Upgrade LZ4, remove extraneous files, put tap-mac into ext/ to declutter root.

ext/lz4/bench.c

@@ -1,477 +0,0 @@
-/*
-    bench.c - Demo program to benchmark open-source compression algorithm
-    Copyright (C) Yann Collet 2012
-	GPL v2 License
-
-    This program is free software; you can redistribute it and/or modify
-    it under the terms of the GNU General Public License as published by
-    the Free Software Foundation; either version 2 of the License, or
-    (at your option) any later version.
-
-    This program is distributed in the hope that it will be useful,
-    but WITHOUT ANY WARRANTY; without even the implied warranty of
-    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License along
-    with this program; if not, write to the Free Software Foundation, Inc.,
-    51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
-
-	You can contact the author at :
-	- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
-	- LZ4 source repository : http://code.google.com/p/lz4/
-*/
-
-//**************************************
-// Compiler Options
-//**************************************
-// Disable some Visual warning messages
-#define _CRT_SECURE_NO_WARNINGS
-#define _CRT_SECURE_NO_DEPRECATE     // VS2005
-
-// Unix Large Files support (>4GB)
-#if (defined(__sun__) && (!defined(__LP64__)))   // Sun Solaris 32-bits requires specific definitions
-#  define _LARGEFILE_SOURCE 
-#  define FILE_OFFSET_BITS=64
-#elif ! defined(__LP64__)                        // No point defining Large file for 64 bit
-#  define _LARGEFILE64_SOURCE
-#endif
-
-// S_ISREG & gettimeofday() are not supported by MSVC
-#if defined(_MSC_VER)
-#  define S_ISREG(x) (((x) & S_IFMT) == S_IFREG)
-#  define BMK_LEGACY_TIMER 1
-#endif
-
-// GCC does not support _rotl outside of Windows
-#if !defined(_WIN32)
-#  define _rotl(x,r) ((x << r) | (x >> (32 - r)))
-#endif
-
-
-//**************************************
-// Includes
-//**************************************
-#include <stdlib.h>      // malloc
-#include <stdio.h>       // fprintf, fopen, ftello64
-#include <sys/types.h>   // stat64
-#include <sys/stat.h>    // stat64
-
-// Use ftime() if gettimeofday() is not available on your target
-#if defined(BMK_LEGACY_TIMER)   
-#  include <sys/timeb.h>   // timeb, ftime
-#else
-#  include <sys/time.h>    // gettimeofday
-#endif
-
-#include "lz4.h"
-#define COMPRESSOR0 LZ4_compress
-#include "lz4hc.h"
-#define COMPRESSOR1 LZ4_compressHC
-#define DEFAULTCOMPRESSOR LZ4_compress
-
-
-
-//**************************************
-// Basic Types
-//**************************************
-#if defined(_MSC_VER)    // Visual Studio does not support 'stdint' natively
-#define BYTE	unsigned __int8
-#define U16		unsigned __int16
-#define U32		unsigned __int32
-#define S32		__int32
-#define U64		unsigned __int64
-#else
-#include <stdint.h>
-#define BYTE	uint8_t
-#define U16		uint16_t
-#define U32		uint32_t
-#define S32		int32_t
-#define U64		uint64_t
-#endif
-
-
-//**************************************
-// Constants
-//**************************************
-#define NBLOOPS		3
-#define TIMELOOP	2000
-
-#define KNUTH		2654435761U
-#define MAX_MEM		(1984<<20)
-#define DEFAULT_CHUNKSIZE   (8<<20)
-
-
-//**************************************
-// Local structures
-//**************************************
-struct chunkParameters
-{
-	U32   id;
-	char* inputBuffer;
-	char* outputBuffer;
-	int   inputSize;
-	int   outputSize;
-};
-
-struct compressionParameters
-{
-	int (*compressionFunction)(const char*, char*, int);
-	int (*decompressionFunction)(const char*, char*, int);
-};
-
-
-//**************************************
-// MACRO
-//**************************************
-#define DISPLAY(...) fprintf(stderr, __VA_ARGS__)
-
-
-
-//**************************************
-// Benchmark Parameters
-//**************************************
-static int chunkSize = DEFAULT_CHUNKSIZE;
-static int nbIterations = NBLOOPS;
-static int BMK_pause = 0;
-
-void BMK_SetBlocksize(int bsize)
-{
-	chunkSize = bsize;
-	DISPLAY("-Using Block Size of %i KB-", chunkSize>>10);
-}
-
-void BMK_SetNbIterations(int nbLoops)
-{
-	nbIterations = nbLoops;
-	DISPLAY("- %i iterations-", nbIterations);
-}
-
-void BMK_SetPause()
-{
-    BMK_pause = 1;
-}
-
-//*********************************************************
-//  Private functions
-//*********************************************************
-
-#if defined(BMK_LEGACY_TIMER)
-
-static int BMK_GetMilliStart()
-{
-  // Based on Legacy ftime()
-  // Rolls over every ~ 12.1 days (0x100000/24/60/60)
-  // Use GetMilliSpan to correct for rollover
-  struct timeb tb;
-  int nCount;
-  ftime( &tb );
-  nCount = (int) (tb.millitm + (tb.time & 0xfffff) * 1000);
-  return nCount;
-}
-
-#else
-
-static int BMK_GetMilliStart()
-{
-  // Based on newer gettimeofday()
-  // Use GetMilliSpan to correct for rollover
-  struct timeval tv;
-  int nCount;
-  gettimeofday(&tv, NULL);
-  nCount = (int) (tv.tv_usec/1000 + (tv.tv_sec & 0xfffff) * 1000);
-  return nCount;
-}
-
-#endif
-
-
-static int BMK_GetMilliSpan( int nTimeStart )
-{
-  int nSpan = BMK_GetMilliStart() - nTimeStart;
-  if ( nSpan < 0 )
-    nSpan += 0x100000 * 1000;
-  return nSpan;
-}
-
-
-static U32 BMK_checksum_MMH3A (char* buff, U32 length)
-{
-  const BYTE* data = (const BYTE*)buff;
-  const int nblocks = length >> 2;
-
-  U32 h1 = KNUTH;
-  U32 c1 = 0xcc9e2d51;
-  U32 c2 = 0x1b873593;
-
-  const U32* blocks = (const U32*)(data + nblocks*4);
-  int i;
-
-  for(i = -nblocks; i; i++)
-  {
-    U32 k1 = blocks[i];
-
-    k1 *= c1;
-    k1 = _rotl(k1,15);
-    k1 *= c2;
-
-    h1 ^= k1;
-    h1 = _rotl(h1,13);
-    h1 = h1*5+0xe6546b64;
-  }
-
-  {
-	  const BYTE* tail = (const BYTE*)(data + nblocks*4);
-	  U32 k1 = 0;
-
-	  switch(length & 3)
-	  {
-	  case 3: k1 ^= tail[2] << 16;
-	  case 2: k1 ^= tail[1] << 8;
-	  case 1: k1 ^= tail[0];
-			  k1 *= c1; k1 = _rotl(k1,15); k1 *= c2; h1 ^= k1;
-	  };
-  }
-
-  h1 ^= length;
-  h1 ^= h1 >> 16;
-  h1 *= 0x85ebca6b;
-  h1 ^= h1 >> 13;
-  h1 *= 0xc2b2ae35;
-  h1 ^= h1 >> 16;
-
-  return h1;
-}
-
-
-static size_t BMK_findMaxMem(U64 requiredMem)
-{
-	size_t step = (64U<<20);   // 64 MB
-	BYTE* testmem=NULL;
-
-	requiredMem = (((requiredMem >> 25) + 1) << 26);
-	if (requiredMem > MAX_MEM) requiredMem = MAX_MEM;
-
-	requiredMem += 2*step;
-	while (!testmem)
-	{
-		requiredMem -= step;
-		testmem = malloc ((size_t)requiredMem);
-	}
-
-	free (testmem);
-	return (size_t) (requiredMem - step);
-}
-
-
-static U64 BMK_GetFileSize(char* infilename)
-{
-	int r;
-#if defined(_MSC_VER)
-	struct _stat64 statbuf;
-	r = _stat64(infilename, &statbuf);
-#else
-	struct stat statbuf;
-	r = stat(infilename, &statbuf);
-#endif
-	if (r || !S_ISREG(statbuf.st_mode)) return 0;   // No good...
-	return (U64)statbuf.st_size;
-}
-
-
-//*********************************************************
-//  Public function
-//*********************************************************
-
-int BMK_benchFile(char** fileNamesTable, int nbFiles, int cLevel)
-{
-  int fileIdx=0;
-  FILE* fileIn;
-  char* infilename;
-  U64 largefilesize;
-  size_t benchedsize;
-  int nbChunks;
-  int maxCChunkSize;
-  size_t readSize;
-  char* in_buff;
-  char* out_buff; int out_buff_size;
-  struct chunkParameters* chunkP;
-  U32 crcc, crcd=0;
-  struct compressionParameters compP;
-
-  U64 totals = 0;
-  U64 totalz = 0;
-  double totalc = 0.;
-  double totald = 0.;
-
-
-  // Init
-  switch (cLevel)
-  {
-#ifdef COMPRESSOR0
-  case 0 : compP.compressionFunction = COMPRESSOR0; break;
-#endif
-#ifdef COMPRESSOR1
-  case 1 : compP.compressionFunction = COMPRESSOR1; break;
-#endif
-  default : compP.compressionFunction = DEFAULTCOMPRESSOR;
-  }
-  compP.decompressionFunction = LZ4_uncompress;
-
-  // Loop for each file
-  while (fileIdx<nbFiles)
-  {
-	  // Check file existence
-	  infilename = fileNamesTable[fileIdx++];
-	  fileIn = fopen( infilename, "rb" );
-	  if (fileIn==NULL)
-	  {
-		DISPLAY( "Pb opening %s\n", infilename);
-		return 11;
-	  }
-
-	  // Memory allocation & restrictions
-	  largefilesize = BMK_GetFileSize(infilename);
-	  benchedsize = (size_t) BMK_findMaxMem(largefilesize) / 2;
-	  if ((U64)benchedsize > largefilesize) benchedsize = (size_t)largefilesize;
-	  if (benchedsize < largefilesize)
-	  {
-		  DISPLAY("Not enough memory for '%s' full size; testing %i MB only...\n", infilename, (int)(benchedsize>>20));
-	  }
-
-	  // Alloc
-	  chunkP = (struct chunkParameters*) malloc(((benchedsize / chunkSize)+1) * sizeof(struct chunkParameters));
-	  in_buff = malloc((size_t )benchedsize);
-	  nbChunks = (int) (benchedsize / chunkSize) + 1;
-	  maxCChunkSize = LZ4_compressBound(chunkSize);
-	  out_buff_size = nbChunks * maxCChunkSize;
-	  out_buff = malloc((size_t )out_buff_size);
-
-
-	  if(!in_buff || !out_buff)
-	  {
-		DISPLAY("\nError: not enough memory!\n");
-		free(in_buff);
-		free(out_buff);
-		fclose(fileIn);
-		return 12;
-	  }
-
-	  // Init chunks data
-	  {
-		  int i;
-		  size_t remaining = benchedsize;
-		  char* in = in_buff;
-		  char* out = out_buff;
-		  for (i=0; i<nbChunks; i++)
-		  {
-			  chunkP[i].id = i;
-			  chunkP[i].inputBuffer = in; in += chunkSize;
-			  if ((int)remaining > chunkSize) { chunkP[i].inputSize = chunkSize; remaining -= chunkSize; } else { chunkP[i].inputSize = (int)remaining; remaining = 0; }
-			  chunkP[i].outputBuffer = out; out += maxCChunkSize;
-			  chunkP[i].outputSize = 0;
-		  }
-	  }
-
-	  // Fill input buffer
-	  DISPLAY("Loading %s...       \r", infilename);
-	  readSize = fread(in_buff, 1, benchedsize, fileIn);
-	  fclose(fileIn);
-
-	  if(readSize != benchedsize)
-	  {
-		DISPLAY("\nError: problem reading file '%s' !!    \n", infilename);
-		free(in_buff);
-		free(out_buff);
-		return 13;
-	  }
-
-	  // Calculating input Checksum
-	  crcc = BMK_checksum_MMH3A(in_buff, (unsigned int)benchedsize);
-
-
-	  // Bench
-	  {
-		int loopNb, nb_loops, chunkNb;
-		size_t cSize=0;
-	    int milliTime;
-		double fastestC = 100000000., fastestD = 100000000.;
-		double ratio=0.;
-
-		DISPLAY("\r%79s\r", "");
-		for (loopNb = 1; loopNb <= nbIterations; loopNb++)
-		{
-		  // Compression
-		  DISPLAY("%1i-%-14.14s : %9i ->\r", loopNb, infilename, (int)benchedsize);
-		  { size_t i; for (i=0; i<benchedsize; i++) out_buff[i]=(char)i; }     // warmimg up memory
-
-		  nb_loops = 0;
-		  milliTime = BMK_GetMilliStart();
-		  while(BMK_GetMilliStart() == milliTime);
-		  milliTime = BMK_GetMilliStart();
-		  while(BMK_GetMilliSpan(milliTime) < TIMELOOP)
-		  {
-			for (chunkNb=0; chunkNb<nbChunks; chunkNb++)
-				chunkP[chunkNb].outputSize = compP.compressionFunction(chunkP[chunkNb].inputBuffer, chunkP[chunkNb].outputBuffer, chunkP[chunkNb].inputSize);
-			nb_loops++;
-		  }
-		  milliTime = BMK_GetMilliSpan(milliTime);
-
-		  if ((double)milliTime < fastestC*nb_loops) fastestC = (double)milliTime/nb_loops;
-		  cSize=0; for (chunkNb=0; chunkNb<nbChunks; chunkNb++) cSize += chunkP[chunkNb].outputSize;
-		  ratio = (double)cSize/(double)benchedsize*100.;
-
-		  DISPLAY("%1i-%-14.14s : %9i -> %9i (%5.2f%%),%7.1f MB/s\r", loopNb, infilename, (int)benchedsize, (int)cSize, ratio, (double)benchedsize / fastestC / 1000.);
-
-		  // Decompression
-		  { size_t i; for (i=0; i<benchedsize; i++) in_buff[i]=0; }     // zeroing area, for CRC checking
-
-		  nb_loops = 0;
-		  milliTime = BMK_GetMilliStart();
-		  while(BMK_GetMilliStart() == milliTime);
-		  milliTime = BMK_GetMilliStart();
-		  while(BMK_GetMilliSpan(milliTime) < TIMELOOP)
-		  {
-			for (chunkNb=0; chunkNb<nbChunks; chunkNb++)
-				chunkP[chunkNb].outputSize = LZ4_uncompress(chunkP[chunkNb].outputBuffer, chunkP[chunkNb].inputBuffer, chunkP[chunkNb].inputSize);
-				//chunkP[chunkNb].inputSize = LZ4_uncompress_unknownOutputSize(chunkP[chunkNb].outputBuffer, chunkP[chunkNb].inputBuffer, chunkP[chunkNb].outputSize, chunkSize);
-			nb_loops++;
-		  }
-		  milliTime = BMK_GetMilliSpan(milliTime);
-
-		  if ((double)milliTime < fastestD*nb_loops) fastestD = (double)milliTime/nb_loops;
-		  DISPLAY("%1i-%-14.14s : %9i -> %9i (%5.2f%%),%7.1f MB/s ,%7.1f MB/s\r", loopNb, infilename, (int)benchedsize, (int)cSize, ratio, (double)benchedsize / fastestC / 1000., (double)benchedsize / fastestD / 1000.);
-
-		  // CRC Checking
-		  crcd = BMK_checksum_MMH3A(in_buff, (unsigned int)benchedsize);
-		  if (crcc!=crcd) { DISPLAY("\n!!! WARNING !!! %14s : Invalid Checksum : %x != %x\n", infilename, (unsigned)crcc, (unsigned)crcd); break; }
-		}
-
-	    if (crcc==crcd)
-		{
-			if (ratio<100.)
-				DISPLAY("%-16.16s : %9i -> %9i (%5.2f%%),%7.1f MB/s ,%7.1f MB/s\n", infilename, (int)benchedsize, (int)cSize, ratio, (double)benchedsize / fastestC / 1000., (double)benchedsize / fastestD / 1000.);
-			else
-				DISPLAY("%-16.16s : %9i -> %9i (%5.1f%%),%7.1f MB/s ,%7.1f MB/s \n", infilename, (int)benchedsize, (int)cSize, ratio, (double)benchedsize / fastestC / 1000., (double)benchedsize / fastestD / 1000.);
-		}
-		totals += benchedsize;
-		totalz += cSize;
-		totalc += fastestC;
-		totald += fastestD;
-	  }
-
-	  free(in_buff);
-	  free(out_buff);
-	  free(chunkP);
-  }
-
-  if (nbFiles > 1)
-		printf("%-16.16s :%10llu ->%10llu (%5.2f%%), %6.1f MB/s , %6.1f MB/s\n", "  TOTAL", (long long unsigned int)totals, (long long unsigned int)totalz, (double)totalz/(double)totals*100., (double)totals/totalc/1000., (double)totals/totald/1000.);
-
-  if (BMK_pause) { printf("press enter...\n"); getchar(); }
-
-  return 0;
-}
-
-
-

ext/lz4/bench.h

@@ -1,41 +0,0 @@
-/*
-    bench.h - Demo program to benchmark open-source compression algorithm
-    Copyright (C) Yann Collet 2012
-
-    This program is free software; you can redistribute it and/or modify
-    it under the terms of the GNU General Public License as published by
-    the Free Software Foundation; either version 2 of the License, or
-    (at your option) any later version.
-
-    This program is distributed in the hope that it will be useful,
-    but WITHOUT ANY WARRANTY; without even the implied warranty of
-    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License along
-    with this program; if not, write to the Free Software Foundation, Inc.,
-    51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
-
-	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 BMK_benchFile(char** fileNamesTable, int nbFiles, int cLevel);
-
-// Parameters
-void BMK_SetBlocksize(int bsize);
-void BMK_SetNbIterations(int nbLoops);
-void BMK_SetPause();
-
-
-
-#if defined (__cplusplus)
-}
-#endif

ext/lz4/fuzzer.c

@@ -1,227 +0,0 @@
-/*
-    fuzzer.c - Fuzzer test tool for LZ4
-    Copyright (C) Andrew Mahone - Yann Collet 2012
-	Original code by Andrew Mahone / Modified by Yann Collet
-	GPL v2 License
-
-    This program is free software; you can redistribute it and/or modify
-    it under the terms of the GNU General Public License as published by
-    the Free Software Foundation; either version 2 of the License, or
-    (at your option) any later version.
-
-    This program is distributed in the hope that it will be useful,
-    but WITHOUT ANY WARRANTY; without even the implied warranty of
-    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License along
-    with this program; if not, write to the Free Software Foundation, Inc.,
-    51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
-
-	You can contact the author at :
-	- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
-	- LZ4 source repository : http://code.google.com/p/lz4/
-*/
-
-//**************************************
-// Remove Visual warning messages
-//**************************************
-#define _CRT_SECURE_NO_WARNINGS  // fgets
-
-
-//**************************************
-// Includes
-//**************************************
-#include <stdlib.h>
-#include <stdio.h>      // fgets, sscanf
-#include <sys/timeb.h>  // timeb
-#include "lz4.h"
-
-
-//**************************************
-// Constants
-//**************************************
-#define NB_ATTEMPTS (1<<18)
-#define LEN ((1<<15))
-#define SEQ_POW 2
-#define NUM_SEQ (1 << SEQ_POW)
-#define SEQ_MSK ((NUM_SEQ) - 1)
-#define MOD_SEQ(x) ((((x) >> 8) & 255) == 0)
-#define NEW_SEQ(x) ((((x) >> 10) %10) == 0)
-#define PAGE_SIZE 4096
-#define ROUND_PAGE(x) (((x) + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1))
-#define PRIME1   2654435761U
-#define PRIME2   2246822519U
-#define PRIME3   3266489917U
-
-
-
-//*********************************************************
-//  Functions
-//*********************************************************
-static int FUZ_GetMilliStart()
-{
-  struct timeb tb;
-  int nCount;
-  ftime( &tb );
-  nCount = (int) (tb.millitm + (tb.time & 0xfffff) * 1000);
-  return nCount;
-}
-
-static int FUZ_GetMilliSpan( int nTimeStart )
-{
-  int nSpan = FUZ_GetMilliStart() - nTimeStart;
-  if ( nSpan < 0 )
-    nSpan += 0x100000 * 1000;
-  return nSpan;
-}
-
-
-unsigned int FUZ_rand(unsigned int* src)
-{
-	*src =  ((*src) * PRIME1) + PRIME2;
-	return *src;
-}
-
-
-int test_canary(unsigned char *buf) {
-        int i;
-        for (i = 0; i < 2048; i++)
-                if (buf[i] != buf[i + 2048])
-                        return 0;
-        return 1;
-}
-
-int FUZ_SecurityTest()
-{
-  char* output;
-  char* input;
-  int i, r;
-
-  printf("Starting security tests...");
-  input = (char*) malloc (20<<20);
-  output = (char*) malloc (20<<20);
-  input[0] = 0x0F;
-  input[1] = 0x00;
-  input[2] = 0x00;
-  for(i = 3; i < 16840000; i++)
-    input[i] = 0xff;
-  r = LZ4_uncompress(input, output, 20<<20);
-
-  free(input);
-  free(output);
-  printf(" Completed (r=%i)\n",r);
-  return 0;
-}
-
-
-//int main(int argc, char *argv[]) {
-int main() {
-        unsigned long long bytes = 0;
-        unsigned long long cbytes = 0;
-        unsigned char buf[LEN];
-        unsigned char testOut[LEN+1];
-#       define FUZ_max   LZ4_COMPRESSBOUND(LEN)
-#       define FUZ_avail ROUND_PAGE(FUZ_max)
-        const int off_full = FUZ_avail - FUZ_max;
-        unsigned char cbuf[FUZ_avail + PAGE_SIZE];
-		unsigned int seed, cur_seq=PRIME3, seeds[NUM_SEQ], timestamp=FUZ_GetMilliStart();
-        int i, j, k, ret, len;
-		char userInput[30] = {0};
-
-		printf("starting LZ4 fuzzer\n");
-		printf("Select an Initialisation number (default : random) : ");
-		fflush(stdout);
-		if ( fgets(userInput, sizeof userInput, stdin) )
-		{
-			if ( sscanf(userInput, "%d", &seed) == 1 ) {}
-			else seed = FUZ_GetMilliSpan(timestamp);
-		}
-		printf("Seed = %u\n", seed);
-
-		FUZ_SecurityTest();
-
-		for (i = 0; i < 2048; i++)
-                cbuf[FUZ_avail + i] = cbuf[FUZ_avail + 2048 + i] = FUZ_rand(&seed) >> 16;
-
-        for (i = 0; i < NB_ATTEMPTS; i++) 
-        {
-			printf("\r%7i /%7i\r", i, NB_ATTEMPTS);
-			
-			FUZ_rand(&seed);
-            for (j = 0; j < NUM_SEQ; j++) {
-                    seeds[j] = FUZ_rand(&seed) << 8;
-                    seeds[j] ^= (FUZ_rand(&seed) >> 8) & 65535;
-            }
-            for (j = 0; j < LEN; j++) {
-                    k = FUZ_rand(&seed);
-                    if (j == 0 || NEW_SEQ(k))
-                            cur_seq = seeds[(FUZ_rand(&seed) >> 16) & SEQ_MSK];
-                    if (MOD_SEQ(k)) {
-                            k = (FUZ_rand(&seed) >> 16) & SEQ_MSK;
-                            seeds[k] = FUZ_rand(&seed) << 8;
-                            seeds[k] ^= (FUZ_rand(&seed) >> 8) & 65535;
-                    }
-                    buf[j] = FUZ_rand(&cur_seq) >> 16;
-            }
-
-			// Test compression
-            ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[off_full], LEN, FUZ_max);
-			if (ret == 0) { printf("compression failed despite sufficient space: seed %u, len %d\n", seed, LEN); goto _output_error; }
-            len = ret;
-
-			// Test decoding with output size being exactly what's necessary => must work
-			ret = LZ4_uncompress((char*)&cbuf[off_full], (char*)testOut, LEN);
-			if (ret<0) { printf("decompression failed despite correct space: seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test decoding with one byte missing => must fail
-			ret = LZ4_uncompress((char*)&cbuf[off_full], (char*)testOut, LEN-1);
-			if (ret>=0) { printf("decompression should have failed, due to Output Size being too small : seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test decoding with one byte too much => must fail
-			ret = LZ4_uncompress((char*)&cbuf[off_full], (char*)testOut, LEN+1);
-			if (ret>=0) { printf("decompression should have failed, due to Output Size being too large : seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test decoding with enough output size => must work
-			ret = LZ4_uncompress_unknownOutputSize((char*)&cbuf[off_full], (char*)testOut, len, LEN+1);
-			if (ret<0) { printf("decompression failed despite sufficient space: seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test decoding with output size being exactly what's necessary => must work
-			ret = LZ4_uncompress_unknownOutputSize((char*)&cbuf[off_full], (char*)testOut, len, LEN);
-			if (ret<0) { printf("decompression failed despite sufficient space: seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test decoding with output size being one byte too short => must fail
-			ret = LZ4_uncompress_unknownOutputSize((char*)&cbuf[off_full], (char*)testOut, len, LEN-1);
-			if (ret>=0) { printf("decompression should have failed, due to Output Size being too small : seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test decoding with input size being one byte too short => must fail
-			ret = LZ4_uncompress_unknownOutputSize((char*)&cbuf[off_full], (char*)testOut, len-1, LEN);
-			if (ret>=0) { printf("decompression should have failed, due to input size being too small : seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test decoding with input size being one byte too large => must fail
-			ret = LZ4_uncompress_unknownOutputSize((char*)&cbuf[off_full], (char*)testOut, len+1, LEN);
-			if (ret>=0) { printf("decompression should have failed, due to input size being too large : seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test compression with output size being exactly what's necessary (should work)
-            ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-len], LEN, len);
-            if (!test_canary(&cbuf[FUZ_avail])) { printf("compression overran output buffer: seed %u, len %d, olen %d\n", seed, LEN, len); goto _output_error; }
-            if (ret == 0) { printf("compression failed despite sufficient space: seed %u, len %d\n", seed, LEN); goto _output_error; }
-
-			// Test compression with just one missing byte into output buffer => must fail
-            ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-(len-1)], LEN, len-1);
-            if (ret) { printf("compression overran output buffer: seed %u, len %d, olen %d => ret %d", seed, LEN, len-1, ret); goto _output_error; }
-            if (!test_canary(&cbuf[FUZ_avail])) { printf("compression overran output buffer: seed %u, len %d, olen %d", seed, LEN, len-1); goto _output_error; }
-
-			bytes += LEN;
-            cbytes += len;
-        }
-
-		printf("all tests completed successfully \n");
-        printf("compression ratio: %0.3f%%\n", (double)cbytes/bytes*100);
-		getchar();
-        return 0;
-
-_output_error:
-		getchar();
-		return 1;
-}

ext/lz4/lz4.h

@@ -1,7 +1,7 @@
 /*
    LZ4 - Fast LZ compression algorithm
    Header File
-   Copyright (C) 2011-2012, Yann Collet.
+   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
@@ -42,7 +42,7 @@ extern "C" {
 // Compiler Options
 //**************************************
 #if defined(_MSC_VER) && !defined(__cplusplus)   // Visual Studio
-#  define inline __inline           // Visual is not C99, but supports some kind of inline
+#  define inline __inline           // Visual C is not C99, but supports some kind of inline
 #endif
 
 
@@ -50,79 +50,156 @@ extern "C" {
 // Simple Functions
 //****************************
 
-int LZ4_compress   (const char* source, char* dest, int isize);
-int LZ4_uncompress (const char* source, char* dest, int osize);
+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 'isize' bytes from 'source' into 'dest'.
+    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()
-
-    isize  : is the input size. Max supported value is ~1.9GB
+    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_uncompress() :
-    osize  : is the output size, therefore the original size
-    return : the number of bytes read in the source buffer
-             If the source stream is malformed, the function will stop decoding and return a negative result, indicating the byte position of the faulty instruction
-             This function never writes outside of provided buffers, and never modifies input buffer.
-    note : destination buffer must be already allocated.
-           its size must be a minimum of 'osize' bytes.
+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
 //****************************
-
-static inline int LZ4_compressBound(int isize)   { return ((isize) + ((isize)/255) + 16); }
-#define           LZ4_COMPRESSBOUND(    isize)            ((isize) + ((isize)/255) + 16)
+#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,
-	but macro is also provided when results need to be evaluated at compile time (such as table size allocation).
+    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 ~1.9GB
+    isize  : is the input size. Max supported value is LZ4_MAX_INPUT_SIZE
     return : maximum output size in a "worst case" scenario
-    note : this function is limited by "int" range (2^31-1)
+             or 0, if input size is too large ( > LZ4_MAX_INPUT_SIZE)
 */
 
 
-int LZ4_compress_limitedOutput   (const char* source, char* dest, int isize, int maxOutputSize);
+int LZ4_compress_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
 
 /*
 LZ4_compress_limitedOutput() :
-    Compress 'isize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
+    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.
 
-    isize  : is the input size. Max supported value is ~1.9GB
+    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_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize);
+int LZ4_decompress_fast (const char* source, char* dest, int outputSize);
 
 /*
-LZ4_uncompress_unknownOutputSize() :
-    isize  : is the input size, therefore the compressed size
-    maxOutputSize : is the size of the destination buffer (which must be already allocated)
+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)
-             If the source stream is malformed, the function will stop decoding and return a negative result, indicating the byte position of the faulty instruction
-             This function never writes beyond dest + maxOutputSize, and is therefore protected against malicious data packets
-    note   : Destination buffer must be already allocated.
-             This version is slightly slower than LZ4_uncompress()
+       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

ext/lz4/lz4_format_description.txt

@@ -1,121 +0,0 @@
-LZ4 Format Description
-Last revised: 2012-02-27
-Author : Y. Collet
-
-
-
-This small specification intents to provide enough information
-to anyone willing to produce LZ4-compatible compressed streams
-using any programming language.
-
-LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
-The most important design principle behind LZ4 is simplicity.
-It helps to create an easy to read and maintain source code.
-It also helps later on for optimisations, compactness, and speed.
-There is no entropy encoder backend nor framing layer.
-The latter is assumed to be handled by other parts of the system.
-
-This document only describes the format,
-not how the LZ4 compressor nor decompressor actually work.
-The correctness of the decompressor should not depend
-on implementation details of the compressor, and vice versa.
-
-
-
--- Compressed stream format --
-
-An LZ4 compressed stream is composed of sequences.
-Schematically, a sequence is a suite of literals, followed by a match copy.
-
-Each sequence starts with a token.
-The token is a one byte value, separated into two 4-bits fields.
-Therefore each field ranges from 0 to 15.
-
-
-The first field uses the 4 high-bits of the token.
-It provides the length of literals to follow.
-(Note : a literal is a not-compressed byte).
-If the field value is 0, then there is no literal.
-If it is 15, then we need to add some more bytes to indicate the full length.
-Each additionnal byte then represent a value from 0 to 255,
-which is added to the previous value to produce a total length.
-When the byte value is 255, another byte is output.
-There can be any number of bytes following the token. There is no "size limit".
-(Sidenote this is why a not-compressible input stream is expanded by 0.4%).
-
-Example 1 : A length of 48 will be represented as :
-- 15 : value for the 4-bits High field
-- 33 : (=48-15) remaining length to reach 48
-
-Example 2 : A length of 280 will be represented as :
-- 15  : value for the 4-bits High field
-- 255 : following byte is maxed, since 280-15 >= 255
-- 10  : (=280 - 15 - 255) ) remaining length to reach 280
-
-Example 3 : A length of 15 will be represented as :
-- 15 : value for the 4-bits High field
-- 0  : (=15-15) yes, the zero must be output
-
-Following the token and optional length bytes, are the literals themselves.
-They are exactly as numerous as previously decoded (length of literals).
-It's possible that there are zero literal.
-
-
-Following the literals is the match copy operation.
-
-It starts by the offset.
-This is a 2 bytes value, in little endian format :
-the lower byte is the first one in the stream.
-
-The offset represents the position of the match to be copied from.
-1 means "current position - 1 byte".
-The maximum offset value is 65535, 65536 cannot be coded.
-Note that 0 is an invalid value, not used. 
-
-Then we need to extract the match length.
-For this, we use the second token field, the low 4-bits.
-Value, obviously, ranges from 0 to 15.
-However here, 0 means that the copy operation will be minimal.
-The minimum length of a match, called minmatch, is 4. 
-As a consequence, a 0 value means 4 bytes, and a value of 15 means 19+ bytes.
-Similar to literal length, on reaching the highest possible value (15), 
-we output additional bytes, one at a time, with values ranging from 0 to 255.
-They are added to total to provide the final match length.
-A 255 value means there is another byte to read and add.
-There is no limit to the number of optional bytes that can be output this way.
-(This points towards a maximum achievable compression ratio of ~250).
-
-With the offset and the matchlength,
-the decoder can now proceed to copy the data from the already decoded buffer.
-On decoding the matchlength, we reach the end of the compressed sequence,
-and therefore start another one.
-
-
--- Parsing restrictions --
-
-There are specific parsing rules to respect in order to remain compatible
-with assumptions made by the decoder :
-1) The last 5 bytes are always literals
-2) The last match must start at least 12 bytes before end of stream
-Consequently, a file with less than 13 bytes cannot be compressed.
-These rules are in place to ensure that the decoder
-will never read beyond the input buffer, nor write beyond the output buffer.
-
-Note that the last sequence is also incomplete,
-and stops right after literals.
-
-
--- Additional notes --
-
-There is no assumption nor limits to the way the compressor
-searches and selects matches within the source stream.
-It could be a fast scan, a multi-probe, a full search using BST,
-standard hash chains or MMC, well whatever.
-
-Advanced parsing strategies can also be implemented, such as lazy match,
-or full optimal parsing.
-
-All these trade-off offer distinctive speed/memory/compression advantages.
-Whatever the method used by the compressor, its result will be decodable
-by any LZ4 decoder if it follows the format specification described above.
-

ext/lz4/lz4demo.c

@@ -1,402 +0,0 @@
-/*
-    LZ4Demo - Demo CLI program using LZ4 compression
-    Copyright (C) Yann Collet 2011-2012
-	GPL v2 License
-
-    This program is free software; you can redistribute it and/or modify
-    it under the terms of the GNU General Public License as published by
-    the Free Software Foundation; either version 2 of the License, or
-    (at your option) any later version.
-
-    This program is distributed in the hope that it will be useful,
-    but WITHOUT ANY WARRANTY; without even the implied warranty of
-    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-    GNU General Public License for more details.
-
-    You should have received a copy of the GNU General Public License along
-    with this program; if not, write to the Free Software Foundation, Inc.,
-    51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
-
-	You can contact the author at :
-	- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
-	- LZ4 source repository : http://code.google.com/p/lz4/
-*/
-/*
-	Note : this is *only* a demo program, an example to show how LZ4 can be used.
-	It is not considered part of LZ4 compression library.
-	The license of LZ4 is BSD.
-	The license of the demo program is GPL.
-*/
-
-//**************************************
-// Compiler Options
-//**************************************
-// Disable some Visual warning messages
-#define _CRT_SECURE_NO_WARNINGS
-#define _CRT_SECURE_NO_DEPRECATE     // VS2005
-
-
-//****************************
-// Includes
-//****************************
-#include <stdio.h>		// fprintf, fopen, fread, _fileno(?)
-#include <stdlib.h>		// malloc
-#include <string.h>		// strcmp
-#include <time.h>		// clock
-#ifdef _WIN32
-#include <io.h>			// _setmode
-#include <fcntl.h>		// _O_BINARY
-#endif
-#include "lz4.h"
-#include "lz4hc.h"
-#include "bench.h"
-
-
-//**************************************
-// Compiler-specific functions
-//**************************************
-#define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
-
-#if defined(_MSC_VER)    // Visual Studio
-#define swap32 _byteswap_ulong
-#elif GCC_VERSION >= 403
-#define swap32 __builtin_bswap32
-#else
-static inline unsigned int swap32(unsigned int x) {
-			return	((x << 24) & 0xff000000 ) |
-				((x <<  8) & 0x00ff0000 ) |
-				((x >>  8) & 0x0000ff00 ) |
-				((x >> 24) & 0x000000ff );
-		 }
-#endif
-
-
-//****************************
-// Constants
-//****************************
-#define COMPRESSOR_NAME "Compression CLI using LZ4 algorithm"
-#define COMPRESSOR_VERSION ""
-#define COMPILED __DATE__
-#define AUTHOR "Yann Collet"
-#define EXTENSION ".lz4"
-#define WELCOME_MESSAGE "*** %s %s, by %s (%s) ***\n", COMPRESSOR_NAME, COMPRESSOR_VERSION, AUTHOR, COMPILED
-
-#define CHUNKSIZE (8<<20)    // 8 MB
-#define CACHELINE 64
-#define ARCHIVE_MAGICNUMBER 0x184C2102
-#define ARCHIVE_MAGICNUMBER_SIZE 4
-
-
-//**************************************
-// Architecture Macros
-//**************************************
-static const int one = 1;
-#define CPU_LITTLE_ENDIAN  (*(char*)(&one))
-#define CPU_BIG_ENDIAN     (!CPU_LITTLE_ENDIAN)
-#define LITTLE_ENDIAN32(i) if (CPU_BIG_ENDIAN) { i = swap32(i); }
-
-
-//**************************************
-// Macros
-//**************************************
-#define DISPLAY(...) fprintf(stderr, __VA_ARGS__)
-
-
-//****************************
-// Functions
-//****************************
-int usage(char* exename)
-{
-	DISPLAY( "Usage :\n");
-	DISPLAY( "      %s [arg] input output\n", exename);
-	DISPLAY( "Arguments :\n");
-	DISPLAY( " -c0: Fast compression (default) \n");
-	DISPLAY( " -c1: High compression \n");
-	DISPLAY( " -d : decompression \n");
-	DISPLAY( " -b#: Benchmark files, using # compression level\n");
-	DISPLAY( " -t : check compressed file \n");
-	DISPLAY( " -h : help (this text)\n");
-	DISPLAY( "input  : can be 'stdin' (pipe) or a filename\n");
-	DISPLAY( "output : can be 'stdout'(pipe) or a filename or 'null'\n");
-	return 0;
-}
-
-
-int badusage(char* exename)
-{
-	DISPLAY("Wrong parameters\n");
-	usage(exename);
-	return 0;
-}
-
-
-
-int get_fileHandle(char* input_filename, char* output_filename, FILE** pfinput, FILE** pfoutput)
-{
-	char stdinmark[] = "stdin";
-	char stdoutmark[] = "stdout";
-
-	if (!strcmp (input_filename, stdinmark)) {
-		DISPLAY( "Using stdin for input\n");
-		*pfinput = stdin;
-#ifdef _WIN32 // Need to set stdin/stdout to binary mode specifically for windows
-		_setmode( _fileno( stdin ), _O_BINARY );
-#endif
-	} else {
-		*pfinput = fopen( input_filename, "rb" );
-	}
-
-	if (!strcmp (output_filename, stdoutmark)) {
-		DISPLAY( "Using stdout for output\n");
-		*pfoutput = stdout;
-#ifdef _WIN32 // Need to set stdin/stdout to binary mode specifically for windows
-		_setmode( _fileno( stdout ), _O_BINARY );
-#endif
-	} else {
-		*pfoutput = fopen( output_filename, "wb" );
-	}
-
-	if ( *pfinput==0 ) { DISPLAY( "Pb opening %s\n", input_filename);  return 2; }
-	if ( *pfoutput==0) { DISPLAY( "Pb opening %s\n", output_filename); return 3; }
-
-	return 0;
-}
-
-
-
-int compress_file(char* input_filename, char* output_filename, int compressionlevel)
-{
-	int (*compressionFunction)(const char*, char*, int);
-	unsigned long long filesize = 0;
-	unsigned long long compressedfilesize = ARCHIVE_MAGICNUMBER_SIZE;
-	unsigned int u32var;
-	char* in_buff;
-	char* out_buff;
-	FILE* finput;
-	FILE* foutput;
-	int r;
-	int displayLevel = (compressionlevel>0);
-	clock_t start, end;
-	size_t sizeCheck;
-
-
-	// Init
-	switch (compressionlevel)
-	{
-	case 0 : compressionFunction = LZ4_compress; break;
-	case 1 : compressionFunction = LZ4_compressHC; break;
-	default : compressionFunction = LZ4_compress;
-	}
-	start = clock();
-	r = get_fileHandle(input_filename, output_filename, &finput, &foutput);
-	if (r) return r;
-
-	// Allocate Memory
-	in_buff = (char*)malloc(CHUNKSIZE);
-	out_buff = (char*)malloc(LZ4_compressBound(CHUNKSIZE));
-	if (!in_buff || !out_buff) { DISPLAY("Allocation error : not enough memory\n"); return 8; }
-
-	// Write Archive Header
-	u32var = ARCHIVE_MAGICNUMBER;
-	LITTLE_ENDIAN32(u32var);
-	*(unsigned int*)out_buff = u32var;
-	sizeCheck = fwrite(out_buff, 1, ARCHIVE_MAGICNUMBER_SIZE, foutput);
-	if (sizeCheck!=ARCHIVE_MAGICNUMBER_SIZE) { DISPLAY("write error\n"); return 10; }
-
-	// Main Loop
-	while (1)
-	{
-		int outSize;
-		// Read Block
-	    int inSize = (int) fread(in_buff, (size_t)1, (size_t)CHUNKSIZE, finput);
-		if( inSize<=0 ) break;
-		filesize += inSize;
-		if (displayLevel) DISPLAY("Read : %i MB  \r", (int)(filesize>>20));
-
-		// Compress Block
-		outSize = compressionFunction(in_buff, out_buff+4, inSize);
-		compressedfilesize += outSize+4;
-		if (displayLevel) DISPLAY("Read : %i MB  ==> %.2f%%\r", (int)(filesize>>20), (double)compressedfilesize/filesize*100);
-
-		// Write Block
-		LITTLE_ENDIAN32(outSize);
-		* (unsigned int*) out_buff = outSize;
-		LITTLE_ENDIAN32(outSize);
-		sizeCheck = fwrite(out_buff, 1, outSize+4, foutput);
-		if (sizeCheck!=(size_t)(outSize+4)) { DISPLAY("write error\n"); return 11; }
-	}
-
-	// Status
-	end = clock();
-	DISPLAY( "Compressed %llu bytes into %llu bytes ==> %.2f%%\n",
-		(unsigned long long) filesize, (unsigned long long) compressedfilesize, (double)compressedfilesize/filesize*100);
-	{
-		double seconds = (double)(end - start)/CLOCKS_PER_SEC;
-		DISPLAY( "Done in %.2f s ==> %.2f MB/s\n", seconds, (double)filesize / seconds / 1024 / 1024);
-	}
-
-	// Close & Free
-	free(in_buff);
-	free(out_buff);
-	fclose(finput);
-	fclose(foutput);
-
-	return 0;
-}
-
-
-int decode_file(char* input_filename, char* output_filename)
-{
-	unsigned long long filesize = 0;
-	char* in_buff;
-	char* out_buff;
-	size_t uselessRet;
-	int sinkint;
-	unsigned int chunkSize;
-	FILE* finput;
-	FILE* foutput;
-	clock_t start, end;
-	int r;
-	size_t sizeCheck;
-
-
-	// Init
-	start = clock();
-	r = get_fileHandle(input_filename, output_filename, &finput, &foutput);
-	if (r) return r;
-
-	// Allocate Memory
-	in_buff = (char*)malloc(LZ4_compressBound(CHUNKSIZE));
-	out_buff = (char*)malloc(CHUNKSIZE);
-	if (!in_buff || !out_buff) { DISPLAY("Allocation error : not enough memory\n"); return 7; }
-
-	// Check Archive Header
-	chunkSize = 0;
-	uselessRet = fread(&chunkSize, 1, ARCHIVE_MAGICNUMBER_SIZE, finput);
-	LITTLE_ENDIAN32(chunkSize);
-	if (chunkSize != ARCHIVE_MAGICNUMBER) { DISPLAY("Unrecognized header : file cannot be decoded\n"); return 6; }
-
-	// Main Loop
-	while (1)
-	{
-		// Block Size
-		uselessRet = fread(&chunkSize, 1, 4, finput);
-		if( uselessRet==0 ) break;   // Nothing to read : file read is completed
-		LITTLE_ENDIAN32(chunkSize);
-		if (chunkSize == ARCHIVE_MAGICNUMBER) 
-			continue;   // appended compressed stream
-		
-		// Read Block
-	    uselessRet = fread(in_buff, 1, chunkSize, finput);
-
-		// Decode Block
-		sinkint = LZ4_uncompress_unknownOutputSize(in_buff, out_buff, chunkSize, CHUNKSIZE);
-		if (sinkint < 0) { DISPLAY("Decoding Failed ! Corrupted input !\n"); return 9; }
-		filesize += sinkint;
-
-		// Write Block
-		sizeCheck = fwrite(out_buff, 1, sinkint, foutput);
-		if (sizeCheck != (size_t)sinkint) { DISPLAY("write error\n"); return 12; }
-	}
-
-	// Status
-	end = clock();
-	DISPLAY( "Successfully decoded %llu bytes \n", (unsigned long long)filesize);
-	{
-		double seconds = (double)(end - start)/CLOCKS_PER_SEC;
-		DISPLAY( "Done in %.2f s ==> %.2f MB/s\n", seconds, (double)filesize / seconds / 1024 / 1024);
-	}
-
-	// Close & Free
-	free(in_buff);
-	free(out_buff);
-	fclose(finput);
-	fclose(foutput);
-
-	return 0;
-}
-
-
-int main(int argc, char** argv)
-{
-  int i,
-	  cLevel=0,
-	  decode=0,
-	  bench=0,
-	  filenamesStart=2;
-  char* exename=argv[0];
-  char* input_filename=0;
-  char* output_filename=0;
-#ifdef _WIN32
-  char nulmark[] = "nul";
-#else
-  char nulmark[] = "/dev/null";
-#endif
-  char nullinput[] = "null";
-
-  // Welcome message
-  DISPLAY( WELCOME_MESSAGE);
-
-  if (argc<2) { badusage(exename); return 1; }
-
-  for(i=1; i<argc; i++)
-  {
-    char* argument = argv[i];
-
-    if(!argument) continue;   // Protection if argument empty
-
-	// Select command
-	if (argument[0]=='-')
-	{
-		argument ++;
-
-		// Display help on usage
-		if ( argument[0] =='h' ) { usage(exename); return 0; }
-
-		// Compression (default)
-		if ( argument[0] =='c' ) { if (argument[1] >='0') cLevel=argument[1] - '0'; continue; }
-
-		// Decoding
-		if ( argument[0] =='d' ) { decode=1; continue; }
-
-		// Bench
-		if ( argument[0] =='b' ) { bench=1; if (argument[1] >= '0') cLevel=argument[1] - '0'; continue; } 
-
-		// Modify Block Size (benchmark only)
-		if ( argument[0] =='B' ) { int B = argument[1] - '0'; int S = 1 << (10 + 2*B); BMK_SetBlocksize(S); continue; }
-
-		// Modify Nb Iterations (benchmark only)
-		if ( argument[0] =='i' ) { int iters = argument[1] - '0'; BMK_SetNbIterations(iters); continue; }
-
-		// Pause at the end (benchmark only)
-		if ( argument[0] =='p' ) { BMK_SetPause(); continue; }
-
-        // Test
-		if ( argument[0] =='t' ) { decode=1; output_filename=nulmark; continue; }
-	}
-
-	// first provided filename is input
-    if (!input_filename) { input_filename=argument; filenamesStart=i; continue; }
-
-	// second provided filename is output
-    if (!output_filename)
-	{
-		output_filename=argument;
-		if (!strcmp (output_filename, nullinput)) output_filename = nulmark;
-		continue;
-	}
-  }
-
-  // No input filename ==> Error
-  if(!input_filename) { badusage(exename); return 1; }
-
-  if (bench) return BMK_benchFile(argv+filenamesStart, argc-filenamesStart, cLevel);
-
-  // No output filename ==> Error
-  if (!output_filename) { badusage(exename); return 1; }
-
-  if (decode) return decode_file(input_filename, output_filename);
-
-  return compress_file(input_filename, output_filename, cLevel);   // Compression is 'default' action
-
-}

ext/lz4/lz4hc.c

@@ -1,6 +1,6 @@
 /*
    LZ4 HC - High Compression Mode of LZ4
-   Copyright (C) 2011-2012, Yann Collet.
+   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
@@ -31,12 +31,24 @@
    - 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(__x86_64) || defined(__amd64__) || defined(__amd64) || defined(__ppc64__) || defined(_WIN64) || defined(__LP64__) || defined(_LP64) )   // Detects 64 bits mode
+#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
@@ -52,7 +64,7 @@
 #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(__ppc__) || defined(_POWER) || defined(__powerpc__) || defined(_ARCH_PPC) || defined(__PPC__) || defined(__PPC) || defined(PPC) || defined(__powerpc__) || defined(__powerpc) || defined(powerpc) \
+   || defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) \
    || defined(__hpux)  || defined(__hppa) \
    || defined(_MIPSEB) || defined(__s390__)
 #  define LZ4_BIG_ENDIAN 1
@@ -76,78 +88,84 @@
 //**************************************
 // Compiler Options
 //**************************************
-#if __STDC_VERSION__ >= 199901L    // C99
+#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L   // C99
   /* "restrict" is a known keyword */
 #else
 #  define restrict  // Disable restrict
 #endif
 
-#ifdef _MSC_VER
-#  define inline __inline             // Visual is not C99, but supports some kind of inline
-#  define forceinline __forceinline   
-#  include <intrin.h>                 // For Visual 2005
-#  if LZ4_ARCH64	// 64-bit
+#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
+#  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 forceinline inline __attribute__((always_inline))
+#    define FORCE_INLINE static inline __attribute__((always_inline))
 #  else
-#    define forceinline inline
+#    define FORCE_INLINE static inline
 #  endif
 #endif
 
 #ifdef _MSC_VER  // Visual Studio
-#define lz4_bswap16(x) _byteswap_ushort(x)
+#  define lz4_bswap16(x) _byteswap_ushort(x)
 #else
-#define lz4_bswap16(x)  ((unsigned short int) ((((x) >> 8) & 0xffu) | (((x) & 0xffu) << 8)))
+#  define lz4_bswap16(x)  ((unsigned short int) ((((x) >> 8) & 0xffu) | (((x) & 0xffu) << 8)))
 #endif
 
 
 //**************************************
 // Includes
 //**************************************
-#include <stdlib.h>   // calloc, free
-#include <string.h>   // memset, memcpy
 #include "lz4hc.h"
-
-#define ALLOCATOR(s) calloc(1,s)
-#define FREEMEM free
-#define MEM_INIT memset
+#include "lz4.h"
 
 
 //**************************************
 // Basic Types
 //**************************************
-#if defined(_MSC_VER)    // Visual Studio does not support 'stdint' natively
-#define BYTE	unsigned __int8
-#define U16		unsigned __int16
-#define U32		unsigned __int32
-#define S32		__int32
-#define U64		unsigned __int64
+#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
-#include <stdint.h>
-#define BYTE	uint8_t
-#define U16		uint16_t
-#define U32		uint32_t
-#define S32		int32_t
-#define U64		uint64_t
+  typedef unsigned char       BYTE;
+  typedef unsigned short      U16;
+  typedef unsigned int        U32;
+  typedef   signed int        S32;
+  typedef unsigned long long  U64;
 #endif
 
-#ifndef LZ4_FORCE_UNALIGNED_ACCESS
-#pragma pack(push, 1) 
+#if defined(__GNUC__)  && !defined(LZ4_FORCE_UNALIGNED_ACCESS)
+#  define _PACKED __attribute__ ((packed))
+#else
+#  define _PACKED
 #endif
 
-typedef struct _U16_S { U16 v; } U16_S;
-typedef struct _U32_S { U32 v; } U32_S;
-typedef struct _U64_S { U64 v; } U64_S;
+#if !defined(LZ4_FORCE_UNALIGNED_ACCESS) && !defined(__GNUC__)
+#  ifdef __IBMC__
+#    pragma pack(1)
+#  else
+#    pragma pack(push, 1)
+#  endif
+#endif
 
-#ifndef LZ4_FORCE_UNALIGNED_ACCESS
-#pragma pack(pop) 
+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)
@@ -182,34 +200,40 @@ typedef struct _U64_S { U64 v; } U64_S;
 #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
+#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; }
+#  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
 
 
@@ -218,7 +242,9 @@ typedef struct _U64_S { U64 v; } U64_S;
 //************************************************************
 typedef struct 
 {
+    const BYTE* inputBuffer;
     const BYTE* base;
+    const BYTE* end;
     HTYPE hashTable[HASHTABLESIZE];
     U16 chainTable[MAXD];
     const BYTE* nextToUpdate;
@@ -230,11 +256,11 @@ typedef struct
 //**************************************
 #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))
+#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))
 
 
 //**************************************
@@ -242,99 +268,98 @@ typedef struct
 //**************************************
 #if LZ4_ARCH64
 
-inline static int LZ4_NbCommonBytes (register U64 val)
+FORCE_INLINE int LZ4_NbCommonBytes (register U64 val)
 {
 #if defined(LZ4_BIG_ENDIAN)
-    #if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+#  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)
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
     return (__builtin_clzll(val) >> 3); 
-    #else
+#  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
+#  endif
 #else
-    #if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+#  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)
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
     return (__builtin_ctzll(val) >> 3); 
-    #else
+#  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
 #endif
 }
 
 #else
 
-inline static int LZ4_NbCommonBytes (register U32 val)
+FORCE_INLINE int LZ4_NbCommonBytes (register U32 val)
 {
 #if defined(LZ4_BIG_ENDIAN)
-    #if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+#  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)
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
     return (__builtin_clz(val) >> 3); 
-    #else
+#  else
     int r;
     if (!(val>>16)) { r=2; val>>=8; } else { r=0; val>>=24; }
     r += (!val);
     return r;
-    #endif
+#  endif
 #else
-    #if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT)
+#  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)
+#  elif defined(__GNUC__) && ((__GNUC__ * 100 + __GNUC_MINOR__) >= 304) && !defined(LZ4_FORCE_SW_BITCOUNT)
     return (__builtin_ctz(val) >> 3); 
-    #else
+#  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
 }
 
 #endif
 
 
-inline static int LZ4HC_Init (LZ4HC_Data_Structure* hc4, const BYTE* base)
+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 + LZ4_ARCH64;
+    hc4->nextToUpdate = base + 1;
     hc4->base = base;
-    return 1;
+    hc4->inputBuffer = base;
+    hc4->end = base;
 }
 
 
-inline static void* LZ4HC_Create (const BYTE* base)
+void* LZ4_createHC (const char* inputBuffer)
 {
     void* hc4 = ALLOCATOR(sizeof(LZ4HC_Data_Structure));
-
-    LZ4HC_Init ((LZ4HC_Data_Structure*)hc4, base);
+    LZ4_initHC ((LZ4HC_Data_Structure*)hc4, (const BYTE*)inputBuffer);
     return hc4;
 }
 
 
-inline static int LZ4HC_Free (void** LZ4HC_Data)
+int LZ4_freeHC (void* LZ4HC_Data)
 {
-    FREEMEM(*LZ4HC_Data);
-    *LZ4HC_Data = NULL;
-    return (1);
+    FREEMEM(LZ4HC_Data);
+    return (0);
 }
 
 
 // Update chains up to ip (excluded)
-forceinline static void LZ4HC_Insert (LZ4HC_Data_Structure* hc4, const BYTE* ip)
+FORCE_INLINE void LZ4HC_Insert (LZ4HC_Data_Structure* hc4, const BYTE* ip)
 {
     U16*   chainTable = hc4->chainTable;
     HTYPE* HashTable  = hc4->hashTable;
@@ -342,17 +367,37 @@ forceinline static void LZ4HC_Insert (LZ4HC_Data_Structure* hc4, const BYTE* ip)
 
     while(hc4->nextToUpdate < ip)
     {
-        const BYTE* p = hc4->nextToUpdate;
+        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)] = (p) - base;
+        HashTable[HASH_VALUE(p)] = (HTYPE)((p) - base);
         hc4->nextToUpdate++;
     }
 }
 
 
-forceinline static size_t LZ4HC_CommonLength (const BYTE* p1, const BYTE* p2, const BYTE* const matchlimit)
+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;
 
@@ -370,7 +415,7 @@ forceinline static size_t LZ4HC_CommonLength (const BYTE* p1, const BYTE* p2, co
 }
 
 
-forceinline static int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4, const BYTE* ip, const BYTE* const matchlimit, const BYTE** matchpos)
+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;
@@ -378,7 +423,7 @@ forceinline static int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4,
     INITBASE(base,hc4->base);
     int nbAttempts=MAX_NB_ATTEMPTS;
     size_t repl=0, ml=0;
-    U16 delta;
+    U16 delta=0;  // useless assignment, to remove an uninitialization warning
 
     // HC4 match finder
     LZ4HC_Insert(hc4, ip);
@@ -387,7 +432,7 @@ forceinline static int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4,
 #define REPEAT_OPTIMIZATION
 #ifdef REPEAT_OPTIMIZATION
     // Detect repetitive sequences of length <= 4
-    if (ref >= ip-4)               // potential repetition
+    if ((U32)(ip-ref) <= 4)        // potential repetition
     {
         if (A32(ref) == A32(ip))   // confirmed
         {
@@ -399,7 +444,7 @@ forceinline static int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4,
     }
 #endif
 
-    while ((ref >= ip-MAX_DISTANCE) && (nbAttempts))
+    while (((U32)(ip-ref) <= MAX_DISTANCE) && (nbAttempts))
     {
         nbAttempts--;
         if (*(ref+ml) == *(ip+ml))
@@ -427,7 +472,7 @@ forceinline static int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4,
         do
         {
             DELTANEXT(ptr) = delta;    
-            HashTable[HASH_VALUE(ptr)] = (ptr) - base;     // Head of chain
+            HashTable[HASH_VALUE(ptr)] = (HTYPE)((ptr) - base);     // Head of chain
             ptr++;
         } while(ptr < end);
         hc4->nextToUpdate = end;
@@ -438,7 +483,7 @@ forceinline static int LZ4HC_InsertAndFindBestMatch (LZ4HC_Data_Structure* hc4,
 }
 
 
-forceinline static int LZ4HC_InsertAndGetWiderMatch (LZ4HC_Data_Structure* hc4, const BYTE* ip, const BYTE* startLimit, const BYTE* matchlimit, int longest, const BYTE** matchpos, const BYTE** startpos)
+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;
@@ -451,7 +496,7 @@ forceinline static int LZ4HC_InsertAndGetWiderMatch (LZ4HC_Data_Structure* hc4,
     LZ4HC_Insert(hc4, ip);
     ref = HASH_POINTER(ip);
 
-    while ((ref >= ip-MAX_DISTANCE) && (nbAttempts))
+    while (((U32)(ip-ref) <= MAX_DISTANCE) && (nbAttempts))
     {
         nbAttempts--;
         if (*(startLimit + longest) == *(ref - delta + longest))
@@ -481,7 +526,7 @@ _endCount:
             const BYTE* ipt = ip + MINMATCH + LZ4HC_CommonLength(ip+MINMATCH, ref+MINMATCH, matchlimit);
 #endif
 
-            while ((startt>startLimit) && (reft > hc4->base) && (startt[-1] == reft[-1])) {startt--; reft--;}
+            while ((startt>startLimit) && (reft > hc4->inputBuffer) && (startt[-1] == reft[-1])) {startt--; reft--;}
 
             if ((ipt-startt) > longest)
             {
@@ -497,16 +542,26 @@ _endCount:
 }
 
 
-forceinline static int LZ4_encodeSequence(const BYTE** ip, BYTE** op, const BYTE** anchor, int ml, const BYTE* ref)
+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, len; 
+    int length;
     BYTE* token;
 
     // Encode Literal length
     length = (int)(*ip - *anchor);
     token = (*op)++;
-    if (length>=(int)RUN_MASK) { *token=(RUN_MASK<<ML_BITS); len = length-RUN_MASK; for(; len > 254 ; len-=255) *(*op)++ = 255;  *(*op)++ = (BYTE)len; } 
-    else *token = (length<<ML_BITS);
+    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);
@@ -515,36 +570,39 @@ forceinline static int LZ4_encodeSequence(const BYTE** ip, BYTE** op, const BYTE
     LZ4_WRITE_LITTLEENDIAN_16(*op,(U16)(*ip-ref));
 
     // Encode MatchLength
-    len = (int)(ml-MINMATCH);
-    if (len>=(int)ML_MASK) { *token+=ML_MASK; len-=ML_MASK; for(; len > 509 ; len-=510) { *(*op)++ = 255; *(*op)++ = 255; } if (len > 254) { len-=255; *(*op)++ = 255; } *(*op)++ = (BYTE)len; } 
-    else *token += len;	
+    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 += ml;
+    *ip += matchLength;
     *anchor = *ip; 
 
     return 0;
 }
 
 
-//****************************
-// Compression CODE
-//****************************
-
-int LZ4_compressHCCtx(LZ4HC_Data_Structure* ctx,
+static int LZ4HC_compress_generic (
+                 void* ctxvoid,
                  const char* source, 
                  char* dest,
-                 int isize)
-{	
+                 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 + isize;
+    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;
+    int   ml, ml2, ml3, ml0;
     const BYTE* ref=NULL;
     const BYTE* start2=NULL;
     const BYTE* ref2=NULL;
@@ -553,6 +611,11 @@ int LZ4_compressHCCtx(LZ4HC_Data_Structure* ctx,
     const BYTE* start0;
     const BYTE* ref0;
 
+
+    // Ensure blocks follow each other
+    if (ip != ctx->end) return 0;
+    ctx->end += inputSize;
+
     ip++;
 
     // Main Loop
@@ -573,7 +636,7 @@ _Search2:
 
         if (ml2 == ml)  // No better match
         {
-            LZ4_encodeSequence(&ip, &op, &anchor, ml, ref);
+            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
             continue;
         }
 
@@ -625,9 +688,9 @@ _Search3:
             // ip & ref are known; Now for ml
             if (start2 < ip+ml)  ml = (int)(start2 - ip);
             // Now, encode 2 sequences
-            LZ4_encodeSequence(&ip, &op, &anchor, ml, ref);
+            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
             ip = start2;
-            LZ4_encodeSequence(&ip, &op, &anchor, ml2, ref2);
+            if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml2, ref2, limit, oend)) return 0;
             continue;
         }
 
@@ -649,7 +712,7 @@ _Search3:
                     }
                 }
 
-                LZ4_encodeSequence(&ip, &op, &anchor, ml, ref);
+                if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
                 ip  = start3;
                 ref = ref3;
                 ml  = ml3;
@@ -688,7 +751,7 @@ _Search3:
                 ml = (int)(start2 - ip);
             }
         }
-        LZ4_encodeSequence(&ip, &op, &anchor, ml, ref);
+        if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) return 0;
 
         ip = start2;
         ref = ref2;
@@ -705,8 +768,9 @@ _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++ = (lastRun<<ML_BITS);
+        else *op++ = (BYTE)(lastRun<<ML_BITS);
         memcpy(op, anchor, iend - anchor);
         op += iend-anchor;
     } 
@@ -716,15 +780,38 @@ _Search3:
 }
 
 
-int LZ4_compressHC(const char* source, 
-                 char* dest,
-                 int isize)
+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 = LZ4HC_Create((const BYTE*)source);
-    int result = LZ4_compressHCCtx(ctx, source, dest, isize);
-    LZ4HC_Free (&ctx);
+    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);
+}
 

ext/lz4/lz4hc.h

@@ -1,7 +1,7 @@
 /*
    LZ4 HC - High Compression Mode of LZ4
    Header File
-   Copyright (C) 2011-2012, Yann Collet.
+   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
@@ -39,19 +39,70 @@ extern "C" {
 #endif
 
 
-int LZ4_compressHC (const char* source, char* dest, int isize);
-
+int LZ4_compressHC (const char* source, char* dest, int inputSize);
 /*
 LZ4_compressHC :
-	return : the number of bytes in compressed buffer dest
-	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")
+    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 regular LZ4 source code (see "lz4.h") (BSD license)
+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.
 */