*
+ * LZ4_compress_continue
+ * Compress data block 'source', using blocks compressed before as dictionary to improve compression ratio
+ * Previous data blocks are assumed to still be present at their previous location.
+ */
+int LZ4_compress_continue (LZ4_stream_t* LZ4_stream, const char* source, char* dest, int inputSize);
+
+/*
+ * LZ4_compress_limitedOutput_continue
+ * Same as before, but also specify a maximum target compressed size (maxOutputSize)
+ * If objective cannot be met, compression exits, and returns a zero.
+ */
+int LZ4_compress_limitedOutput_continue (LZ4_stream_t* LZ4_stream, const char* source, char* dest, int inputSize, int maxOutputSize);
+
+/*
+ * LZ4_saveDict
+ * If previously compressed data block is not guaranteed to remain available at its memory location
+ * save it into a safe place (char* safeBuffer)
+ * Note : you don't need to call LZ4_loadDict() afterwards,
+ *        dictionary is immediately usable, you can therefore call again LZ4_compress_continue()
+ * Return : 1 if OK, 0 if error
+ * Note : any dictSize > 64 KB will be interpreted as 64KB.
+ */
+int LZ4_saveDict (LZ4_stream_t* LZ4_stream, char* safeBuffer, int dictSize);
+
+
+/************************************************
+  Experimental Streaming Decompression Functions
+************************************************/
+
+#define LZ4_STREAMDECODESIZE_U32 4
+#define LZ4_STREAMDECODESIZE     (LZ4_STREAMDECODESIZE_U32 * sizeof(unsigned int))
+/*
+ * LZ4_streamDecode_t
+ * information structure to track an LZ4 stream.
+ * important : init this structure content using LZ4_setStreamDecode or memset() before first use !
+ */
+typedef struct { unsigned int table[LZ4_STREAMDECODESIZE_U32]; } LZ4_streamDecode_t;
+
+/*
+ * LZ4_setStreamDecode
+ * Use this function to instruct where to find the dictionary.
+ * This function can be used to specify a static dictionary,
+ * or to instruct where to find some previously decoded data saved into a different memory space.
+ * Setting a size of 0 is allowed (same effect as no dictionary).
+ * Return : 1 if OK, 0 if error
+ */
+int LZ4_setStreamDecode (LZ4_streamDecode_t* LZ4_streamDecode, const char* dictionary, int dictSize);
+
+/*
+ * If you prefer dynamic allocation methods,
+ * LZ4_createStreamDecode will allocate and initialize an LZ4_streamDecode_t structure
+ * LZ4_freeStreamDecode releases its memory.
+ */
+LZ4_streamDecode_t* LZ4_createStreamDecode(void);
+int                 LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream);
+
+/*
+*_continue() :
+    These decoding functions allow decompression of multiple blocks in "streaming" mode.
+    Previously decoded blocks must still be available at the memory position where they were decoded.
+    If it's not possible, save the relevant part of decoded data into a safe buffer,
+    and indicate where its new address using LZ4_setDictDecode()
+*/
+int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int compressedSize, int maxOutputSize);
+int LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int originalSize);
+
+
+/*
+Advanced decoding functions :
+*_usingDict() :
+    These decoding functions work the same as
+    a combination of LZ4_setDictDecode() followed by LZ4_decompress_x_continue()
+    all together into a single function call.
+    It doesn't use nor update an LZ4_streamDecode_t structure.
+*/
+int LZ4_decompress_safe_usingDict (const char* source, char* dest, int compressedSize, int maxOutputSize, const char* dictStart, int dictSize);
+int LZ4_decompress_fast_usingDict (const char* source, char* dest, int originalSize, const char* dictStart, int dictSize);
+
+
+
+/**************************************
+   Obsolete Functions
+**************************************/
+/*
+Obsolete decompression functions
+These function names are deprecated and should no longer be used.
+They are only provided here for compatibility with older user programs.
+- LZ4_uncompress is the same as LZ4_decompress_fast
+- LZ4_uncompress_unknownOutputSize is the same as LZ4_decompress_safe
+These function prototypes are now disabled; uncomment them if you really need them.
+It is highly recommended to stop using these functions and migrated to newer ones */
+/* int LZ4_uncompress (const char* source, char* dest, int outputSize); */
+/* int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize); */
+
+/*
+ * If you prefer dynamic allocation methods,
+ * LZ4_createStreamDecode()
+ * provides a pointer (void*) towards an initialized LZ4_streamDecode_t structure.
+ * LZ4_free just frees it.
+ */
+/* void* LZ4_createStreamDecode(void); */
+/*int   LZ4_free (void* LZ4_stream);    yes, it's the same one as for compression */
+
+/* Obsolete streaming functions; use new streaming interface whenever possible */
+void* LZ4_create (const char* inputBuffer);
+int   LZ4_sizeofStreamState(void);
+int   LZ4_resetStreamState(void* state, const char* inputBuffer);
+char* LZ4_slideInputBuffer (void* state);
+
+/* Obsolete streaming decoding functions */
+int LZ4_decompress_safe_withPrefix64k (const char* source, char* dest, int compressedSize, int maxOutputSize);
+int LZ4_decompress_fast_withPrefix64k (const char* source, char* dest, int originalSize);
+
+
+#if defined (__cplusplus)
+}
+#endif

ext/lz4/lz4hc.c

@@ -1,817 +1,892 @@
-/*
-   LZ4 HC - High Compression Mode of LZ4
-   Copyright (C) 2011-2013, 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 :
-   - LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
-   - LZ4 source repository : http://code.google.com/p/lz4/
-*/
-
-//**************************************
-// Memory routines
-//**************************************
-#include <stdlib.h>   // calloc, free
-#define ALLOCATOR(s)  calloc(1,s)
-#define FREEMEM       free
-#include <string.h>   // memset, memcpy
-#define MEM_INIT      memset
-
-
-//**************************************
-// CPU Feature Detection
-//**************************************
-// 32 or 64 bits ?
-#if (defined(__x86_64__) || defined(_M_X64) || defined(_WIN64) \
-  || defined(__powerpc64__) || defined(__ppc64__) || defined(__PPC64__) \
-  || defined(__64BIT__) || defined(_LP64) || defined(__LP64__) \
-  || defined(__ia64) || defined(__itanium__) || defined(_M_IA64) )   // Detects 64 bits mode
-#  define LZ4_ARCH64 1
-#else
-#  define LZ4_ARCH64 0
-#endif
-
-// Little Endian or Big Endian ?
-// Overwrite the #define below if you know your architecture endianess
-#if defined (__GLIBC__)
-#  include <endian.h>
-#  if (__BYTE_ORDER == __BIG_ENDIAN)
-#     define LZ4_BIG_ENDIAN 1
-#  endif
-#elif (defined(__BIG_ENDIAN__) || defined(__BIG_ENDIAN) || defined(_BIG_ENDIAN)) && !(defined(__LITTLE_ENDIAN__) || defined(__LITTLE_ENDIAN) || defined(_LITTLE_ENDIAN))
-#  define LZ4_BIG_ENDIAN 1
-#elif defined(__sparc) || defined(__sparc__) \
-   || defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) \
-   || defined(__hpux)  || defined(__hppa) \
-   || defined(_MIPSEB) || defined(__s390__)
-#  define LZ4_BIG_ENDIAN 1
-#else
-// Little Endian assumed. PDP Endian and other very rare endian format are unsupported.
-#endif
-
-// 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
-#if defined(__ARM_FEATURE_UNALIGNED)
-#  define LZ4_FORCE_UNALIGNED_ACCESS 1
-#endif
-
-// Define this parameter if your target system or compiler does not support hardware bit count
-#if defined(_MSC_VER) && defined(_WIN32_WCE)            // Visual Studio for Windows CE does not support Hardware bit count
-#  define LZ4_FORCE_SW_BITCOUNT
-#endif
-
-
-//**************************************
-// Compiler Options
-//**************************************
-#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L   // C99
-  /* "restrict" is a known keyword */
-#else
-#  define restrict  // Disable restrict
-#endif
-
-#ifdef _MSC_VER    // Visual Studio
-#  define FORCE_INLINE static __forceinline
-#  include <intrin.h>                    // For Visual 2005
-#  if LZ4_ARCH64   // 64-bits
-#    pragma intrinsic(_BitScanForward64) // For Visual 2005
-#    pragma intrinsic(_BitScanReverse64) // For Visual 2005
-#  else            // 32-bits
-#    pragma intrinsic(_BitScanForward)   // For Visual 2005
-#    pragma intrinsic(_BitScanReverse)   // For Visual 2005
-#  endif
-#  pragma warning(disable : 4127)        // disable: C4127: conditional expression is constant
-#  pragma warning(disable : 4701)        // disable: C4701: potentially uninitialized local variable used
-#else 
-#  ifdef __GNUC__
-#    define FORCE_INLINE static inline __attribute__((always_inline))
-#  else
-#    define FORCE_INLINE static inline
-#  endif
-#endif
-
-#ifdef _MSC_VER  // Visual Studio
-#  define lz4_bswap16(x) _byteswap_ushort(x)
-#else
-#  define lz4_bswap16(x)  ((unsigned short int) ((((x) >> 8) & 0xffu) | (((x) & 0xffu) << 8)))
-#endif
-
-
-//**************************************
-// Includes
-//**************************************
-#include "lz4hc.h"
-#include "lz4.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(LZ4_FORCE_UNALIGNED_ACCESS)
-#  define _PACKED __attribute__ ((packed))
-#else
-#  define _PACKED
-#endif
-
-#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
-#  ifdef __IBMC__
-#    pragma pack(1)
-#  else
-#    pragma pack(push, 1)
-#  endif
-#endif
-
-typedef struct _U16_S { U16 v; } _PACKED U16_S;
-typedef struct _U32_S { U32 v; } _PACKED U32_S;
-typedef struct _U64_S { U64 v; } _PACKED U64_S;
-
-#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
-#  pragma pack(pop)
-#endif
-
-#define A64(x) (((U64_S *)(x))->v)
-#define A32(x) (((U32_S *)(x))->v)
-#define A16(x) (((U16_S *)(x))->v)
-
-
-//**************************************
-// Constants
-//**************************************
-#define MINMATCH 4
-
-#define DICTIONARY_LOGSIZE 16
-#define MAXD (1<<DICTIONARY_LOGSIZE)
-#define MAXD_MASK ((U32)(MAXD - 1))
-#define MAX_DISTANCE (MAXD - 1)
-
-#define HASH_LOG (DICTIONARY_LOGSIZE-1)
-#define HASHTABLESIZE (1 << HASH_LOG)
-#define HASH_MASK (HASHTABLESIZE - 1)
-
-#define MAX_NB_ATTEMPTS 256
-
-#define ML_BITS  4
-#define ML_MASK  (size_t)((1U<<ML_BITS)-1)
-#define RUN_BITS (8-ML_BITS)
-#define RUN_MASK ((1U<<RUN_BITS)-1)
-
-#define COPYLENGTH 8
-#define LASTLITERALS 5
-#define MFLIMIT (COPYLENGTH+MINMATCH)
-#define MINLENGTH (MFLIMIT+1)
-#define OPTIMAL_ML (int)((ML_MASK-1)+MINMATCH)
-
-#define KB *(1U<<10)
-#define MB *(1U<<20)
-#define GB *(1U<<30)
-
-
-//**************************************
-// Architecture-specific macros
-//**************************************
-#if LZ4_ARCH64   // 64-bit
-#  define STEPSIZE 8
-#  define LZ4_COPYSTEP(s,d)     A64(d) = A64(s); d+=8; s+=8;
-#  define LZ4_COPYPACKET(s,d)   LZ4_COPYSTEP(s,d)
-#  define UARCH U64
-#  define AARCH A64
-#  define HTYPE                 U32
-#  define INITBASE(b,s)         const BYTE* const b = s
-#else   // 32-bit
-#  define STEPSIZE 4
-#  define LZ4_COPYSTEP(s,d)     A32(d) = A32(s); d+=4; s+=4;
-#  define LZ4_COPYPACKET(s,d)   LZ4_COPYSTEP(s,d); LZ4_COPYSTEP(s,d);
-#  define UARCH U32
-#  define AARCH A32
-//#  define HTYPE                 const BYTE*
-//#  define INITBASE(b,s)         const int b = 0
-#  define HTYPE                 U32
-#  define INITBASE(b,s)         const BYTE* const b = s
-#endif
-
-#if defined(LZ4_BIG_ENDIAN)
-#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { U16 v = A16(p); v = lz4_bswap16(v); d = (s) - v; }
-#  define LZ4_WRITE_LITTLEENDIAN_16(p,i)  { U16 v = (U16)(i); v = lz4_bswap16(v); A16(p) = v; p+=2; }
-#else   // Little Endian
-#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { d = (s) - A16(p); }
-#  define LZ4_WRITE_LITTLEENDIAN_16(p,v)  { A16(p) = v; p+=2; }
-#endif
-
-
-//************************************************************
-// Local Types
-//************************************************************
-typedef struct 
-{
-    const BYTE* inputBuffer;
-    const BYTE* base;
-    const BYTE* end;
-    HTYPE hashTable[HASHTABLESIZE];
-    U16 chainTable[MAXD];
-    const BYTE* nextToUpdate;
-} LZ4HC_Data_Structure;
-
-
-//**************************************
-// Macros
-//**************************************
-#define LZ4_WILDCOPY(s,d,e)    do { LZ4_COPYPACKET(s,d) } while (d<e);
-#define LZ4_BLINDCOPY(s,d,l)   { BYTE* e=d+l; LZ4_WILDCOPY(s,d,e); d=e; }
-#define HASH_FUNCTION(i)       (((i) * 2654435761U) >> ((MINMATCH*8)-HASH_LOG))
-#define HASH_VALUE(p)          HASH_FUNCTION(A32(p))
-#define HASH_POINTER(p)        (HashTable[HASH_VALUE(p)] + base)
-#define DELTANEXT(p)           chainTable[(size_t)(p) & MAXD_MASK] 
-#define GETNEXT(p)             ((p) - (size_t)DELTANEXT(p))
-
-
-//**************************************
-// Private functions
-//**************************************
-#if LZ4_ARCH64
-
-FORCE_INLINE int LZ4_NbCommonBytes (register U64 val)
-{
-#if defined(LZ4_BIG_ENDIAN)
-#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r = 0;
-    _BitScanReverse64( &r, val );
-    return (int)(r>>3);
-#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_clzll(val) >> 3); 
-#  else
-    int r;
-    if (!(val>>32)) { r=4; } else { r=0; val>>=32; }
-    if (!(val>>16)) { r+=2; val>>=8; } else { val>>=24; }
-    r += (!val);
-    return r;
-#  endif
-#else
-#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r = 0;
-    _BitScanForward64( &r, val );
-    return (int)(r>>3);
-#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_ctzll(val) >> 3); 
-#  else
-    static const int DeBruijnBytePos[64] = { 0, 0, 0, 0, 0, 1, 1, 2, 0, 3, 1, 3, 1, 4, 2, 7, 0, 2, 3, 6, 1, 5, 3, 5, 1, 3, 4, 4, 2, 5, 6, 7, 7, 0, 1, 2, 3, 3, 4, 6, 2, 6, 5, 5, 3, 4, 5, 6, 7, 1, 2, 4, 6, 4, 4, 5, 7, 2, 6, 5, 7, 6, 7, 7 };
-    return DeBruijnBytePos[((U64)((val & -val) * 0x0218A392CDABBD3F)) >> 58];
-#  endif
-#endif
-}
-
-#else
-
-FORCE_INLINE int LZ4_NbCommonBytes (register U32 val)
-{
-#if defined(LZ4_BIG_ENDIAN)
-#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r;
-    _BitScanReverse( &r, val );
-    return (int)(r>>3);
-#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_clz(val) >> 3); 
-#  else
-    int r;
-    if (!(val>>16)) { r=2; val>>=8; } else { r=0; val>>=24; }
-    r += (!val);
-    return r;
-#  endif
-#else
-#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    unsigned long r;
-    _BitScanForward( &r, val );
-    return (int)(r>>3);
-#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
-    return (__builtin_ctz(val) >> 3); 
-#  else
-    static const int DeBruijnBytePos[32] = { 0, 0, 3, 0, 3, 1, 3, 0, 3, 2, 2, 1, 3, 2, 0, 1, 3, 3, 1, 2, 2, 2, 2, 0, 3, 1, 2, 0, 1, 0, 1, 1 };
-    return DeBruijnBytePos[((U32)((val & -(S32)val) * 0x077CB531U)) >> 27];
-#  endif
-#endif
-}
-
-#endif
-
-
-FORCE_INLINE void LZ4_initHC (LZ4HC_Data_Structure* hc4, const BYTE* base)
-{
-    MEM_INIT((void*)hc4->hashTable, 0, sizeof(hc4->hashTable));
-    MEM_INIT(hc4->chainTable, 0xFF, sizeof(hc4->chainTable));
-    hc4->nextToUpdate = base + 1;
-    hc4->base = base;
-    hc4->inputBuffer = base;
-    hc4->end = base;
-}
-
-
-void* LZ4_createHC (const char* inputBuffer)
-{
-    void* hc4 = ALLOCATOR(sizeof(LZ4HC_Data_Structure));
-    LZ4_initHC ((LZ4HC_Data_Structure*)hc4, (const BYTE*)inputBuffer);
-    return hc4;
-}
-
-
-int LZ4_freeHC (void* LZ4HC_Data)
-{
-    FREEMEM(LZ4HC_Data);
-    return (0);
-}
-
-
-// Update chains up to ip (excluded)
-FORCE_INLINE void LZ4HC_Insert (LZ4HC_Data_Structure* hc4, const BYTE* ip)
-{
-    U16*   chainTable = hc4->chainTable;
-    HTYPE* HashTable  = hc4->hashTable;
-    INITBASE(base,hc4->base);
-
-    while(hc4->nextToUpdate < ip)
-    {
-        const BYTE* const p = hc4->nextToUpdate;
-        size_t delta = (p) - HASH_POINTER(p); 
-        if (delta>MAX_DISTANCE) delta = MAX_DISTANCE; 
-        DELTANEXT(p) = (U16)delta; 
-        HashTable[HASH_VALUE(p)] = (HTYPE)((p) - base);
-        hc4->nextToUpdate++;
-    }
-}
-
-
-char* LZ4_slideInputBufferHC(void* LZ4HC_Data)
-{
-    LZ4HC_Data_Structure* hc4 = (LZ4HC_Data_Structure*)LZ4HC_Data;
-    U32 distance = (U32)(hc4->end - hc4->inputBuffer) - 64 KB;
-    distance = (distance >> 16) << 16;   // Must be a multiple of 64 KB
-    LZ4HC_Insert(hc4, hc4->end - MINMATCH);
-    memcpy((void*)(hc4->end - 64 KB - distance), (const void*)(hc4->end - 64 KB), 64 KB);
-    hc4->nextToUpdate -= distance;
-    hc4->base -= distance;
-    if ((U32)(hc4->inputBuffer - hc4->base) > 1 GB + 64 KB)   // Avoid overflow
-    {
-        int i;
-        hc4->base += 1 GB;
-        for (i=0; i<HASHTABLESIZE; i++) hc4->hashTable[i] -= 1 GB;
-    }
-    hc4->end -= distance;
-    return (char*)(hc4->end);
-}
-
-
-FORCE_INLINE size_t LZ4HC_CommonLength (const BYTE* p1, const BYTE* p2, const BYTE* const matchlimit)
-{
-    const BYTE* p1t = p1;
-
-    while (p1t<matchlimit-(STEPSIZE-1))
-    {
-        UARCH diff = AARCH(p2) ^ AARCH(p1t);
-        if (!diff) { p1t+=STEPSIZE; p2+=STEPSIZE; continue; }
-        p1t += LZ4_NbCommonBytes(diff);
-        return (p1t - p1);
-    }
-    if (LZ4_ARCH64) if ((p1t<(matchlimit-3)) && (A32(p2) == A32(p1t))) { p1t+=4; p2+=4; }
-    if ((p1t<(matchlimit-1)) && (A16(p2) == A16(p1t))) { p1t+=2; p2+=2; }
-    if ((p1t<matchlimit) && (*p2 == *p1t)) p1t++;
-    return (p1t - p1);
-}
-
-
-FORCE_INLINE int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4, const BYTE* ip, const BYTE* const matchlimit, const BYTE** matchpos)
-{
-    U16* const chainTable = hc4->chainTable;
-    HTYPE* const HashTable = hc4->hashTable;
-    const BYTE* ref;
-    INITBASE(base,hc4->base);
-    int nbAttempts=MAX_NB_ATTEMPTS;
-    size_t repl=0, ml=0;
-    U16 delta=0;  // useless assignment, to remove an uninitialization warning
-
-    // HC4 match finder
-    LZ4HC_Insert(hc4, ip);
-    ref = HASH_POINTER(ip);
-
-#define REPEAT_OPTIMIZATION
-#ifdef REPEAT_OPTIMIZATION
-    // Detect repetitive sequences of length <= 4
-    if ((U32)(ip-ref) <= 4)        // potential repetition
-    {
-        if (A32(ref) == A32(ip))   // confirmed
-        {
-            delta = (U16)(ip-ref);
-            repl = ml  = LZ4HC_CommonLength(ip+MINMATCH, ref+MINMATCH, matchlimit) + MINMATCH;
-            *matchpos = ref;
-        }
-        ref = GETNEXT(ref);
-    }
-#endif
-
-    while (((U32)(ip-ref) <= MAX_DISTANCE) && (nbAttempts))
-    {
-        nbAttempts--;
-        if (*(ref+ml) == *(ip+ml))
-        if (A32(ref) == A32(ip))
-        {
-            size_t mlt = LZ4HC_CommonLength(ip+MINMATCH, ref+MINMATCH, matchlimit) + MINMATCH;
-            if (mlt > ml) { ml = mlt; *matchpos = ref; }
-        }
-        ref = GETNEXT(ref);
-    }
-
-#ifdef REPEAT_OPTIMIZATION
-    // Complete table
-    if (repl)
-    {
-        const BYTE* ptr = ip;
-        const BYTE* end;
-
-        end = ip + repl - (MINMATCH-1);
-        while(ptr < end-delta)
-        {
-            DELTANEXT(ptr) = delta;    // Pre-Load
-            ptr++;
-        }
-        do
-        {
-            DELTANEXT(ptr) = delta;    
-            HashTable[HASH_VALUE(ptr)] = (HTYPE)((ptr) - base);     // Head of chain
-            ptr++;
-        } while(ptr < end);
-        hc4->nextToUpdate = end;
-    }
-#endif 
-
-    return (int)ml;
-}
-
-
-FORCE_INLINE int LZ4HC_InsertAndGetWiderMatch (LZ4HC_Data_Structure* hc4, const BYTE* ip, const BYTE* startLimit, const BYTE* matchlimit, int longest, const BYTE** matchpos, const BYTE** startpos)
-{
-    U16* const  chainTable = hc4->chainTable;
-    HTYPE* const HashTable = hc4->hashTable;
-    INITBASE(base,hc4->base);
-    const BYTE*  ref;
-    int nbAttempts = MAX_NB_ATTEMPTS;
-    int delta = (int)(ip-startLimit);
-
-    // First Match
-    LZ4HC_Insert(hc4, ip);
-    ref = HASH_POINTER(ip);
-
-    while (((U32)(ip-ref) <= MAX_DISTANCE) && (nbAttempts))
-    {
-        nbAttempts--;
-        if (*(startLimit + longest) == *(ref - delta + longest))
-        if (A32(ref) == A32(ip))
-        {
-#if 1
-            const BYTE* reft = ref+MINMATCH;
-            const BYTE* ipt = ip+MINMATCH;
-            const BYTE* startt = ip;
-
-            while (ipt<matchlimit-(STEPSIZE-1))
-            {
-                UARCH diff = AARCH(reft) ^ AARCH(ipt);
-                if (!diff) { ipt+=STEPSIZE; reft+=STEPSIZE; continue; }
-                ipt += LZ4_NbCommonBytes(diff);
-                goto _endCount;
-            }
-            if (LZ4_ARCH64) if ((ipt<(matchlimit-3)) && (A32(reft) == A32(ipt))) { ipt+=4; reft+=4; }
-            if ((ipt<(matchlimit-1)) && (A16(reft) == A16(ipt))) { ipt+=2; reft+=2; }
-            if ((ipt<matchlimit) && (*reft == *ipt)) ipt++;
-_endCount:
-            reft = ref;
-#else
-            // Easier for code maintenance, but unfortunately slower too
-            const BYTE* startt = ip;
-            const BYTE* reft = ref;
-            const BYTE* ipt = ip + MINMATCH + LZ4HC_CommonLength(ip+MINMATCH, ref+MINMATCH, matchlimit);
-#endif
-
-            while ((startt>startLimit) && (reft > hc4->inputBuffer) && (startt[-1] == reft[-1])) {startt--; reft--;}
-
-            if ((ipt-startt) > longest)
-            {
-                longest = (int)(ipt-startt);
-                *matchpos = reft;
-                *startpos = startt;
-            }
-        }
-        ref = GETNEXT(ref);
-    }
-
-    return longest;
-}
-
-
-typedef enum { noLimit = 0, limitedOutput = 1 } limitedOutput_directive;
-
-FORCE_INLINE int LZ4HC_encodeSequence (
-                       const BYTE** ip,
-                       BYTE** op,
-                       const BYTE** anchor,
-                       int matchLength,
-                       const BYTE* ref,
-                       limitedOutput_directive limitedOutputBuffer,
-                       BYTE* oend)
-{
-    int length;
-    BYTE* token;
-
-    // Encode Literal length
-    length = (int)(*ip - *anchor);
-    token = (*op)++;
-    if ((limitedOutputBuffer) && ((*op + length + (2 + 1 + LASTLITERALS) + (length>>8)) > oend)) return 1;   // Check output limit
-    if (length>=(int)RUN_MASK) { int len; *token=(RUN_MASK<<ML_BITS); len = length-RUN_MASK; for(; len > 254 ; len-=255) *(*op)++ = 255;  *(*op)++ = (BYTE)len; } 
-    else *token = (BYTE)(length<<ML_BITS);
-
-    // Copy Literals
-    LZ4_BLINDCOPY(*anchor, *op, length);
-
-    // Encode Offset
-    LZ4_WRITE_LITTLEENDIAN_16(*op,(U16)(*ip-ref));
-
-    // Encode MatchLength
-    length = (int)(matchLength-MINMATCH);
-    if ((limitedOutputBuffer) && (*op + (1 + LASTLITERALS) + (length>>8) > oend)) return 1;   // Check output limit
-    if (length>=(int)ML_MASK) { *token+=ML_MASK; length-=ML_MASK; for(; length > 509 ; length-=510) { *(*op)++ = 255; *(*op)++ = 255; } if (length > 254) { length-=255; *(*op)++ = 255; } *(*op)++ = (BYTE)length; } 
-    else *token += (BYTE)(length);	
-
-    // Prepare next loop
-    *ip += matchLength;
-    *anchor = *ip; 
-
-    return 0;
-}
-
-
-static int LZ4HC_compress_generic (
-                 void* ctxvoid,
-                 const char* source, 
-                 char* dest,
-                 int inputSize,
-                 int maxOutputSize,
-                 limitedOutput_directive limit
-                )
-{
-    LZ4HC_Data_Structure* ctx = (LZ4HC_Data_Structure*) ctxvoid;
-    const BYTE* ip = (const BYTE*) source;
-    const BYTE* anchor = ip;
-    const BYTE* const iend = ip + inputSize;
-    const BYTE* const mflimit = iend - MFLIMIT;
-    const BYTE* const matchlimit = (iend - LASTLITERALS);
-
-    BYTE* op = (BYTE*) dest;
-    BYTE* const oend = op + maxOutputSize;
-
-    int   ml, ml2, ml3, ml0;
-    const BYTE* ref=NULL;
-    const BYTE* start2=NULL;
-    const BYTE* ref2=NULL;
-    const BYTE* start3=NULL;
-    const BYTE* ref3=NULL;
-    const BYTE* start0;
-    const BYTE* ref0;
-
-
-    // Ensure blocks follow each other
-    if (ip != ctx->end) return 0;
-    ctx->end += inputSize;
-
-    ip++;
-
-    // Main Loop
-    while (ip < mflimit)
-    {
-        ml = LZ4HC_InsertAndFindBestMatch (ctx, ip, matchlimit, (&ref));
-        if (!ml) { ip++; continue; }
-
-        // saved, in case we would skip too much
-        start0 = ip;
-        ref0 = ref;
-        ml0 = ml;
-
-_Search2:
-        if (ip+ml < mflimit)
-            ml2 = LZ4HC_InsertAndGetWiderMatch(ctx, ip + ml - 2, ip + 1, matchlimit, ml, &ref2, &start2);
-        else ml2 = ml;
-
-        if (ml2 == ml)  // No better match
-        {
-            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
-            continue;
-        }
-
-        if (start0 < ip)
-        {
-            if (start2 < ip + ml0)   // empirical
-            {
-                ip = start0;
-                ref = ref0;
-                ml = ml0;
-            }
-        }
-
-        // Here, start0==ip
-        if ((start2 - ip) < 3)   // First Match too small : removed
-        {
-            ml = ml2;
-            ip = start2;
-            ref =ref2;
-            goto _Search2;
-        }
-
-_Search3:
-        // Currently we have :
-        // ml2 > ml1, and
-        // ip1+3 <= ip2 (usually < ip1+ml1)
-        if ((start2 - ip) < OPTIMAL_ML)
-        {
-            int correction;
-            int new_ml = ml;
-            if (new_ml > OPTIMAL_ML) new_ml = OPTIMAL_ML;
-            if (ip+new_ml > start2 + ml2 - MINMATCH) new_ml = (int)(start2 - ip) + ml2 - MINMATCH;
-            correction = new_ml - (int)(start2 - ip);
-            if (correction > 0)
-            {
-                start2 += correction;
-                ref2 += correction;
-                ml2 -= correction;
-            }
-        }
-        // Now, we have start2 = ip+new_ml, with new_ml = min(ml, OPTIMAL_ML=18)
-
-        if (start2 + ml2 < mflimit)
-            ml3 = LZ4HC_InsertAndGetWiderMatch(ctx, start2 + ml2 - 3, start2, matchlimit, ml2, &ref3, &start3);
-        else ml3 = ml2;
-
-        if (ml3 == ml2) // No better match : 2 sequences to encode
-        {
-            // ip & ref are known; Now for ml
-            if (start2 < ip+ml)  ml = (int)(start2 - ip);
-            // Now, encode 2 sequences
-            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
-            ip = start2;
-            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml2, ref2, limit, oend)) return 0;
-            continue;
-        }
-
-        if (start3 < ip+ml+3) // Not enough space for match 2 : remove it
-        {
-            if (start3 >= (ip+ml)) // can write Seq1 immediately ==> Seq2 is removed, so Seq3 becomes Seq1
-            {
-                if (start2 < ip+ml)
-                {
-                    int correction = (int)(ip+ml - start2);
-                    start2 += correction;
-                    ref2 += correction;
-                    ml2 -= correction;
-                    if (ml2 < MINMATCH)
-                    {
-                        start2 = start3;
-                        ref2 = ref3;
-                        ml2 = ml3;
-                    }
-                }
-
-                if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
-                ip  = start3;
-                ref = ref3;
-                ml  = ml3;
-
-                start0 = start2;
-                ref0 = ref2;
-                ml0 = ml2;
-                goto _Search2;
-            }
-
-            start2 = start3;
-            ref2 = ref3;
-            ml2 = ml3;
-            goto _Search3;
-        }
-
-        // OK, now we have 3 ascending matches; let's write at least the first one
-        // ip & ref are known; Now for ml
-        if (start2 < ip+ml)
-        {
-            if ((start2 - ip) < (int)ML_MASK)
-            {
-                int correction;
-                if (ml > OPTIMAL_ML) ml = OPTIMAL_ML;
-                if (ip + ml > start2 + ml2 - MINMATCH) ml = (int)(start2 - ip) + ml2 - MINMATCH;
-                correction = ml - (int)(start2 - ip);
-                if (correction > 0)
-                {
-                    start2 += correction;
-                    ref2 += correction;
-                    ml2 -= correction;
-                }
-            }
-            else
-            {
-                ml = (int)(start2 - ip);
-            }
-        }
-        if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
-
-        ip = start2;
-        ref = ref2;
-        ml = ml2;
-
-        start2 = start3;
-        ref2 = ref3;
-        ml2 = ml3;
-
-        goto _Search3;
-
-    }
-
-    // Encode Last Literals
-    {
-        int lastRun = (int)(iend - anchor);
-        if ((limit) && (((char*)op - dest) + lastRun + 1 + ((lastRun+255-RUN_MASK)/255) > (U32)maxOutputSize)) return 0;  // Check output limit
-        if (lastRun>=(int)RUN_MASK) { *op++=(RUN_MASK<<ML_BITS); lastRun-=RUN_MASK; for(; lastRun > 254 ; lastRun-=255) *op++ = 255; *op++ = (BYTE) lastRun; } 
-        else *op++ = (BYTE)(lastRun<<ML_BITS);
-        memcpy(op, anchor, iend - anchor);
-        op += iend-anchor;
-    } 
-
-    // End
-    return (int) (((char*)op)-dest);
-}
-
-
-int LZ4_compressHC(const char* source, char* dest, int inputSize)
-{
-    void* ctx = LZ4_createHC(source);
-    int result;
-    if (ctx==NULL) return 0;
-
-    result = LZ4HC_compress_generic (ctx, source, dest, inputSize, 0, noLimit);
-
-    LZ4_freeHC(ctx);
-    return result;
-}
-
-int LZ4_compressHC_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize)
-{
-    return LZ4HC_compress_generic (LZ4HC_Data, source, dest, inputSize, 0, noLimit);
-}
-
-
-int LZ4_compressHC_limitedOutput(const char* source, char* dest, int inputSize, int maxOutputSize)
-{
-    void* ctx = LZ4_createHC(source);
-    int result;
-    if (ctx==NULL) return 0;
-
-    result = LZ4HC_compress_generic (ctx, source, dest, inputSize, maxOutputSize, limitedOutput);
-
-    LZ4_freeHC(ctx);
-    return result;
-}
-
-int LZ4_compressHC_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize)
-{
-    return LZ4HC_compress_generic (LZ4HC_Data, source, dest, inputSize, maxOutputSize, limitedOutput);
-}
-
+/*
+   LZ4 HC - High Compression Mode of LZ4
+   Copyright (C) 2011-2014, 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 :
+   - LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
+   - LZ4 source repository : http://code.google.com/p/lz4/
+*/
+
+
+
+/**************************************
+   Tuning Parameter
+**************************************/
+#define LZ4HC_DEFAULT_COMPRESSIONLEVEL 8
+
+
+/**************************************
+   Memory routines
+**************************************/
+#include <stdlib.h>   /* calloc, free */
+#define ALLOCATOR(s)  calloc(1,s)
+#define FREEMEM       free
+#include <string.h>   /* memset, memcpy */
+#define MEM_INIT      memset
+
+
+/**************************************
+   CPU Feature Detection
+**************************************/
+/* 32 or 64 bits ? */
+#if (defined(__x86_64__) || defined(_M_X64) || defined(_WIN64) \
+  || defined(__powerpc64__) || defined(__powerpc64le__) \
+  || defined(__ppc64__) || defined(__ppc64le__) \
+  || defined(__PPC64__) || defined(__PPC64LE__) \
+  || defined(__ia64) || defined(__itanium__) || defined(_M_IA64) )   /* Detects 64 bits mode */
+#  define LZ4_ARCH64 1
+#else
+#  define LZ4_ARCH64 0
+#endif
+
+/*
+ * Little Endian or Big Endian ?
+ * Overwrite the #define below if you know your architecture endianess
+ */
+#include <stdlib.h>   /* Apparently required to detect endianess */
+#if defined (__GLIBC__)
+#  include <endian.h>
+#  if (__BYTE_ORDER == __BIG_ENDIAN)
+#     define LZ4_BIG_ENDIAN 1
+#  endif
+#elif (defined(__BIG_ENDIAN__) || defined(__BIG_ENDIAN) || defined(_BIG_ENDIAN)) && !(defined(__LITTLE_ENDIAN__) || defined(__LITTLE_ENDIAN) || defined(_LITTLE_ENDIAN))
+#  define LZ4_BIG_ENDIAN 1
+#elif defined(__sparc) || defined(__sparc__) \
+   || defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) \
+   || defined(__hpux)  || defined(__hppa) \
+   || defined(_MIPSEB) || defined(__s390__)
+#  define LZ4_BIG_ENDIAN 1
+#else
+/* Little Endian assumed. PDP Endian and other very rare endian format are unsupported. */
+#endif
+
+/*
+ * 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
+ */
+#if defined(__ARM_FEATURE_UNALIGNED)
+#  define LZ4_FORCE_UNALIGNED_ACCESS 1
+#endif
+
+/* Define this parameter if your target system or compiler does not support hardware bit count */
+#if defined(_MSC_VER) && defined(_WIN32_WCE)            /* Visual Studio for Windows CE does not support Hardware bit count */
+#  define LZ4_FORCE_SW_BITCOUNT
+#endif
+
+
+/**************************************
+ Compiler Options
+**************************************/
+#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L)   /* C99 */
+/* "restrict" is a known keyword */
+#else
+#  define restrict /* Disable restrict */
+#endif
+
+#ifdef _MSC_VER    /* Visual Studio */
+#  define FORCE_INLINE static __forceinline
+#  include <intrin.h>                    /* For Visual 2005 */
+#  if LZ4_ARCH64   /* 64-bits */
+#    pragma intrinsic(_BitScanForward64) /* For Visual 2005 */
+#    pragma intrinsic(_BitScanReverse64) /* For Visual 2005 */
+#  else            /* 32-bits */
+#    pragma intrinsic(_BitScanForward)   /* For Visual 2005 */
+#    pragma intrinsic(_BitScanReverse)   /* For Visual 2005 */
+#  endif
+#  pragma warning(disable : 4127)        /* disable: C4127: conditional expression is constant */
+#  pragma warning(disable : 4701)        /* disable: C4701: potentially uninitialized local variable used */
+#else
+#  ifdef __GNUC__
+#    define FORCE_INLINE static inline __attribute__((always_inline))
+#  else
+#    define FORCE_INLINE static inline
+#  endif
+#endif
+
+#ifdef _MSC_VER  /* Visual Studio */
+#  define lz4_bswap16(x) _byteswap_ushort(x)
+#else
+#  define lz4_bswap16(x)  ((unsigned short int) ((((x) >> 8) & 0xffu) | (((x) & 0xffu) << 8)))
+#endif
+
+
+/**************************************
+   Includes
+**************************************/
+#include "lz4hc.h"
+#include "lz4.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(LZ4_FORCE_UNALIGNED_ACCESS)
+#  define _PACKED __attribute__ ((packed))
+#else
+#  define _PACKED
+#endif
+
+#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
+#  ifdef __IBMC__
+#    pragma pack(1)
+#  else
+#    pragma pack(push, 1)
+#  endif
+#endif
+
+typedef struct _U16_S { U16 v; } _PACKED U16_S;
+typedef struct _U32_S { U32 v; } _PACKED U32_S;
+typedef struct _U64_S { U64 v; } _PACKED U64_S;
+
+#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
+#  pragma pack(pop)
+#endif
+
+#define A64(x) (((U64_S *)(x))->v)
+#define A32(x) (((U32_S *)(x))->v)
+#define A16(x) (((U16_S *)(x))->v)
+
+
+/**************************************
+   Constants
+**************************************/
+#define MINMATCH 4
+
+#define DICTIONARY_LOGSIZE 16
+#define MAXD (1<<DICTIONARY_LOGSIZE)
+#define MAXD_MASK ((U32)(MAXD - 1))
+#define MAX_DISTANCE (MAXD - 1)
+
+#define HASH_LOG (DICTIONARY_LOGSIZE-1)
+#define HASHTABLESIZE (1 << HASH_LOG)
+#define HASH_MASK (HASHTABLESIZE - 1)
+
+#define ML_BITS  4
+#define ML_MASK  (size_t)((1U<<ML_BITS)-1)
+#define RUN_BITS (8-ML_BITS)
+#define RUN_MASK ((1U<<RUN_BITS)-1)
+
+#define COPYLENGTH 8
+#define LASTLITERALS 5
+#define MFLIMIT (COPYLENGTH+MINMATCH)
+#define MINLENGTH (MFLIMIT+1)
+#define OPTIMAL_ML (int)((ML_MASK-1)+MINMATCH)
+
+#define KB *(1U<<10)
+#define MB *(1U<<20)
+#define GB *(1U<<30)
+
+
+/**************************************
+   Architecture-specific macros
+**************************************/
+#if LZ4_ARCH64   /* 64-bit */
+#  define STEPSIZE 8
+#  define LZ4_COPYSTEP(s,d)     A64(d) = A64(s); d+=8; s+=8;
+#  define LZ4_COPYPACKET(s,d)   LZ4_COPYSTEP(s,d)
+#  define AARCH A64
+#  define HTYPE                 U32
+#  define INITBASE(b,s)         const BYTE* const b = s
+#else            /* 32-bit */
+#  define STEPSIZE 4
+#  define LZ4_COPYSTEP(s,d)     A32(d) = A32(s); d+=4; s+=4;
+#  define LZ4_COPYPACKET(s,d)   LZ4_COPYSTEP(s,d); LZ4_COPYSTEP(s,d);
+#  define AARCH A32
+#  define HTYPE                 U32
+#  define INITBASE(b,s)         const BYTE* const b = s
+#endif
+
+#if defined(LZ4_BIG_ENDIAN)
+#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { U16 v = A16(p); v = lz4_bswap16(v); d = (s) - v; }
+#  define LZ4_WRITE_LITTLEENDIAN_16(p,i)  { U16 v = (U16)(i); v = lz4_bswap16(v); A16(p) = v; p+=2; }
+#else      /* Little Endian */
+#  define LZ4_READ_LITTLEENDIAN_16(d,s,p) { d = (s) - A16(p); }
+#  define LZ4_WRITE_LITTLEENDIAN_16(p,v)  { A16(p) = v; p+=2; }
+#endif
+
+
+/**************************************
+   Local Types
+**************************************/
+typedef struct
+{
+    const BYTE* inputBuffer;
+    const BYTE* base;
+    const BYTE* end;
+    HTYPE hashTable[HASHTABLESIZE];
+    U16 chainTable[MAXD];
+    const BYTE* nextToUpdate;
+} LZ4HC_Data_Structure;
+
+
+/**************************************
+   Macros
+**************************************/
+#define LZ4_WILDCOPY(s,d,e)    do { LZ4_COPYPACKET(s,d) } while (d<e);
+#define LZ4_BLINDCOPY(s,d,l)   { BYTE* e=d+l; LZ4_WILDCOPY(s,d,e); d=e; }
+#define HASH_FUNCTION(i)       (((i) * 2654435761U) >> ((MINMATCH*8)-HASH_LOG))
+#define HASH_VALUE(p)          HASH_FUNCTION(A32(p))
+#define HASH_POINTER(p)        (HashTable[HASH_VALUE(p)] + base)
+#define DELTANEXT(p)           chainTable[(size_t)(p) & MAXD_MASK]
+#define GETNEXT(p)             ((p) - (size_t)DELTANEXT(p))
+
+
+/**************************************
+ Private functions
+**************************************/
+#if LZ4_ARCH64
+
+FORCE_INLINE int LZ4_NbCommonBytes (register U64 val)
+{
+#if defined(LZ4_BIG_ENDIAN)
+#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r = 0;
+    _BitScanReverse64( &r, val );
+    return (int)(r>>3);
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_clzll(val) >> 3);
+#  else
+    int r;
+    if (!(val>>32)) { r=4; } else { r=0; val>>=32; }
+    if (!(val>>16)) { r+=2; val>>=8; } else { val>>=24; }
+    r += (!val);
+    return r;
+#  endif
+#else
+#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r = 0;
+    _BitScanForward64( &r, val );
+    return (int)(r>>3);
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_ctzll(val) >> 3);
+#  else
+    static const int DeBruijnBytePos[64] = { 0, 0, 0, 0, 0, 1, 1, 2, 0, 3, 1, 3, 1, 4, 2, 7, 0, 2, 3, 6, 1, 5, 3, 5, 1, 3, 4, 4, 2, 5, 6, 7, 7, 0, 1, 2, 3, 3, 4, 6, 2, 6, 5, 5, 3, 4, 5, 6, 7, 1, 2, 4, 6, 4, 4, 5, 7, 2, 6, 5, 7, 6, 7, 7 };
+    return DeBruijnBytePos[((U64)((val & -val) * 0x0218A392CDABBD3F)) >> 58];
+#  endif
+#endif
+}
+
+#else
+
+FORCE_INLINE int LZ4_NbCommonBytes (register U32 val)
+{
+#if defined(LZ4_BIG_ENDIAN)
+#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r;
+    _BitScanReverse( &r, val );
+    return (int)(r>>3);
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_clz(val) >> 3);
+#  else
+    int r;
+    if (!(val>>16)) { r=2; val>>=8; } else { r=0; val>>=24; }
+    r += (!val);
+    return r;
+#  endif
+#else
+#  if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    unsigned long r;
+    _BitScanForward( &r, val );
+    return (int)(r>>3);
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
+    return (__builtin_ctz(val) >> 3);
+#  else
+    static const int DeBruijnBytePos[32] = { 0, 0, 3, 0, 3, 1, 3, 0, 3, 2, 2, 1, 3, 2, 0, 1, 3, 3, 1, 2, 2, 2, 2, 0, 3, 1, 2, 0, 1, 0, 1, 1 };
+    return DeBruijnBytePos[((U32)((val & -(S32)val) * 0x077CB531U)) >> 27];
+#  endif
+#endif
+}
+
+#endif
+
+
+int LZ4_sizeofStreamStateHC()
+{
+    return sizeof(LZ4HC_Data_Structure);
+}
+
+FORCE_INLINE void LZ4_initHC (LZ4HC_Data_Structure* hc4, const BYTE* base)
+{
+    MEM_INIT((void*)hc4->hashTable, 0, sizeof(hc4->hashTable));
+    MEM_INIT(hc4->chainTable, 0xFF, sizeof(hc4->chainTable));
+    hc4->nextToUpdate = base + 1;
+    hc4->base = base;
+    hc4->inputBuffer = base;
+    hc4->end = base;
+}
+
+int LZ4_resetStreamStateHC(void* state, const char* inputBuffer)
+{
+    if ((((size_t)state) & (sizeof(void*)-1)) != 0) return 1;   /* Error : pointer is not aligned for pointer (32 or 64 bits) */
+    LZ4_initHC((LZ4HC_Data_Structure*)state, (const BYTE*)inputBuffer);
+    return 0;
+}
+
+
+void* LZ4_createHC (const char* inputBuffer)
+{
+    void* hc4 = ALLOCATOR(sizeof(LZ4HC_Data_Structure));
+    LZ4_initHC ((LZ4HC_Data_Structure*)hc4, (const BYTE*)inputBuffer);
+    return hc4;
+}
+
+
+int LZ4_freeHC (void* LZ4HC_Data)
+{
+    FREEMEM(LZ4HC_Data);
+    return (0);
+}
+
+
+/* Update chains up to ip (excluded) */
+FORCE_INLINE void LZ4HC_Insert (LZ4HC_Data_Structure* hc4, const BYTE* ip)
+{
+    U16*   chainTable = hc4->chainTable;
+    HTYPE* HashTable  = hc4->hashTable;
+    INITBASE(base,hc4->base);
+
+    while(hc4->nextToUpdate < ip)
+    {
+        const BYTE* const p = hc4->nextToUpdate;
+        size_t delta = (p) - HASH_POINTER(p);
+        if (delta>MAX_DISTANCE) delta = MAX_DISTANCE;
+        DELTANEXT(p) = (U16)delta;
+        HashTable[HASH_VALUE(p)] = (HTYPE)((p) - base);
+        hc4->nextToUpdate++;
+    }
+}
+
+
+char* LZ4_slideInputBufferHC(void* LZ4HC_Data)
+{
+    LZ4HC_Data_Structure* hc4 = (LZ4HC_Data_Structure*)LZ4HC_Data;
+    U32 distance = (U32)(hc4->end - hc4->inputBuffer) - 64 KB;
+    distance = (distance >> 16) << 16;   /* Must be a multiple of 64 KB */
+    LZ4HC_Insert(hc4, hc4->end - MINMATCH);
+    memcpy((void*)(hc4->end - 64 KB - distance), (const void*)(hc4->end - 64 KB), 64 KB);
+    hc4->nextToUpdate -= distance;
+    hc4->base -= distance;
+    if ((U32)(hc4->inputBuffer - hc4->base) > 1 GB + 64 KB)   /* Avoid overflow */
+    {
+        int i;
+        hc4->base += 1 GB;
+        for (i=0; i<HASHTABLESIZE; i++) hc4->hashTable[i] -= 1 GB;
+    }
+    hc4->end -= distance;
+    return (char*)(hc4->end);
+}
+
+
+FORCE_INLINE size_t LZ4HC_CommonLength (const BYTE* p1, const BYTE* p2, const BYTE* const matchlimit)
+{
+    const BYTE* p1t = p1;
+
+    while (p1t<matchlimit-(STEPSIZE-1))
+    {
+        size_t diff = AARCH(p2) ^ AARCH(p1t);
+        if (!diff) { p1t+=STEPSIZE; p2+=STEPSIZE; continue; }
+        p1t += LZ4_NbCommonBytes(diff);
+        return (p1t - p1);
+    }
+    if (LZ4_ARCH64) if ((p1t<(matchlimit-3)) && (A32(p2) == A32(p1t))) { p1t+=4; p2+=4; }
+    if ((p1t<(matchlimit-1)) && (A16(p2) == A16(p1t))) { p1t+=2; p2+=2; }
+    if ((p1t<matchlimit) && (*p2 == *p1t)) p1t++;
+    return (p1t - p1);
+}
+
+
+FORCE_INLINE int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4, const BYTE* ip, const BYTE* const matchlimit, const BYTE** matchpos, const int maxNbAttempts)
+{
+    U16* const chainTable = hc4->chainTable;
+    HTYPE* const HashTable = hc4->hashTable;
+    const BYTE* ref;
+    INITBASE(base,hc4->base);
+    int nbAttempts=maxNbAttempts;
+    size_t repl=0, ml=0;
+    U16 delta=0;  /* useless assignment, to remove an uninitialization warning */
+
+    /* HC4 match finder */
+    LZ4HC_Insert(hc4, ip);
+    ref = HASH_POINTER(ip);
+
+#define REPEAT_OPTIMIZATION
+#ifdef REPEAT_OPTIMIZATION
+    /* Detect repetitive sequences of length <= 4 */
+    if ((U32)(ip-ref) <= 4)        /* potential repetition */
+    {
+        if (A32(ref) == A32(ip))   /* confirmed */
+        {
+            delta = (U16)(ip-ref);
+            repl = ml  = LZ4HC_CommonLength(ip+MINMATCH, ref+MINMATCH, matchlimit) + MINMATCH;
+            *matchpos = ref;
+        }
+        ref = GETNEXT(ref);
+    }
+#endif
+
+    while (((U32)(ip-ref) <= MAX_DISTANCE) && (nbAttempts))
+    {
+        nbAttempts--;
+        if (*(ref+ml) == *(ip+ml))
+        if (A32(ref) == A32(ip))
+        {
+            size_t mlt = LZ4HC_CommonLength(ip+MINMATCH, ref+MINMATCH, matchlimit) + MINMATCH;
+            if (mlt > ml) { ml = mlt; *matchpos = ref; }
+        }
+        ref = GETNEXT(ref);
+    }
+
+#ifdef REPEAT_OPTIMIZATION
+    /* Complete table */
+    if (repl)
+    {
+        const BYTE* ptr = ip;
+        const BYTE* end;
+
+        end = ip + repl - (MINMATCH-1);
+        while(ptr < end-delta)
+        {
+            DELTANEXT(ptr) = delta;    /* Pre-Load */
+            ptr++;
+        }
+        do
+        {
+            DELTANEXT(ptr) = delta;
+            HashTable[HASH_VALUE(ptr)] = (HTYPE)((ptr) - base);     /* Head of chain */
+            ptr++;
+        } while(ptr < end);
+        hc4->nextToUpdate = end;
+    }
+#endif
+
+    return (int)ml;
+}
+
+
+FORCE_INLINE int LZ4HC_InsertAndGetWiderMatch (LZ4HC_Data_Structure* hc4, const BYTE* ip, const BYTE* startLimit, const BYTE* matchlimit, int longest, const BYTE** matchpos, const BYTE** startpos, const int maxNbAttempts)
+{
+    U16* const  chainTable = hc4->chainTable;
+    HTYPE* const HashTable = hc4->hashTable;
+    INITBASE(base,hc4->base);
+    const BYTE*  ref;
+    int nbAttempts = maxNbAttempts;
+    int delta = (int)(ip-startLimit);
+
+    /* First Match */
+    LZ4HC_Insert(hc4, ip);
+    ref = HASH_POINTER(ip);
+
+    while (((U32)(ip-ref) <= MAX_DISTANCE) && (nbAttempts))
+    {
+        nbAttempts--;
+        if (*(startLimit + longest) == *(ref - delta + longest))
+        if (A32(ref) == A32(ip))
+        {
+#if 1
+            const BYTE* reft = ref+MINMATCH;
+            const BYTE* ipt = ip+MINMATCH;
+            const BYTE* startt = ip;
+
+            while (ipt<matchlimit-(STEPSIZE-1))
+            {
+                size_t diff = AARCH(reft) ^ AARCH(ipt);
+                if (!diff) { ipt+=STEPSIZE; reft+=STEPSIZE; continue; }
+                ipt += LZ4_NbCommonBytes(diff);
+                goto _endCount;
+            }
+            if (LZ4_ARCH64) if ((ipt<(matchlimit-3)) && (A32(reft) == A32(ipt))) { ipt+=4; reft+=4; }
+            if ((ipt<(matchlimit-1)) && (A16(reft) == A16(ipt))) { ipt+=2; reft+=2; }
+            if ((ipt<matchlimit) && (*reft == *ipt)) ipt++;
+_endCount:
+            reft = ref;
+#else
+            /* Easier for code maintenance, but unfortunately slower too */
+            const BYTE* startt = ip;
+            const BYTE* reft = ref;
+            const BYTE* ipt = ip + MINMATCH + LZ4HC_CommonLength(ip+MINMATCH, ref+MINMATCH, matchlimit);
+#endif
+
+            while ((startt>startLimit) && (reft > hc4->inputBuffer) && (startt[-1] == reft[-1])) {startt--; reft--;}
+
+            if ((ipt-startt) > longest)
+            {
+                longest = (int)(ipt-startt);
+                *matchpos = reft;
+                *startpos = startt;
+            }
+        }
+        ref = GETNEXT(ref);
+    }
+
+    return longest;
+}
+
+
+typedef enum { noLimit = 0, limitedOutput = 1 } limitedOutput_directive;
+
+FORCE_INLINE int LZ4HC_encodeSequence (
+                       const BYTE** ip,
+                       BYTE** op,
+                       const BYTE** anchor,
+                       int matchLength,
+                       const BYTE* ref,
+                       limitedOutput_directive limitedOutputBuffer,
+                       BYTE* oend)
+{
+    int length;
+    BYTE* token;
+
+    /* Encode Literal length */
+    length = (int)(*ip - *anchor);
+    token = (*op)++;
+    if ((limitedOutputBuffer) && ((*op + length + (2 + 1 + LASTLITERALS) + (length>>8)) > oend)) return 1;   /* Check output limit */
+    if (length>=(int)RUN_MASK) { int len; *token=(RUN_MASK<<ML_BITS); len = length-RUN_MASK; for(; len > 254 ; len-=255) *(*op)++ = 255;  *(*op)++ = (BYTE)len; }
+    else *token = (BYTE)(length<<ML_BITS);
+
+    /* Copy Literals */
+    LZ4_BLINDCOPY(*anchor, *op, length);
+
+    /* Encode Offset */
+    LZ4_WRITE_LITTLEENDIAN_16(*op,(U16)(*ip-ref));
+
+    /* Encode MatchLength */
+    length = (int)(matchLength-MINMATCH);
+    if ((limitedOutputBuffer) && (*op + (1 + LASTLITERALS) + (length>>8) > oend)) return 1;   /* Check output limit */
+    if (length>=(int)ML_MASK) { *token+=ML_MASK; length-=ML_MASK; for(; length > 509 ; length-=510) { *(*op)++ = 255; *(*op)++ = 255; } if (length > 254) { length-=255; *(*op)++ = 255; } *(*op)++ = (BYTE)length; }
+    else *token += (BYTE)(length);
+
+    /* Prepare next loop */
+    *ip += matchLength;
+    *anchor = *ip;
+
+    return 0;
+}
+
+
+#define MAX_COMPRESSION_LEVEL 16
+static int LZ4HC_compress_generic (
+                 void* ctxvoid,
+                 const char* source,
+                 char* dest,
+                 int inputSize,
+                 int maxOutputSize,
+                 int compressionLevel,
+                 limitedOutput_directive limit
+                )
+{
+    LZ4HC_Data_Structure* ctx = (LZ4HC_Data_Structure*) ctxvoid;
+    const BYTE* ip = (const BYTE*) source;
+    const BYTE* anchor = ip;
+    const BYTE* const iend = ip + inputSize;
+    const BYTE* const mflimit = iend - MFLIMIT;
+    const BYTE* const matchlimit = (iend - LASTLITERALS);
+
+    BYTE* op = (BYTE*) dest;
+    BYTE* const oend = op + maxOutputSize;
+
+    const int maxNbAttempts = compressionLevel > MAX_COMPRESSION_LEVEL ? 1 << MAX_COMPRESSION_LEVEL : compressionLevel ? 1<<(compressionLevel-1) : 1<<LZ4HC_DEFAULT_COMPRESSIONLEVEL;
+    int   ml, ml2, ml3, ml0;
+    const BYTE* ref=NULL;
+    const BYTE* start2=NULL;
+    const BYTE* ref2=NULL;
+    const BYTE* start3=NULL;
+    const BYTE* ref3=NULL;
+    const BYTE* start0;
+    const BYTE* ref0;
+
+
+    /* Ensure blocks follow each other */
+    if (ip != ctx->end) return 0;
+    ctx->end += inputSize;
+
+    ip++;
+
+    /* Main Loop */
+    while (ip < mflimit)
+    {
+        ml = LZ4HC_InsertAndFindBestMatch (ctx, ip, matchlimit, (&ref), maxNbAttempts);
+        if (!ml) { ip++; continue; }
+
+        /* saved, in case we would skip too much */
+        start0 = ip;
+        ref0 = ref;
+        ml0 = ml;
+
+_Search2:
+        if (ip+ml < mflimit)
+            ml2 = LZ4HC_InsertAndGetWiderMatch(ctx, ip + ml - 2, ip + 1, matchlimit, ml, &ref2, &start2, maxNbAttempts);
+        else ml2 = ml;
+
+        if (ml2 == ml)  /* No better match */
+        {
+            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
+            continue;
+        }
+
+        if (start0 < ip)
+        {
+            if (start2 < ip + ml0)   /* empirical */
+            {
+                ip = start0;
+                ref = ref0;
+                ml = ml0;
+            }
+        }
+
+        /* Here, start0==ip */
+        if ((start2 - ip) < 3)   /* First Match too small : removed */
+        {
+            ml = ml2;
+            ip = start2;
+            ref =ref2;
+            goto _Search2;
+        }
+
+_Search3:
+        /*
+         * Currently we have :
+         * ml2 > ml1, and
+         * ip1+3 <= ip2 (usually < ip1+ml1)
+         */
+        if ((start2 - ip) < OPTIMAL_ML)
+        {
+            int correction;
+            int new_ml = ml;
+            if (new_ml > OPTIMAL_ML) new_ml = OPTIMAL_ML;
+            if (ip+new_ml > start2 + ml2 - MINMATCH) new_ml = (int)(start2 - ip) + ml2 - MINMATCH;
+            correction = new_ml - (int)(start2 - ip);
+            if (correction > 0)
+            {
+                start2 += correction;
+                ref2 += correction;
+                ml2 -= correction;
+            }
+        }
+        /* Now, we have start2 = ip+new_ml, with new_ml = min(ml, OPTIMAL_ML=18) */
+
+        if (start2 + ml2 < mflimit)
+            ml3 = LZ4HC_InsertAndGetWiderMatch(ctx, start2 + ml2 - 3, start2, matchlimit, ml2, &ref3, &start3, maxNbAttempts);
+        else ml3 = ml2;
+
+        if (ml3 == ml2) /* No better match : 2 sequences to encode */
+        {
+            /* ip & ref are known; Now for ml */
+            if (start2 < ip+ml)  ml = (int)(start2 - ip);
+            /* Now, encode 2 sequences */
+            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
+            ip = start2;
+            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml2, ref2, limit, oend)) return 0;
+            continue;
+        }
+
+        if (start3 < ip+ml+3) /* Not enough space for match 2 : remove it */
+        {
+            if (start3 >= (ip+ml)) /* can write Seq1 immediately ==> Seq2 is removed, so Seq3 becomes Seq1 */
+            {
+                if (start2 < ip+ml)
+                {
+                    int correction = (int)(ip+ml - start2);
+                    start2 += correction;
+                    ref2 += correction;
+                    ml2 -= correction;
+                    if (ml2 < MINMATCH)
+                    {
+                        start2 = start3;
+                        ref2 = ref3;
+                        ml2 = ml3;
+                    }
+                }
+
+                if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
+                ip  = start3;
+                ref = ref3;
+                ml  = ml3;
+
+                start0 = start2;
+                ref0 = ref2;
+                ml0 = ml2;
+                goto _Search2;
+            }
+
+            start2 = start3;
+            ref2 = ref3;
+            ml2 = ml3;
+            goto _Search3;
+        }
+
+        /*
+         * OK, now we have 3 ascending matches; let's write at least the first one
+         * ip & ref are known; Now for ml
+         */
+        if (start2 < ip+ml)
+        {
+            if ((start2 - ip) < (int)ML_MASK)
+            {
+                int correction;
+                if (ml > OPTIMAL_ML) ml = OPTIMAL_ML;
+                if (ip + ml > start2 + ml2 - MINMATCH) ml = (int)(start2 - ip) + ml2 - MINMATCH;
+                correction = ml - (int)(start2 - ip);
+                if (correction > 0)
+                {
+                    start2 += correction;
+                    ref2 += correction;
+                    ml2 -= correction;
+                }
+            }
+            else
+            {
+                ml = (int)(start2 - ip);
+            }
+        }
+        if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
+
+        ip = start2;
+        ref = ref2;
+        ml = ml2;
+
+        start2 = start3;
+        ref2 = ref3;
+        ml2 = ml3;
+
+        goto _Search3;
+
+    }
+
+    /* Encode Last Literals */
+    {
+        int lastRun = (int)(iend - anchor);
+        if ((limit) && (((char*)op - dest) + lastRun + 1 + ((lastRun+255-RUN_MASK)/255) > (U32)maxOutputSize)) return 0;  /* Check output limit */
+        if (lastRun>=(int)RUN_MASK) { *op++=(RUN_MASK<<ML_BITS); lastRun-=RUN_MASK; for(; lastRun > 254 ; lastRun-=255) *op++ = 255; *op++ = (BYTE) lastRun; }
+        else *op++ = (BYTE)(lastRun<<ML_BITS);
+        memcpy(op, anchor, iend - anchor);
+        op += iend-anchor;
+    }
+
+    /* End */
+    return (int) (((char*)op)-dest);
+}
+
+
+int LZ4_compressHC2(const char* source, char* dest, int inputSize, int compressionLevel)
+{
+    void* ctx = LZ4_createHC(source);
+    int result;
+    if (ctx==NULL) return 0;
+
+    result = LZ4HC_compress_generic (ctx, source, dest, inputSize, 0, compressionLevel, noLimit);
+
+    LZ4_freeHC(ctx);
+    return result;
+}
+
+int LZ4_compressHC(const char* source, char* dest, int inputSize) { return LZ4_compressHC2(source, dest, inputSize, 0); }
+
+int LZ4_compressHC2_limitedOutput(const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel)
+{
+    void* ctx = LZ4_createHC(source);
+    int result;
+    if (ctx==NULL) return 0;
+
+    result = LZ4HC_compress_generic (ctx, source, dest, inputSize, maxOutputSize, compressionLevel, limitedOutput);
+
+    LZ4_freeHC(ctx);
+    return result;
+}
+
+int LZ4_compressHC_limitedOutput(const char* source, char* dest, int inputSize, int maxOutputSize)
+{
+    return LZ4_compressHC2_limitedOutput(source, dest, inputSize, maxOutputSize, 0);
+}
+
+
+/*****************************
+   Using external allocation
+*****************************/
+int LZ4_sizeofStateHC() { return sizeof(LZ4HC_Data_Structure); }
+
+
+int LZ4_compressHC2_withStateHC (void* state, const char* source, char* dest, int inputSize, int compressionLevel)
+{
+    if (((size_t)(state)&(sizeof(void*)-1)) != 0) return 0;   /* Error : state is not aligned for pointers (32 or 64 bits) */
+    LZ4_initHC ((LZ4HC_Data_Structure*)state, (const BYTE*)source);
+    return LZ4HC_compress_generic (state, source, dest, inputSize, 0, compressionLevel, noLimit);
+}
+
+int LZ4_compressHC_withStateHC (void* state, const char* source, char* dest, int inputSize)
+{ return LZ4_compressHC2_withStateHC (state, source, dest, inputSize, 0); }
+
+
+int LZ4_compressHC2_limitedOutput_withStateHC (void* state, const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel)
+{
+    if (((size_t)(state)&(sizeof(void*)-1)) != 0) return 0;   /* Error : state is not aligned for pointers (32 or 64 bits) */
+    LZ4_initHC ((LZ4HC_Data_Structure*)state, (const BYTE*)source);
+    return LZ4HC_compress_generic (state, source, dest, inputSize, maxOutputSize, compressionLevel, limitedOutput);
+}
+
+int LZ4_compressHC_limitedOutput_withStateHC (void* state, const char* source, char* dest, int inputSize, int maxOutputSize)
+{ return LZ4_compressHC2_limitedOutput_withStateHC (state, source, dest, inputSize, maxOutputSize, 0); }
+
+
+/****************************
+   Stream functions
+****************************/
+
+int LZ4_compressHC_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize)
+{
+    return LZ4HC_compress_generic (LZ4HC_Data, source, dest, inputSize, 0, 0, noLimit);
+}
+
+int LZ4_compressHC2_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int compressionLevel)
+{
+    return LZ4HC_compress_generic (LZ4HC_Data, source, dest, inputSize, 0, compressionLevel, noLimit);
+}
+
+int LZ4_compressHC_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize)
+{
+    return LZ4HC_compress_generic (LZ4HC_Data, source, dest, inputSize, maxOutputSize, 0, limitedOutput);
+}
+
+int LZ4_compressHC2_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel)
+{
+    return LZ4HC_compress_generic (LZ4HC_Data, source, dest, inputSize, maxOutputSize, compressionLevel, limitedOutput);
+}

ext/lz4/lz4hc.h

@@ -1,111 +1,173 @@
-/*
-   LZ4 HC - High Compression Mode of LZ4
-   Header File
-   Copyright (C) 2011-2013, 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 :
-   - LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
-   - LZ4 source repository : http://code.google.com/p/lz4/
-*/
-#pragma once
-
-
-#if defined (__cplusplus)
-extern "C" {
-#endif
-
-
-int LZ4_compressHC (const char* source, char* dest, int inputSize);
-/*
-LZ4_compressHC :
-    return : the number of bytes in compressed buffer dest
-             or 0 if compression fails.
-    note : destination buffer must be already allocated. 
-        To avoid any problem, size it to handle worst cases situations (input data not compressible)
-        Worst case size evaluation is provided by function LZ4_compressBound() (see "lz4.h")
-*/
-
-int LZ4_compressHC_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
-/*
-LZ4_compress_limitedOutput() :
-    Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
-    If it cannot achieve it, compression will stop, and result of the function will be zero.
-    This function never writes outside of provided output buffer.
-
-    inputSize  : Max supported value is 1 GB
-    maxOutputSize : is maximum allowed size into the destination buffer (which must be already allocated)
-    return : the number of output bytes written in buffer 'dest'
-             or 0 if compression fails.
-*/
-
-
-/* Note :
-Decompression functions are provided within LZ4 source code (see "lz4.h") (BSD license)
-*/
-
-
-/* Advanced Functions */
-
-void* LZ4_createHC (const char* inputBuffer);
-int   LZ4_compressHC_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize);
-int   LZ4_compressHC_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize);
-char* LZ4_slideInputBufferHC (void* LZ4HC_Data);
-int   LZ4_freeHC (void* LZ4HC_Data);
-
-/* 
-These functions allow the compression of dependent blocks, where each block benefits from prior 64 KB within preceding blocks.
-In order to achieve this, it is necessary to start creating the LZ4HC Data Structure, thanks to the function :
-
-void* LZ4_createHC (const char* inputBuffer);
-The result of the function is the (void*) pointer on the LZ4HC Data Structure.
-This pointer will be needed in all other functions.
-If the pointer returned is NULL, then the allocation has failed, and compression must be aborted.
-The only parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
-The input buffer must be already allocated, and size at least 192KB.
-'inputBuffer' will also be the 'const char* source' of the first block.
-
-All blocks are expected to lay next to each other within the input buffer, starting from 'inputBuffer'.
-To compress each block, use either LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue().
-Their behavior are identical to LZ4_compressHC() or LZ4_compressHC_limitedOutput(), 
-but require the LZ4HC Data Structure as their first argument, and check that each block starts right after the previous one.
-If next block does not begin immediately after the previous one, the compression will fail (return 0).
-
-When it's no longer possible to lay the next block after the previous one (not enough space left into input buffer), a call to : 
-char* LZ4_slideInputBufferHC(void* LZ4HC_Data);
-must be performed. It will typically copy the latest 64KB of input at the beginning of input buffer.
-Note that, for this function to work properly, minimum size of an input buffer must be 192KB.
-==> The memory position where the next input data block must start is provided as the result of the function.
-
-Compression can then resume, using LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue(), as usual.
-
-When compression is completed, a call to LZ4_freeHC() will release the memory used by the LZ4HC Data Structure.
-*/
-
-
-#if defined (__cplusplus)
-}
-#endif
+/*
+   LZ4 HC - High Compression Mode of LZ4
+   Header File
+   Copyright (C) 2011-2014, 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 :
+   - LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
+   - LZ4 source repository : http://code.google.com/p/lz4/
+*/
+#pragma once
+
+
+#if defined (__cplusplus)
+extern "C" {
+#endif
+
+
+int LZ4_compressHC (const char* source, char* dest, int inputSize);
+/*
+LZ4_compressHC :
+    return : the number of bytes in compressed buffer dest
+             or 0 if compression fails.
+    note : destination buffer must be already allocated.
+        To avoid any problem, size it to handle worst cases situations (input data not compressible)
+        Worst case size evaluation is provided by function LZ4_compressBound() (see "lz4.h")
+*/
+
+int LZ4_compressHC_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
+/*
+LZ4_compress_limitedOutput() :
+    Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
+    If it cannot achieve it, compression will stop, and result of the function will be zero.
+    This function never writes outside of provided output buffer.
+
+    inputSize  : Max supported value is 1 GB
+    maxOutputSize : is maximum allowed size into the destination buffer (which must be already allocated)
+    return : the number of output bytes written in buffer 'dest'
+             or 0 if compression fails.
+*/
+
+
+int LZ4_compressHC2 (const char* source, char* dest, int inputSize, int compressionLevel);
+int LZ4_compressHC2_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel);
+/*
+    Same functions as above, but with programmable 'compressionLevel'.
+    Recommended values are between 4 and 9, although any value between 0 and 16 will work.
+    'compressionLevel'==0 means use default 'compressionLevel' value.
+    Values above 16 behave the same as 16.
+    Equivalent variants exist for all other compression functions below.
+*/
+
+/* Note :
+Decompression functions are provided within LZ4 source code (see "lz4.h") (BSD license)
+*/
+
+
+/**************************************
+   Using an external allocation
+**************************************/
+int LZ4_sizeofStateHC(void);
+int LZ4_compressHC_withStateHC               (void* state, const char* source, char* dest, int inputSize);
+int LZ4_compressHC_limitedOutput_withStateHC (void* state, const char* source, char* dest, int inputSize, int maxOutputSize);
+
+int LZ4_compressHC2_withStateHC              (void* state, const char* source, char* dest, int inputSize, int compressionLevel);
+int LZ4_compressHC2_limitedOutput_withStateHC(void* state, const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel);
+
+/*
+These functions are provided should you prefer to allocate memory for compression tables with your own allocation methods.
+To know how much memory must be allocated for the compression tables, use :
+int LZ4_sizeofStateHC();
+
+Note that tables must be aligned for pointer (32 or 64 bits), otherwise compression will fail (return code 0).
+
+The allocated memory can be provided to the compressions functions using 'void* state' parameter.
+LZ4_compress_withStateHC() and LZ4_compress_limitedOutput_withStateHC() are equivalent to previously described functions.
+They just use the externally allocated memory area instead of allocating their own (on stack, or on heap).
+*/
+
+
+/**************************************
+   Streaming Functions
+**************************************/
+/* Note : these streaming functions still follows the older model */
+void* LZ4_createHC (const char* inputBuffer);
+int   LZ4_compressHC_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize);
+int   LZ4_compressHC_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize);
+char* LZ4_slideInputBufferHC (void* LZ4HC_Data);
+int   LZ4_freeHC (void* LZ4HC_Data);
+
+int   LZ4_compressHC2_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int compressionLevel);
+int   LZ4_compressHC2_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel);
+
+/*
+These functions allow the compression of dependent blocks, where each block benefits from prior 64 KB within preceding blocks.
+In order to achieve this, it is necessary to start creating the LZ4HC Data Structure, thanks to the function :
+
+void* LZ4_createHC (const char* inputBuffer);
+The result of the function is the (void*) pointer on the LZ4HC Data Structure.
+This pointer will be needed in all other functions.
+If the pointer returned is NULL, then the allocation has failed, and compression must be aborted.
+The only parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
+The input buffer must be already allocated, and size at least 192KB.
+'inputBuffer' will also be the 'const char* source' of the first block.
+
+All blocks are expected to lay next to each other within the input buffer, starting from 'inputBuffer'.
+To compress each block, use either LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue().
+Their behavior are identical to LZ4_compressHC() or LZ4_compressHC_limitedOutput(),
+but require the LZ4HC Data Structure as their first argument, and check that each block starts right after the previous one.
+If next block does not begin immediately after the previous one, the compression will fail (return 0).
+
+When it's no longer possible to lay the next block after the previous one (not enough space left into input buffer), a call to :
+char* LZ4_slideInputBufferHC(void* LZ4HC_Data);
+must be performed. It will typically copy the latest 64KB of input at the beginning of input buffer.
+Note that, for this function to work properly, minimum size of an input buffer must be 192KB.
+==> The memory position where the next input data block must start is provided as the result of the function.
+
+Compression can then resume, using LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue(), as usual.
+
+When compression is completed, a call to LZ4_freeHC() will release the memory used by the LZ4HC Data Structure.
+*/
+
+int LZ4_sizeofStreamStateHC(void);
+int LZ4_resetStreamStateHC(void* state, const char* inputBuffer);
+
+/*
+These functions achieve the same result as :
+void* LZ4_createHC (const char* inputBuffer);
+
+They are provided here to allow the user program to allocate memory using its own routines.
+
+To know how much space must be allocated, use LZ4_sizeofStreamStateHC();
+Note also that space must be aligned for pointers (32 or 64 bits).
+
+Once space is allocated, you must initialize it using : LZ4_resetStreamStateHC(void* state, const char* inputBuffer);
+void* state is a pointer to the space allocated.
+It must be aligned for pointers (32 or 64 bits), and be large enough.
+The parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
+The input buffer must be already allocated, and size at least 192KB.
+'inputBuffer' will also be the 'const char* source' of the first block.
+
+The same space can be re-used multiple times, just by initializing it each time with LZ4_resetStreamState().
+return value of LZ4_resetStreamStateHC() must be 0 is OK.
+Any other value means there was an error (typically, state is not aligned for pointers (32 or 64 bits)).
+*/
+
+
+#if defined (__cplusplus)
+}
+#endif

node/Packet.hpp

@@ -993,7 +993,7 @@ public:
 		if ((compressed())&&(size() >= ZT_PROTO_MIN_PACKET_LENGTH)) {
 			if (size() > ZT_PACKET_IDX_PAYLOAD) {
 				unsigned int compLen = size() - ZT_PACKET_IDX_PAYLOAD;
-				int ucl = LZ4_uncompress_unknownOutputSize((const char *)field(ZT_PACKET_IDX_PAYLOAD,compLen),(char *)buf,compLen,sizeof(buf));
+				int ucl = LZ4_decompress_safe((const char *)field(ZT_PACKET_IDX_PAYLOAD,compLen),(char *)buf,compLen,sizeof(buf));
 				if ((ucl > 0)&&(ucl <= (int)(capacity() - ZT_PACKET_IDX_PAYLOAD))) {
 					setSize((unsigned int)ucl + ZT_PACKET_IDX_PAYLOAD);
 					memcpy(field(ZT_PACKET_IDX_PAYLOAD,(unsigned int)ucl),buf,ucl);