facebook
/
zstd
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

huff0 dynamic reduction

dev
Yann Collet 2016-05-20 14:36:36 +02:00
parent 45f84ab7cc
commit f22a0d653d
7 changed files with 153 additions and 126 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
lib/common/bitstream.h
View File

@ -162,14 +162,12 @@ MEM_STATIC unsigned BIT_highbit32 (register U32 val)
# else /* Software version */
static const unsigned DeBruijnClz[32] = { 0, 9, 1, 10, 13, 21, 2, 29, 11, 14, 16, 18, 22, 25, 3, 30, 8, 12, 20, 28, 15, 17, 24, 7, 19, 27, 23, 6, 26, 5, 4, 31 };
U32 v = val;
unsigned r;
v |= v >> 1;
v |= v >> 2;
v |= v >> 4;
v |= v >> 8;
v |= v >> 16;
r = DeBruijnClz[ (U32) (v * 0x07C4ACDDU) >> 27];
return r;
return DeBruijnClz[ (U32) (v * 0x07C4ACDDU) >> 27];
# endif
}
@ -375,7 +373,7 @@ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, U32 nbBits)
if status == unfinished, internal register is filled with >= (sizeof(bitD->bitContainer)*8 - 7) bits */
MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD)
{
if (bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8)) /* should never happen */
if (bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8)) /* should not happen => corruption detected */
return BIT_DStream_overflow;
if (bitD->ptr >= bitD->start + sizeof(bitD->bitContainer)) {

17
lib/common/fse_static.h
View File

@ -64,19 +64,22 @@ extern "C" {
* FSE advanced API
*******************************************/
size_t FSE_countFast(unsigned* count, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize);
/* same as FSE_count(), but blindly trusts that all byte values within src are <= *maxSymbolValuePtr */
/**< same as FSE_count(), but blindly trusts that all byte values within src are <= *maxSymbolValuePtr */
unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus);
/**< same as FSE_optimalTableLog(), which used `minus==2` */
size_t FSE_buildCTable_raw (FSE_CTable* ct, unsigned nbBits);
/* build a fake FSE_CTable, designed to not compress an input, where each symbol uses nbBits */
/**< build a fake FSE_CTable, designed to not compress an input, where each symbol uses nbBits */
size_t FSE_buildCTable_rle (FSE_CTable* ct, unsigned char symbolValue);
/* build a fake FSE_CTable, designed to compress always the same symbolValue */
/**< build a fake FSE_CTable, designed to compress always the same symbolValue */
size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits);
/* build a fake FSE_DTable, designed to read an uncompressed bitstream where each symbol uses nbBits */
/**< build a fake FSE_DTable, designed to read an uncompressed bitstream where each symbol uses nbBits */
size_t FSE_buildDTable_rle (FSE_DTable* dt, unsigned char symbolValue);
/* build a fake FSE_DTable, designed to always generate the same symbolValue */
/**< build a fake FSE_DTable, designed to always generate the same symbolValue */
/* *****************************************
@ -103,7 +106,7 @@ static void FSE_encodeSymbol(BIT_CStream_t* bitC, FSE_CState_t* CStatePtr, unsig
static void FSE_flushCState(BIT_CStream_t* bitC, const FSE_CState_t* CStatePtr);
/*!
/**<
These functions are inner components of FSE_compress_usingCTable().
They allow the creation of custom streams, mixing multiple tables and bit sources.
@ -163,7 +166,7 @@ static unsigned char FSE_decodeSymbol(FSE_DState_t* DStatePtr, BIT_DStream_t* bi
static unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr);
/*!
/**<
Let's now decompose FSE_decompress_usingDTable() into its unitary components.
You will decode FSE-encoded symbols from the bitStream,
and also any other bitFields you put in, **in reverse order**.

1
lib/common/huf.h
View File

@ -77,6 +77,7 @@ HUF_decompress() :
/* ****************************************
* Tool functions
******************************************/
#define HUF_BLOCKSIZE_MAX (128 * 1024)
size_t HUF_compressBound(size_t size); /**< maximum compressed size */
/* Error Management */

29
lib/common/huf_static.h
View File

@ -85,15 +85,17 @@ size_t HUF_decompress4X6 (void* dst, size_t dstSize, const void* cSrc, size_t cS
/*!
HUF_compress() does the following:
1. count symbol occurrence from source[] into table count[] using FSE_count()
2. build Huffman table from count using HUF_buildCTable()
3. save Huffman table to memory buffer using HUF_writeCTable()
4. encode the data stream using HUF_compress4X_usingCTable()
2. (optional) refine tableLog using HUF_optimalTableLog()
3. build Huffman table from count using HUF_buildCTable()
4. save Huffman table to memory buffer using HUF_writeCTable()
5. encode the data stream using HUF_compress4X_usingCTable()
The following API allows targeting specific sub-functions for advanced tasks.
For example, it's possible to compress several blocks using the same 'CTable',
or to save and regenerate 'CTable' using external methods.
*/
/* FSE_count() : find it within "fse.h" */
unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue);
typedef struct HUF_CElt_s HUF_CElt; /* incomplete type */
size_t HUF_buildCTable (HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits);
size_t HUF_writeCTable (void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog);
@ -137,16 +139,19 @@ size_t HUF_readCTable (HUF_CElt* CTable, unsigned maxSymbolValue, const void* sr
/* **************************************************************
* Constants
****************************************************************/
#define HUF_ABSOLUTEMAX_TABLELOG 16 /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */
#define HUF_MAX_TABLELOG 12 /* max configured tableLog (for static allocation); can be modified up to HUF_ABSOLUTEMAX_TABLELOG */
#define HUF_DEFAULT_TABLELOG HUF_MAX_TABLELOG /* tableLog by default, when not specified */
#define HUF_MAX_SYMBOL_VALUE 255
#if (HUF_MAX_TABLELOG > HUF_ABSOLUTEMAX_TABLELOG)
# error "HUF_MAX_TABLELOG is too large !"
#define HUF_TABLELOG_ABSOLUTEMAX 16 /* absolute limit of HUF_MAX_TABLELOG. Beyond that value, code does not work */
#define HUF_TABLELOG_MAX 12 /* max configured tableLog (for static allocation); can be modified up to HUF_ABSOLUTEMAX_TABLELOG */
#define HUF_TABLELOG_DEFAULT HUF_TABLELOG_MAX /* tableLog by default, when not specified */
#define HUF_SYMBOLVALUE_MAX 255
#if (HUF_TABLELOG_MAX > HUF_TABLELOG_ABSOLUTEMAX)
# error "HUF_TABLELOG_MAX is too large !"
#endif
/* **************************************************************
* Needed by zstd in both compression and decompression
****************************************************************/
/*! HUF_readStats() :
Read compact Huffman tree, saved by HUF_writeCTable().
`huffWeight` is destination buffer.
@ -188,17 +193,17 @@ MEM_STATIC size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
}
/* collect weight stats */
memset(rankStats, 0, (HUF_ABSOLUTEMAX_TABLELOG + 1) * sizeof(U32));
memset(rankStats, 0, (HUF_TABLELOG_ABSOLUTEMAX + 1) * sizeof(U32));
weightTotal = 0;
{ U32 n; for (n=0; n<oSize; n++) {
if (huffWeight[n] >= HUF_ABSOLUTEMAX_TABLELOG) return ERROR(corruption_detected);
if (huffWeight[n] >= HUF_TABLELOG_ABSOLUTEMAX) return ERROR(corruption_detected);
rankStats[huffWeight[n]]++;
weightTotal += (1 << huffWeight[n]) >> 1;
} }
/* get last non-null symbol weight (implied, total must be 2^n) */
{ U32 const tableLog = BIT_highbit32(weightTotal) + 1;
if (tableLog > HUF_ABSOLUTEMAX_TABLELOG) return ERROR(corruption_detected);
if (tableLog > HUF_TABLELOG_ABSOLUTEMAX) return ERROR(corruption_detected);
*tableLogPtr = tableLog;
/* determine last weight */
{ U32 const total = 1 << tableLog;

11
lib/compress/fse_compress.c
View File

@ -442,7 +442,7 @@ FSE_CTable* FSE_createCTable (unsigned maxSymbolValue, unsigned tableLog)
return (FSE_CTable*)malloc(size);
}
void FSE_freeCTable (FSE_CTable* ct) { free(ct); }
void FSE_freeCTable (FSE_CTable* ct) { free(ct); }
/* provides the minimum logSize to safely represent a distribution */
static unsigned FSE_minTableLog(size_t srcSize, unsigned maxSymbolValue)
@ -453,9 +453,9 @@ static unsigned FSE_minTableLog(size_t srcSize, unsigned maxSymbolValue)
return minBits;
}
unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue)
unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus)
{
U32 maxBitsSrc = BIT_highbit32((U32)(srcSize - 1)) - 2;
U32 maxBitsSrc = BIT_highbit32((U32)(srcSize - 1)) - minus;
U32 tableLog = maxTableLog;
U32 minBits = FSE_minTableLog(srcSize, maxSymbolValue);
if (tableLog==0) tableLog = FSE_DEFAULT_TABLELOG;
@ -466,6 +466,11 @@ unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxS
return tableLog;
}
unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue)
{
return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 2);
}
/* Secondary normalization method.
To be used when primary method fails. */

157
lib/compress/huf_compress.c
View File

@ -64,7 +64,7 @@
#include <stdio.h> /* printf (debug) */
#include "huf_static.h"
#include "bitstream.h"
#include "fse.h" /* header compression */
#include "fse_static.h" /* header compression */
/* **************************************************************
@ -73,6 +73,15 @@
#define HUF_STATIC_ASSERT(c) { enum { HUF_static_assert = 1/(int)(!!(c)) }; } /* use only *after* variable declarations */
/* **************************************************************
* Utils
****************************************************************/
unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue)
{
return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 1);
}
/* *******************************************************
* HUF : Huffman block compression
*********************************************************/
@ -94,14 +103,14 @@ typedef struct nodeElt_s {
size_t HUF_writeCTable (void* dst, size_t maxDstSize,
const HUF_CElt* CTable, U32 maxSymbolValue, U32 huffLog)
{
BYTE bitsToWeight[HUF_MAX_TABLELOG + 1];
BYTE huffWeight[HUF_MAX_SYMBOL_VALUE + 1];
BYTE bitsToWeight[HUF_TABLELOG_MAX + 1];
BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1];
U32 n;
BYTE* op = (BYTE*)dst;
size_t size;
/* check conditions */
if (maxSymbolValue > HUF_MAX_SYMBOL_VALUE + 1)
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX + 1)
return ERROR(GENERIC);
/* convert to weight */
@ -159,19 +168,19 @@ size_t HUF_writeCTable (void* dst, size_t maxDstSize,
size_t HUF_readCTable (HUF_CElt* CTable, U32 maxSymbolValue, const void* src, size_t srcSize)
{
BYTE huffWeight[HUF_MAX_SYMBOL_VALUE + 1];
U32 rankVal[HUF_ABSOLUTEMAX_TABLELOG + 1]; /* large enough for values from 0 to 16 */
BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1];
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1]; /* large enough for values from 0 to 16 */
U32 tableLog = 0;
size_t readSize;
U32 nbSymbols = 0;
//memset(huffWeight, 0, sizeof(huffWeight)); /* is not necessary, even though some analyzer complain ... */
/* get symbol weights */
readSize = HUF_readStats(huffWeight, HUF_MAX_SYMBOL_VALUE+1, rankVal, &nbSymbols, &tableLog, src, srcSize);
readSize = HUF_readStats(huffWeight, HUF_SYMBOLVALUE_MAX+1, rankVal, &nbSymbols, &tableLog, src, srcSize);
if (HUF_isError(readSize)) return readSize;
/* check result */
if (tableLog > HUF_MAX_TABLELOG) return ERROR(tableLog_tooLarge);
if (tableLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
if (nbSymbols > maxSymbolValue+1) return ERROR(maxSymbolValue_tooSmall);
/* Prepare base value per rank */
@ -189,12 +198,12 @@ size_t HUF_readCTable (HUF_CElt* CTable, U32 maxSymbolValue, const void* src, si
}}
/* fill val */
{ U16 nbPerRank[HUF_MAX_TABLELOG+1] = {0};
U16 valPerRank[HUF_MAX_TABLELOG+1] = {0};
{ U16 nbPerRank[HUF_TABLELOG_MAX+1] = {0};
U16 valPerRank[HUF_TABLELOG_MAX+1] = {0};
{ U32 n; for (n=0; n<nbSymbols; n++) nbPerRank[CTable[n].nbBits]++; }
/* determine stating value per rank */
{ U16 min = 0;
U32 n; for (n=HUF_MAX_TABLELOG; n>0; n--) {
U32 n; for (n=HUF_TABLELOG_MAX; n>0; n--) {
valPerRank[n] = min; /* get starting value within each rank */
min += nbPerRank[n];
min >>= 1;
@ -229,7 +238,7 @@ static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
/* repay normalized cost */
{ U32 const noSymbol = 0xF0F0F0F0;
U32 rankLast[HUF_MAX_TABLELOG+1];
U32 rankLast[HUF_TABLELOG_MAX+1];
int pos;
/* Get pos of last (smallest) symbol per rank */
@ -253,7 +262,7 @@ static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
if (highTotal <= lowTotal) break;
} }
/* only triggered when no more rank 1 symbol left => find closest one (note : there is necessarily at least one !) */
while ((nBitsToDecrease<=HUF_MAX_TABLELOG) && (rankLast[nBitsToDecrease] == noSymbol)) /* HUF_MAX_TABLELOG test just to please gcc 5+; but it should not be necessary */
while ((nBitsToDecrease<=HUF_TABLELOG_MAX) && (rankLast[nBitsToDecrease] == noSymbol)) /* HUF_MAX_TABLELOG test just to please gcc 5+; but it should not be necessary */
nBitsToDecrease ++;
totalCost -= 1 << (nBitsToDecrease-1);
if (rankLast[nBitsToDecrease-1] == noSymbol)
@ -312,10 +321,10 @@ static void HUF_sort(nodeElt* huffNode, const U32* count, U32 maxSymbolValue)
}
#define STARTNODE (HUF_MAX_SYMBOL_VALUE+1)
#define STARTNODE (HUF_SYMBOLVALUE_MAX+1)
size_t HUF_buildCTable (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits)
{
nodeElt huffNode0[2*HUF_MAX_SYMBOL_VALUE+1 +1];
nodeElt huffNode0[2*HUF_SYMBOLVALUE_MAX+1 +1];
nodeElt* huffNode = huffNode0 + 1;
U32 n, nonNullRank;
int lowS, lowN;
@ -323,8 +332,8 @@ size_t HUF_buildCTable (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U3
U32 nodeRoot;
/* safety checks */
if (maxNbBits == 0) maxNbBits = HUF_DEFAULT_TABLELOG;
if (maxSymbolValue > HUF_MAX_SYMBOL_VALUE) return ERROR(GENERIC);
if (maxNbBits == 0) maxNbBits = HUF_TABLELOG_DEFAULT;
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(GENERIC);
memset(huffNode0, 0, sizeof(huffNode0));
/* sort, decreasing order */
@ -360,9 +369,9 @@ size_t HUF_buildCTable (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U3
maxNbBits = HUF_setMaxHeight(huffNode, nonNullRank, maxNbBits);
/* fill result into tree (val, nbBits) */
{ U16 nbPerRank[HUF_MAX_TABLELOG+1] = {0};
U16 valPerRank[HUF_MAX_TABLELOG+1] = {0};
if (maxNbBits > HUF_MAX_TABLELOG) return ERROR(GENERIC); /* check fit into table */
{ U16 nbPerRank[HUF_TABLELOG_MAX+1] = {0};
U16 valPerRank[HUF_TABLELOG_MAX+1] = {0};
if (maxNbBits > HUF_TABLELOG_MAX) return ERROR(GENERIC); /* check fit into table */
for (n=0; n<=nonNullRank; n++)
nbPerRank[huffNode[n].nbBits]++;
/* determine stating value per rank */
@ -391,10 +400,10 @@ size_t HUF_compressBound(size_t size) { return HUF_COMPRESSBOUND(size); }
#define HUF_FLUSHBITS(s) (fast ? BIT_flushBitsFast(s) : BIT_flushBits(s))
#define HUF_FLUSHBITS_1(stream) \
if (sizeof((stream)->bitContainer)*8 < HUF_MAX_TABLELOG*2+7) HUF_FLUSHBITS(stream)
if (sizeof((stream)->bitContainer)*8 < HUF_TABLELOG_MAX*2+7) HUF_FLUSHBITS(stream)
#define HUF_FLUSHBITS_2(stream) \
if (sizeof((stream)->bitContainer)*8 < HUF_MAX_TABLELOG*4+7) HUF_FLUSHBITS(stream)
if (sizeof((stream)->bitContainer)*8 < HUF_TABLELOG_MAX*4+7) HUF_FLUSHBITS(stream)
size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
{
@ -441,44 +450,47 @@ size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, si
size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
{
size_t segmentSize = (srcSize+3)/4; /* first 3 segments */
size_t const segmentSize = (srcSize+3)/4; /* first 3 segments */
const BYTE* ip = (const BYTE*) src;
const BYTE* const iend = ip + srcSize;
BYTE* const ostart = (BYTE*) dst;
BYTE* const oend = ostart + dstSize;
BYTE* op = ostart;
size_t errorCode;
if (dstSize < 6 + 1 + 1 + 1 + 8) return 0; /* minimum space to compress successfully */
if (srcSize < 12) return 0; /* no saving possible : too small input */
op += 6; /* jumpTable */
errorCode = HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable);
if (HUF_isError(errorCode)) return errorCode;
if (errorCode==0) return 0;
MEM_writeLE16(ostart, (U16)errorCode);
{ size_t const cSize = HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable);
if (HUF_isError(cSize)) return cSize;
if (cSize==0) return 0;
MEM_writeLE16(ostart, (U16)cSize);
op += cSize;
}
ip += segmentSize;
op += errorCode;
errorCode = HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable);
if (HUF_isError(errorCode)) return errorCode;
if (errorCode==0) return 0;
MEM_writeLE16(ostart+2, (U16)errorCode);
{ size_t const cSize = HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable);
if (HUF_isError(cSize)) return cSize;
if (cSize==0) return 0;
MEM_writeLE16(ostart+2, (U16)cSize);
op += cSize;
}
ip += segmentSize;
op += errorCode;
errorCode = HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable);
if (HUF_isError(errorCode)) return errorCode;
if (errorCode==0) return 0;
MEM_writeLE16(ostart+4, (U16)errorCode);
{ size_t const cSize = HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable);
if (HUF_isError(cSize)) return cSize;
if (cSize==0) return 0;
MEM_writeLE16(ostart+4, (U16)cSize);
op += cSize;
}
ip += segmentSize;
op += errorCode;
errorCode = HUF_compress1X_usingCTable(op, oend-op, ip, iend-ip, CTable);
if (HUF_isError(errorCode)) return errorCode;
if (errorCode==0) return 0;
{ size_t const cSize = HUF_compress1X_usingCTable(op, oend-op, ip, iend-ip, CTable);
if (HUF_isError(cSize)) return cSize;
if (cSize==0) return 0;
op += cSize;
}
op += errorCode;
return op-ostart;
}
@ -493,43 +505,46 @@ static size_t HUF_compress_internal (
BYTE* const oend = ostart + dstSize;
BYTE* op = ostart;
U32 count[HUF_MAX_SYMBOL_VALUE+1];
HUF_CElt CTable[HUF_MAX_SYMBOL_VALUE+1];
size_t errorCode;
U32 count[HUF_SYMBOLVALUE_MAX+1];
HUF_CElt CTable[HUF_SYMBOLVALUE_MAX+1];
/* checks & inits */
if (srcSize < 1) return 0; /* Uncompressed - note : 1 means rle, so first byte must be correct */
if (dstSize < 1) return 0; /* not compressible within dst budget */
if (srcSize > 128 * 1024) return ERROR(srcSize_wrong); /* current block size limit */
if (huffLog > HUF_MAX_TABLELOG) return ERROR(tableLog_tooLarge);
if (!maxSymbolValue) maxSymbolValue = HUF_MAX_SYMBOL_VALUE;
if (!huffLog) huffLog = HUF_DEFAULT_TABLELOG;
if (!srcSize) return 0; /* Uncompressed (note : 1 means rle, so first byte must be correct) */
if (!dstSize) return 0; /* cannot fit within dst budget */
if (srcSize > HUF_BLOCKSIZE_MAX) return ERROR(srcSize_wrong); /* current block size limit */
if (huffLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
if (!maxSymbolValue) maxSymbolValue = HUF_SYMBOLVALUE_MAX;
if (!huffLog) huffLog = HUF_TABLELOG_DEFAULT;
/* Scan input and build symbol stats */
errorCode = FSE_count (count, &maxSymbolValue, (const BYTE*)src, srcSize);
if (HUF_isError(errorCode)) return errorCode;
if (errorCode == srcSize) { *ostart = ((const BYTE*)src)[0]; return 1; }
if (errorCode <= (srcSize >> 7)+1) return 0; /* Heuristic : not compressible enough */
{ size_t const largest = FSE_count (count, &maxSymbolValue, (const BYTE*)src, srcSize);
if (HUF_isError(largest)) return largest;
if (largest == srcSize) { *ostart = ((const BYTE*)src)[0]; return 1; } /* rle */
if (largest <= (srcSize >> 7)+1) return 0; /* Fast heuristic : not compressible enough */
}
/* Build Huffman Tree */
errorCode = HUF_buildCTable (CTable, count, maxSymbolValue, huffLog);
if (HUF_isError(errorCode)) return errorCode;
huffLog = (U32)errorCode;
huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue);
{ size_t const maxBits = HUF_buildCTable (CTable, count, maxSymbolValue, huffLog);
if (HUF_isError(maxBits)) return maxBits;
huffLog = (U32)maxBits;
}
/* Write table description header */
errorCode = HUF_writeCTable (op, dstSize, CTable, maxSymbolValue, huffLog);
if (HUF_isError(errorCode)) return errorCode;
if (errorCode + 12 >= srcSize) return 0; /* not useful to try compression */
op += errorCode;
{ size_t const hSize = HUF_writeCTable (op, dstSize, CTable, maxSymbolValue, huffLog);
if (HUF_isError(hSize)) return hSize;
if (hSize + 12 >= srcSize) return 0; /* not useful to try compression */
op += hSize;
}
/* Compress */
if (singleStream)
errorCode = HUF_compress1X_usingCTable(op, oend - op, src, srcSize, CTable); /* single segment */
else
errorCode = HUF_compress4X_usingCTable(op, oend - op, src, srcSize, CTable);
if (HUF_isError(errorCode)) return errorCode;
if (errorCode==0) return 0;
op += errorCode;
{ size_t const cSize = (singleStream) ?
HUF_compress1X_usingCTable(op, oend - op, src, srcSize, CTable) : /* single segment */
HUF_compress4X_usingCTable(op, oend - op, src, srcSize, CTable);
if (HUF_isError(cSize)) return cSize;
if (cSize==0) return 0; /* uncompressible */
op += cSize;
}
/* check compressibility */
if ((size_t)(op-ostart) >= srcSize-1)
@ -556,5 +571,5 @@ size_t HUF_compress2 (void* dst, size_t dstSize,
size_t HUF_compress (void* dst, size_t maxDstSize, const void* src, size_t srcSize)
{
return HUF_compress2(dst, maxDstSize, src, (U32)srcSize, 255, HUF_DEFAULT_TABLELOG);
return HUF_compress2(dst, maxDstSize, src, (U32)srcSize, 255, HUF_TABLELOG_DEFAULT);
}

58
lib/decompress/huf_decompress.c
View File

@ -92,8 +92,8 @@ typedef struct { BYTE symbol; BYTE weight; } sortedSymbol_t;
size_t HUF_readDTableX2 (U16* DTable, const void* src, size_t srcSize)
{
BYTE huffWeight[HUF_MAX_SYMBOL_VALUE + 1];
U32 rankVal[HUF_ABSOLUTEMAX_TABLELOG + 1]; /* large enough for values from 0 to 16 */
BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1];
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1]; /* large enough for values from 0 to 16 */
U32 tableLog = 0;
size_t iSize;
U32 nbSymbols = 0;
@ -105,7 +105,7 @@ size_t HUF_readDTableX2 (U16* DTable, const void* src, size_t srcSize)
HUF_STATIC_ASSERT(sizeof(HUF_DEltX2) == sizeof(U16)); /* if compilation fails here, assertion is false */
//memset(huffWeight, 0, sizeof(huffWeight)); /* is not necessary, even though some analyzer complain ... */
iSize = HUF_readStats(huffWeight, HUF_MAX_SYMBOL_VALUE + 1, rankVal, &nbSymbols, &tableLog, src, srcSize);
iSize = HUF_readStats(huffWeight, HUF_SYMBOLVALUE_MAX + 1, rankVal, &nbSymbols, &tableLog, src, srcSize);
if (HUF_isError(iSize)) return iSize;
/* check result */
@ -148,7 +148,7 @@ static BYTE HUF_decodeSymbolX2(BIT_DStream_t* Dstream, const HUF_DEltX2* dt, con
*ptr++ = HUF_decodeSymbolX2(DStreamPtr, dt, dtLog)
#define HUF_DECODE_SYMBOLX2_1(ptr, DStreamPtr) \
if (MEM_64bits() || (HUF_MAX_TABLELOG<=12)) \
if (MEM_64bits() || (HUF_TABLELOG_MAX<=12)) \
HUF_DECODE_SYMBOLX2_0(ptr, DStreamPtr)
#define HUF_DECODE_SYMBOLX2_2(ptr, DStreamPtr) \
@ -203,7 +203,7 @@ size_t HUF_decompress1X2_usingDTable(
size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_MAX_TABLELOG);
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
const BYTE* ip = (const BYTE*) cSrc;
size_t const errorCode = HUF_readDTableX2 (DTable, cSrc, cSrcSize);
@ -312,7 +312,7 @@ size_t HUF_decompress4X2_usingDTable(
size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_MAX_TABLELOG);
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
const BYTE* ip = (const BYTE*) cSrc;
size_t const errorCode = HUF_readDTableX2 (DTable, cSrc, cSrcSize);
@ -335,7 +335,7 @@ static void HUF_fillDTableX4Level2(HUF_DEltX4* DTable, U32 sizeLog, const U32 co
U32 nbBitsBaseline, U16 baseSeq)
{
HUF_DEltX4 DElt;
U32 rankVal[HUF_ABSOLUTEMAX_TABLELOG + 1];
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1];
/* get pre-calculated rankVal */
memcpy(rankVal, rankValOrigin, sizeof(rankVal));
@ -369,14 +369,14 @@ static void HUF_fillDTableX4Level2(HUF_DEltX4* DTable, U32 sizeLog, const U32 co
}}
}
typedef U32 rankVal_t[HUF_ABSOLUTEMAX_TABLELOG][HUF_ABSOLUTEMAX_TABLELOG + 1];
typedef U32 rankVal_t[HUF_TABLELOG_ABSOLUTEMAX][HUF_TABLELOG_ABSOLUTEMAX + 1];
static void HUF_fillDTableX4(HUF_DEltX4* DTable, const U32 targetLog,
const sortedSymbol_t* sortedList, const U32 sortedListSize,
const U32* rankStart, rankVal_t rankValOrigin, const U32 maxWeight,
const U32 nbBitsBaseline)
{
U32 rankVal[HUF_ABSOLUTEMAX_TABLELOG + 1];
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1];
const int scaleLog = nbBitsBaseline - targetLog; /* note : targetLog >= srcLog, hence scaleLog <= 1 */
const U32 minBits = nbBitsBaseline - maxWeight;
U32 s;
@ -415,10 +415,10 @@ static void HUF_fillDTableX4(HUF_DEltX4* DTable, const U32 targetLog,
size_t HUF_readDTableX4 (U32* DTable, const void* src, size_t srcSize)
{
BYTE weightList[HUF_MAX_SYMBOL_VALUE + 1];
sortedSymbol_t sortedSymbol[HUF_MAX_SYMBOL_VALUE + 1];
U32 rankStats[HUF_ABSOLUTEMAX_TABLELOG + 1] = { 0 };
U32 rankStart0[HUF_ABSOLUTEMAX_TABLELOG + 2] = { 0 };
BYTE weightList[HUF_SYMBOLVALUE_MAX + 1];
sortedSymbol_t sortedSymbol[HUF_SYMBOLVALUE_MAX + 1];
U32 rankStats[HUF_TABLELOG_ABSOLUTEMAX + 1] = { 0 };
U32 rankStart0[HUF_TABLELOG_ABSOLUTEMAX + 2] = { 0 };
U32* const rankStart = rankStart0+1;
rankVal_t rankVal;
U32 tableLog, maxW, sizeOfSort, nbSymbols;
@ -428,10 +428,10 @@ size_t HUF_readDTableX4 (U32* DTable, const void* src, size_t srcSize)
HUF_DEltX4* const dt = ((HUF_DEltX4*)dtPtr) + 1;
HUF_STATIC_ASSERT(sizeof(HUF_DEltX4) == sizeof(U32)); /* if compilation fails here, assertion is false */
if (memLog > HUF_ABSOLUTEMAX_TABLELOG) return ERROR(tableLog_tooLarge);
if (memLog > HUF_TABLELOG_ABSOLUTEMAX) return ERROR(tableLog_tooLarge);
//memset(weightList, 0, sizeof(weightList)); /* is not necessary, even though some analyzer complain ... */
iSize = HUF_readStats(weightList, HUF_MAX_SYMBOL_VALUE + 1, rankStats, &nbSymbols, &tableLog, src, srcSize);
iSize = HUF_readStats(weightList, HUF_SYMBOLVALUE_MAX + 1, rankStats, &nbSymbols, &tableLog, src, srcSize);
if (HUF_isError(iSize)) return iSize;
/* check result */
@ -517,7 +517,7 @@ static U32 HUF_decodeLastSymbolX4(void* op, BIT_DStream_t* DStream, const HUF_DE
ptr += HUF_decodeSymbolX4(ptr, DStreamPtr, dt, dtLog)
#define HUF_DECODE_SYMBOLX4_1(ptr, DStreamPtr) \
if (MEM_64bits() || (HUF_MAX_TABLELOG<=12)) \
if (MEM_64bits() || (HUF_TABLELOG_MAX<=12)) \
ptr += HUF_decodeSymbolX4(ptr, DStreamPtr, dt, dtLog)
#define HUF_DECODE_SYMBOLX4_2(ptr, DStreamPtr) \
@ -580,7 +580,7 @@ size_t HUF_decompress1X4_usingDTable(
size_t HUF_decompress1X4 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX4(DTable, HUF_MAX_TABLELOG);
HUF_CREATE_STATIC_DTABLEX4(DTable, HUF_TABLELOG_MAX);
const BYTE* ip = (const BYTE*) cSrc;
size_t const hSize = HUF_readDTableX4 (DTable, cSrc, cSrcSize);
@ -688,7 +688,7 @@ size_t HUF_decompress4X4_usingDTable(
size_t HUF_decompress4X4 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX4(DTable, HUF_MAX_TABLELOG);
HUF_CREATE_STATIC_DTABLEX4(DTable, HUF_TABLELOG_MAX);
const BYTE* ip = (const BYTE*) cSrc;
size_t hSize = HUF_readDTableX4 (DTable, cSrc, cSrcSize);
@ -716,7 +716,7 @@ static void HUF_fillDTableX6LevelN(HUF_DDescX6* DDescription, HUF_DSeqX6* DSeque
const int scaleLog = nbBitsBaseline - sizeLog; /* note : targetLog >= (nbBitsBaseline-1), hence scaleLog <= 1 */
const int minBits = nbBitsBaseline - maxWeight;
const U32 level = DDesc.nbBytes;
U32 rankVal[HUF_ABSOLUTEMAX_TABLELOG + 1];
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1];
U32 symbolStartPos, s;
/* local rankVal, will be modified */
@ -766,20 +766,20 @@ static void HUF_fillDTableX6LevelN(HUF_DDescX6* DDescription, HUF_DSeqX6* DSeque
/* note : same preparation as X4 */
size_t HUF_readDTableX6 (U32* DTable, const void* src, size_t srcSize)
{
BYTE weightList[HUF_MAX_SYMBOL_VALUE + 1];
sortedSymbol_t sortedSymbol[HUF_MAX_SYMBOL_VALUE + 1];
U32 rankStats[HUF_ABSOLUTEMAX_TABLELOG + 1] = { 0 };
U32 rankStart0[HUF_ABSOLUTEMAX_TABLELOG + 2] = { 0 };
BYTE weightList[HUF_SYMBOLVALUE_MAX + 1];
sortedSymbol_t sortedSymbol[HUF_SYMBOLVALUE_MAX + 1];
U32 rankStats[HUF_TABLELOG_ABSOLUTEMAX + 1] = { 0 };
U32 rankStart0[HUF_TABLELOG_ABSOLUTEMAX + 2] = { 0 };
U32* const rankStart = rankStart0+1;
U32 tableLog, maxW, sizeOfSort, nbSymbols;
rankVal_t rankVal;
const U32 memLog = DTable[0];
size_t iSize;
if (memLog > HUF_ABSOLUTEMAX_TABLELOG) return ERROR(tableLog_tooLarge);
if (memLog > HUF_TABLELOG_ABSOLUTEMAX) return ERROR(tableLog_tooLarge);
//memset(weightList, 0, sizeof(weightList)); /* is not necessary, even though some analyzer complain ... */
iSize = HUF_readStats(weightList, HUF_MAX_SYMBOL_VALUE + 1, rankStats, &nbSymbols, &tableLog, src, srcSize);
iSize = HUF_readStats(weightList, HUF_SYMBOLVALUE_MAX + 1, rankStats, &nbSymbols, &tableLog, src, srcSize);
if (HUF_isError(iSize)) return iSize;
/* check result */
@ -838,7 +838,7 @@ size_t HUF_readDTableX6 (U32* DTable, const void* src, size_t srcSize)
DDesc.nbBits = 0;
DDesc.nbBytes = 0;
HUF_fillDTableX6LevelN(DDescription, DSequence, memLog,
(const U32 (*)[HUF_ABSOLUTEMAX_TABLELOG + 1])rankVal, 0, 1, maxW,
(const U32 (*)[HUF_TABLELOG_ABSOLUTEMAX + 1])rankVal, 0, 1, maxW,
sortedSymbol, sizeOfSort, rankStart0,
tableLog+1, DSeq, DDesc);
}
@ -879,7 +879,7 @@ static U32 HUF_decodeLastSymbolsX6(void* op, U32 const maxL, BIT_DStream_t* DStr
ptr += HUF_decodeSymbolX6(ptr, DStreamPtr, dd, ds, dtLog)
#define HUF_DECODE_SYMBOLX6_1(ptr, DStreamPtr) \
if (MEM_64bits() || (HUF_MAX_TABLELOG<=12)) \
if (MEM_64bits() || (HUF_TABLELOG_MAX<=12)) \
HUF_DECODE_SYMBOLX6_0(ptr, DStreamPtr)
#define HUF_DECODE_SYMBOLX6_2(ptr, DStreamPtr) \
@ -939,7 +939,7 @@ size_t HUF_decompress1X6_usingDTable(
size_t HUF_decompress1X6 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX6(DTable, HUF_MAX_TABLELOG);
HUF_CREATE_STATIC_DTABLEX6(DTable, HUF_TABLELOG_MAX);
const BYTE* ip = (const BYTE*) cSrc;
size_t const hSize = HUF_readDTableX6 (DTable, cSrc, cSrcSize);
@ -1068,7 +1068,7 @@ size_t HUF_decompress4X6_usingDTable(
size_t HUF_decompress4X6 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX6(DTable, HUF_MAX_TABLELOG);
HUF_CREATE_STATIC_DTABLEX6(DTable, HUF_TABLELOG_MAX);
const BYTE* ip = (const BYTE*) cSrc;
size_t const hSize = HUF_readDTableX6 (DTable, cSrc, cSrcSize);