comparison contrib/python-zstandard/zstd/decompress/zstd_decompress.c @ 40121:73fef626dae3

zstandard: vendor python-zstandard 0.10.1

This was just released.

The upstream source distribution from PyPI was extracted. Unwanted
files were removed.

The clang-format ignore list was updated to reflect the new source
of files.

setup.py was updated to pass a new argument to python-zstandard's
function for returning an Extension instance. Upstream had to change
to use relative paths because Python 3.7's packaging doesn't
seem to like absolute paths when defining sources, includes, etc.
The default relative path calculation is relative to setup_zstd.py
which is different from the directory of Mercurial's setup.py.

The project contains a vendored copy of zstandard 1.3.6. The old
version was 1.3.4.

The API should be backwards compatible and nothing in core should
need adjusted. However, there is a new "chunker" API that we
may find useful in places where we want to emit compressed chunks
of a fixed size.

There are a pair of bug fixes in 0.10.0 with regards to
compressobj() and decompressobj() when block flushing is used. I
actually found these bugs when introducing these APIs in Mercurial!
But existing Mercurial code is not affected because we don't
perform block flushing.

# no-check-commit because 3rd party code has different style guidelines

Differential Revision: https://phab.mercurial-scm.org/D4911

40120:89742f1fa6cb

 */
#ifndef ZSTD_MAXWINDOWSIZE_DEFAULT
#  define ZSTD_MAXWINDOWSIZE_DEFAULT (((U32)1 << ZSTD_WINDOWLOG_DEFAULTMAX) + 1)
#endif


/*-*******************************************************
*  Dependencies
*********************************************************/
#include <string.h>      /* memcpy, memmove, memset */
#include "cpu.h"
#include "mem.h"         /* low level memory routines */
#define FSE_STATIC_LINKING_ONLY
#include "fse.h"
#define HUF_STATIC_LINKING_ONLY
#include "huf.h"
#include "zstd_internal.h"

#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
#  include "zstd_legacy.h"
#endif


/*-*************************************
*  Errors
***************************************/

} ZSTD_seqSymbol;

#define SEQSYMBOL_TABLE_SIZE(log)   (1 + (1 << (log)))

typedef struct {
    ZSTD_seqSymbol LLTable[SEQSYMBOL_TABLE_SIZE(LLFSELog)];
    ZSTD_seqSymbol OFTable[SEQSYMBOL_TABLE_SIZE(OffFSELog)];
    ZSTD_seqSymbol MLTable[SEQSYMBOL_TABLE_SIZE(MLFSELog)];
    HUF_DTable hufTable[HUF_DTABLE_SIZE(HufLog)];  /* can accommodate HUF_decompress4X */
    U32 workspace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
    U32 rep[ZSTD_REP_NUM];
} ZSTD_entropyDTables_t;

struct ZSTD_DCtx_s
{
    const ZSTD_seqSymbol* LLTptr;
    const ZSTD_seqSymbol* MLTptr;
    const ZSTD_seqSymbol* OFTptr;
    const HUF_DTable* HUFptr;
    ZSTD_entropyDTables_t entropy;
    const void* previousDstEnd;   /* detect continuity */
    const void* base;             /* start of current segment */
    const void* vBase;            /* virtual start of previous segment if it was just before current one */
    const void* dictEnd;          /* end of previous segment */
    size_t expected;
    ZSTD_frameHeader fParams;
    U64 decodedSize;
    blockType_e bType;            /* used in ZSTD_decompressContinue(), store blockType between block header decoding and block decompression stages */
    ZSTD_dStage stage;
    U32 litEntropy;
    U32 fseEntropy;
    XXH64_state_t xxhState;
    size_t headerSize;
    U32 dictID;
    ZSTD_format_e format;
    const BYTE* litPtr;
    ZSTD_customMem customMem;
    size_t litSize;
    size_t rleSize;
    size_t staticSize;
    int bmi2;                     /* == 1 if the CPU supports BMI2 and 0 otherwise. CPU support is determined dynamically once per context lifetime. */

    /* streaming */
    ZSTD_DDict* ddictLocal;
    const ZSTD_DDict* ddict;
    ZSTD_dStreamStage streamStage;
    char*  inBuff;
    size_t inBuffSize;
    size_t inPos;
    size_t maxWindowSize;

    size_t lhSize;
    void* legacyContext;
    U32 previousLegacyVersion;
    U32 legacyVersion;
    U32 hostageByte;

    /* workspace */
    BYTE litBuffer[ZSTD_BLOCKSIZE_MAX + WILDCOPY_OVERLENGTH];
    BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX];
};  /* typedef'd to ZSTD_DCtx within "zstd.h" */

static size_t ZSTD_startingInputLength(ZSTD_format_e format)
{
    size_t const startingInputLength = (format==ZSTD_f_zstd1_magicless) ?
                    ZSTD_frameHeaderSize_prefix - ZSTD_frameIdSize :
                    ZSTD_frameHeaderSize_prefix;
    ZSTD_STATIC_ASSERT(ZSTD_FRAMEHEADERSIZE_PREFIX >= ZSTD_FRAMEIDSIZE);
    /* only supports formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless */
    assert( (format == ZSTD_f_zstd1) || (format == ZSTD_f_zstd1_magicless) );
    return startingInputLength;

    dctx->format = ZSTD_f_zstd1;  /* ZSTD_decompressBegin() invokes ZSTD_startingInputLength() with argument dctx->format */
    dctx->staticSize  = 0;
    dctx->maxWindowSize = ZSTD_MAXWINDOWSIZE_DEFAULT;
    dctx->ddict       = NULL;
    dctx->ddictLocal  = NULL;
    dctx->inBuff      = NULL;
    dctx->inBuffSize  = 0;
    dctx->outBuffSize = 0;
    dctx->streamStage = zdss_init;
    dctx->bmi2 = ZSTD_cpuid_bmi2(ZSTD_cpuid());
}

ZSTD_DCtx* ZSTD_initStaticDCtx(void *workspace, size_t workspaceSize)
{

    if (!customMem.customAlloc ^ !customMem.customFree) return NULL;

    {   ZSTD_DCtx* const dctx = (ZSTD_DCtx*)ZSTD_malloc(sizeof(*dctx), customMem);
        if (!dctx) return NULL;
        dctx->customMem = customMem;
        dctx->legacyContext = NULL;
        dctx->previousLegacyVersion = 0;
        ZSTD_initDCtx_internal(dctx);
        return dctx;
    }
}

 *  Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0.
 *  Note 2 : Legacy Frame Identifiers are considered valid only if Legacy Support is enabled.
 *  Note 3 : Skippable Frame Identifiers are considered valid. */
unsigned ZSTD_isFrame(const void* buffer, size_t size)
{
    if (size < ZSTD_frameIdSize) return 0;
    {   U32 const magic = MEM_readLE32(buffer);
        if (magic == ZSTD_MAGICNUMBER) return 1;
        if ((magic & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) return 1;
    }
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)

    }
}

/** ZSTD_frameHeaderSize() :
 *  srcSize must be >= ZSTD_frameHeaderSize_prefix.
 * @return : size of the Frame Header */
size_t ZSTD_frameHeaderSize(const void* src, size_t srcSize)
{
    return ZSTD_frameHeaderSize_internal(src, srcSize, ZSTD_f_zstd1);
}


/** ZSTD_getFrameHeader_internal() :
 *  decode Frame Header, or require larger `srcSize`.
 *  note : only works for formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless
 * @return : 0, `zfhPtr` is correctly filled,
 *          >0, `srcSize` is too small, value is wanted `srcSize` amount,
 *           or an error code, which can be tested using ZSTD_isError() */
static size_t ZSTD_getFrameHeader_internal(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format)
{
    const BYTE* ip = (const BYTE*)src;
    size_t const minInputSize = ZSTD_startingInputLength(format);

    if (srcSize < minInputSize) return minInputSize;

    if ( (format != ZSTD_f_zstd1_magicless)
      && (MEM_readLE32(src) != ZSTD_MAGICNUMBER) ) {
        if ((MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
            /* skippable frame */
            if (srcSize < ZSTD_skippableHeaderSize)
                return ZSTD_skippableHeaderSize; /* magic number + frame length */
            memset(zfhPtr, 0, sizeof(*zfhPtr));
            zfhPtr->frameContentSize = MEM_readLE32((const char *)src + ZSTD_frameIdSize);
            zfhPtr->frameType = ZSTD_skippableFrame;
            return 0;
        }
        return ERROR(prefix_unknown);
    }

 * @return : 0, `zfhPtr` is correctly filled,
 *          >0, `srcSize` is too small, value is wanted `srcSize` amount,
 *           or an error code, which can be tested using ZSTD_isError() */
size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize)
{
    return ZSTD_getFrameHeader_internal(zfhPtr, src, srcSize, ZSTD_f_zstd1);
}


/** ZSTD_getFrameContentSize() :
 *  compatible with legacy mode

        if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
            size_t skippableSize;
            if (srcSize < ZSTD_skippableHeaderSize)
                return ERROR(srcSize_wrong);
            skippableSize = MEM_readLE32((const BYTE *)src + ZSTD_frameIdSize)
                          + ZSTD_skippableHeaderSize;
            if (srcSize < skippableSize) {
                return ZSTD_CONTENTSIZE_ERROR;
            }

/** ZSTD_decodeFrameHeader() :
*   `headerSize` must be the size provided by ZSTD_frameHeaderSize().
*   @return : 0 if success, or an error code, which can be tested using ZSTD_isError() */
static size_t ZSTD_decodeFrameHeader(ZSTD_DCtx* dctx, const void* src, size_t headerSize)
{
    size_t const result = ZSTD_getFrameHeader_internal(&(dctx->fParams), src, headerSize, dctx->format);
    if (ZSTD_isError(result)) return result;    /* invalid header */
    if (result>0) return ERROR(srcSize_wrong);  /* headerSize too small */
    if (dctx->fParams.dictID && (dctx->dictID != dctx->fParams.dictID))
        return ERROR(dictionary_wrong);
    if (dctx->fParams.checksumFlag) XXH64_reset(&dctx->xxhState, 0);

static size_t ZSTD_copyRawBlock(void* dst, size_t dstCapacity,
                          const void* src, size_t srcSize)
{
    if (srcSize > dstCapacity) return ERROR(dstSize_tooSmall);
    memcpy(dst, src, srcSize);
    return srcSize;
}

    if (regenSize > dstCapacity) return ERROR(dstSize_tooSmall);
    memset(dst, *(const BYTE*)src, regenSize);
    return regenSize;
}

/*! ZSTD_decodeLiteralsBlock() :
 * @return : nb of bytes read from src (< srcSize )
 *  note : symbol not declared but exposed for fullbench */
size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx* dctx,
                          const void* src, size_t srcSize)   /* note : srcSize < BLOCKSIZE */

        switch(litEncType)
        {
        case set_repeat:
            if (dctx->litEntropy==0) return ERROR(dictionary_corrupted);
            /* fall-through */
        case set_compressed:
            if (srcSize < 5) return ERROR(corruption_detected);   /* srcSize >= MIN_CBLOCK_SIZE == 3; here we need up to 5 for case 3 */
            {   size_t lhSize, litSize, litCSize;
                U32 singleStream=0;
                U32 const lhlCode = (istart[0] >> 2) & 3;

                    break;
                }
                if (litSize > ZSTD_BLOCKSIZE_MAX) return ERROR(corruption_detected);
                if (litCSize + lhSize > srcSize) return ERROR(corruption_detected);

                if (HUF_isError((litEncType==set_repeat) ?
                                    ( singleStream ?
                                        HUF_decompress1X_usingDTable_bmi2(dctx->litBuffer, litSize, istart+lhSize, litCSize, dctx->HUFptr, dctx->bmi2) :
                                        HUF_decompress4X_usingDTable_bmi2(dctx->litBuffer, litSize, istart+lhSize, litCSize, dctx->HUFptr, dctx->bmi2) ) :
                                    ( singleStream ?
                                        HUF_decompress1X2_DCtx_wksp_bmi2(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart+lhSize, litCSize,
                                                                         dctx->entropy.workspace, sizeof(dctx->entropy.workspace), dctx->bmi2) :
                                        HUF_decompress4X_hufOnly_wksp_bmi2(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart+lhSize, litCSize,
                                                                           dctx->entropy.workspace, sizeof(dctx->entropy.workspace), dctx->bmi2))))
                    return ERROR(corruption_detected);

                dctx->litPtr = dctx->litBuffer;
                dctx->litSize = litSize;
                dctx->litEntropy = 1;

 *           or an error code if it fails */
static size_t ZSTD_buildSeqTable(ZSTD_seqSymbol* DTableSpace, const ZSTD_seqSymbol** DTablePtr,
                                 symbolEncodingType_e type, U32 max, U32 maxLog,
                                 const void* src, size_t srcSize,
                                 const U32* baseValue, const U32* nbAdditionalBits,
                                 const ZSTD_seqSymbol* defaultTable, U32 flagRepeatTable)
{
    switch(type)
    {
    case set_rle :
        if (!srcSize) return ERROR(srcSize_wrong);

    case set_basic :
        *DTablePtr = defaultTable;
        return 0;
    case set_repeat:
        if (!flagRepeatTable) return ERROR(corruption_detected);
        return 0;
    case set_compressed :
        {   U32 tableLog;
            S16 norm[MaxSeq+1];
            size_t const headerSize = FSE_readNCount(norm, &max, &tableLog, src, srcSize);

                    27, 28, 29,   30,    31,    32,    33,    34,
                    35, 37, 39,   41,    43,    47,    51,    59,
                    67, 83, 99, 0x83, 0x103, 0x203, 0x403, 0x803,
                    0x1003, 0x2003, 0x4003, 0x8003, 0x10003 };


size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx* dctx, int* nbSeqPtr,
                             const void* src, size_t srcSize)
{
    const BYTE* const istart = (const BYTE* const)src;
    const BYTE* const iend = istart + srcSize;
    const BYTE* ip = istart;
    DEBUGLOG(5, "ZSTD_decodeSeqHeaders");

    /* check */
    if (srcSize < MIN_SEQUENCES_SIZE) return ERROR(srcSize_wrong);

    /* SeqHead */
    {   int nbSeq = *ip++;
        if (!nbSeq) { *nbSeqPtr=0; return 1; }
        if (nbSeq > 0x7F) {
            if (nbSeq == 0xFF) {
                if (ip+2 > iend) return ERROR(srcSize_wrong);
                nbSeq = MEM_readLE16(ip) + LONGNBSEQ, ip+=2;
            } else {
                if (ip >= iend) return ERROR(srcSize_wrong);
                nbSeq = ((nbSeq-0x80)<<8) + *ip++;
            }
        }
        *nbSeqPtr = nbSeq;
    }

    /* FSE table descriptors */
    if (ip+4 > iend) return ERROR(srcSize_wrong); /* minimum possible size */
    {   symbolEncodingType_e const LLtype = (symbolEncodingType_e)(*ip >> 6);
        symbolEncodingType_e const OFtype = (symbolEncodingType_e)((*ip >> 4) & 3);

        /* Build DTables */
        {   size_t const llhSize = ZSTD_buildSeqTable(dctx->entropy.LLTable, &dctx->LLTptr,
                                                      LLtype, MaxLL, LLFSELog,
                                                      ip, iend-ip,
                                                      LL_base, LL_bits,
                                                      LL_defaultDTable, dctx->fseEntropy);
            if (ZSTD_isError(llhSize)) return ERROR(corruption_detected);
            ip += llhSize;
        }

        {   size_t const ofhSize = ZSTD_buildSeqTable(dctx->entropy.OFTable, &dctx->OFTptr,
                                                      OFtype, MaxOff, OffFSELog,
                                                      ip, iend-ip,
                                                      OF_base, OF_bits,
                                                      OF_defaultDTable, dctx->fseEntropy);
            if (ZSTD_isError(ofhSize)) return ERROR(corruption_detected);
            ip += ofhSize;
        }

        {   size_t const mlhSize = ZSTD_buildSeqTable(dctx->entropy.MLTable, &dctx->MLTptr,
                                                      MLtype, MaxML, MLFSELog,
                                                      ip, iend-ip,
                                                      ML_base, ML_bits,
                                                      ML_defaultDTable, dctx->fseEntropy);
            if (ZSTD_isError(mlhSize)) return ERROR(corruption_detected);
            ip += mlhSize;
        }
    }

    return ip-istart;
}

HINT_INLINE
size_t ZSTD_execSequence(BYTE* op,
                         BYTE* const oend, seq_t sequence,
                         const BYTE** litPtr, const BYTE* const litLimit,
                         const BYTE* const base, const BYTE* const vBase, const BYTE* const dictEnd)
{
    BYTE* const oLitEnd = op + sequence.litLength;
    size_t const sequenceLength = sequence.litLength + sequence.matchLength;
    BYTE* const oMatchEnd = op + sequenceLength;   /* risk : address space overflow (32-bits) */
    BYTE* const oend_w = oend - WILDCOPY_OVERLENGTH;

    const BYTE* match = oLitEnd - sequence.offset;

    /* check */
    if (oMatchEnd>oend) return ERROR(dstSize_tooSmall); /* last match must start at a minimum distance of WILDCOPY_OVERLENGTH from oend */
    if (iLitEnd > litLimit) return ERROR(corruption_detected);   /* over-read beyond lit buffer */
    if (oLitEnd>oend_w) return ZSTD_execSequenceLast7(op, oend, sequence, litPtr, litLimit, base, vBase, dictEnd);

    /* copy Literals */
    ZSTD_copy8(op, *litPtr);
    if (sequence.litLength > 8)
        ZSTD_wildcopy(op+8, (*litPtr)+8, sequence.litLength - 8);   /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */
    op = oLitEnd;
    *litPtr = iLitEnd;   /* update for next sequence */

    /* copy Match */
    if (sequence.offset > (size_t)(oLitEnd - base)) {
        /* offset beyond prefix -> go into extDict */
        if (sequence.offset > (size_t)(oLitEnd - vBase))
            return ERROR(corruption_detected);
        match = dictEnd + (match - base);
        if (match + sequence.matchLength <= dictEnd) {
            memmove(oLitEnd, match, sequence.matchLength);
            return sequenceLength;
        }
        /* span extDict & currentPrefixSegment */
        {   size_t const length1 = dictEnd - match;
            memmove(oLitEnd, match, length1);
            op = oLitEnd + length1;
            sequence.matchLength -= length1;
            match = base;
            if (op > oend_w || sequence.matchLength < MINMATCH) {
              U32 i;
              for (i = 0; i < sequence.matchLength; ++i) op[i] = match[i];
              return sequenceLength;
            }

    BYTE* const ostart = (BYTE* const)dst;
    BYTE* const oend = ostart + maxDstSize;
    BYTE* op = ostart;
    const BYTE* litPtr = dctx->litPtr;
    const BYTE* const litEnd = litPtr + dctx->litSize;
    const BYTE* const base = (const BYTE*) (dctx->base);
    const BYTE* const vBase = (const BYTE*) (dctx->vBase);
    const BYTE* const dictEnd = (const BYTE*) (dctx->dictEnd);
    DEBUGLOG(5, "ZSTD_decompressSequences");

    /* Regen sequences */
    if (nbSeq) {
        seqState_t seqState;
        dctx->fseEntropy = 1;

        ZSTD_initFseState(&seqState.stateML, &seqState.DStream, dctx->MLTptr);

        for ( ; (BIT_reloadDStream(&(seqState.DStream)) <= BIT_DStream_completed) && nbSeq ; ) {
            nbSeq--;
            {   seq_t const sequence = ZSTD_decodeSequence(&seqState, isLongOffset);
                size_t const oneSeqSize = ZSTD_execSequence(op, oend, sequence, &litPtr, litEnd, base, vBase, dictEnd);
                DEBUGLOG(6, "regenerated sequence size : %u", (U32)oneSeqSize);
                if (ZSTD_isError(oneSeqSize)) return oneSeqSize;
                op += oneSeqSize;
        }   }

        /* check if reached exact end */
        DEBUGLOG(5, "ZSTD_decompressSequences: after decode loop, remaining nbSeq : %i", nbSeq);
        if (nbSeq) return ERROR(corruption_detected);
        /* save reps for next block */
        { U32 i; for (i=0; i<ZSTD_REP_NUM; i++) dctx->entropy.rep[i] = (U32)(seqState.prevOffset[i]); }
    }

    BYTE* const ostart = (BYTE* const)dst;
    BYTE* const oend = ostart + maxDstSize;
    BYTE* op = ostart;
    const BYTE* litPtr = dctx->litPtr;
    const BYTE* const litEnd = litPtr + dctx->litSize;
    const BYTE* const prefixStart = (const BYTE*) (dctx->base);
    const BYTE* const dictStart = (const BYTE*) (dctx->vBase);
    const BYTE* const dictEnd = (const BYTE*) (dctx->dictEnd);

    /* Regen sequences */
    if (nbSeq) {
#define STORED_SEQS 4

{   /* blockType == blockCompressed */
    const BYTE* ip = (const BYTE*)src;
    /* isLongOffset must be true if there are long offsets.
     * Offsets are long if they are larger than 2^STREAM_ACCUMULATOR_MIN.
     * We don't expect that to be the case in 64-bit mode.
     * In block mode, window size is not known, so we have to be conservative. (note: but it could be evaluated from current-lowLimit)
     */
    ZSTD_longOffset_e const isLongOffset = (ZSTD_longOffset_e)(MEM_32bits() && (!frame || dctx->fParams.windowSize > (1ULL << STREAM_ACCUMULATOR_MIN)));
    DEBUGLOG(5, "ZSTD_decompressBlock_internal (size : %u)", (U32)srcSize);

    if (srcSize >= ZSTD_BLOCKSIZE_MAX) return ERROR(srcSize_wrong);

static void ZSTD_checkContinuity(ZSTD_DCtx* dctx, const void* dst)
{
    if (dst != dctx->previousDstEnd) {   /* not contiguous */
        dctx->dictEnd = dctx->previousDstEnd;
        dctx->vBase = (const char*)dst - ((const char*)(dctx->previousDstEnd) - (const char*)(dctx->base));
        dctx->base = dst;
        dctx->previousDstEnd = dst;
    }
}

size_t ZSTD_decompressBlock(ZSTD_DCtx* dctx,

    dctx->previousDstEnd = (const char*)blockStart + blockSize;
    return blockSize;
}


static size_t ZSTD_generateNxBytes(void* dst, size_t dstCapacity, BYTE byte, size_t length)
{
    if (length > dstCapacity) return ERROR(dstSize_tooSmall);
    memset(dst, byte, length);
    return length;
}

/** ZSTD_findFrameCompressedSize() :
 *  compatible with legacy mode

    if (ZSTD_isLegacy(src, srcSize))
        return ZSTD_findFrameCompressedSizeLegacy(src, srcSize);
#endif
    if ( (srcSize >= ZSTD_skippableHeaderSize)
      && (MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START ) {
        return ZSTD_skippableHeaderSize + MEM_readLE32((const BYTE*)src + ZSTD_frameIdSize);
    } else {
        const BYTE* ip = (const BYTE*)src;
        const BYTE* const ipstart = ip;
        size_t remainingSize = srcSize;
        ZSTD_frameHeader zfh;

        }

        if (zfh.checksumFlag) {   /* Final frame content checksum */
            if (remainingSize < 4) return ERROR(srcSize_wrong);
            ip += 4;
            remainingSize -= 4;
        }

        return ip - ipstart;
    }
}

    *srcPtr = ip;
    *srcSizePtr = remainingSize;
    return op-ostart;
}

static const void* ZSTD_DDictDictContent(const ZSTD_DDict* ddict);
static size_t ZSTD_DDictDictSize(const ZSTD_DDict* ddict);

static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx,
                                        void* dst, size_t dstCapacity,
                                  const void* src, size_t srcSize,
                                  const void* dict, size_t dictSize,
                                  const ZSTD_DDict* ddict)
{
    void* const dststart = dst;
    assert(dict==NULL || ddict==NULL);  /* either dict or ddict set, not both */

    if (ddict) {
        dict = ZSTD_DDictDictContent(ddict);
        dictSize = ZSTD_DDictDictSize(ddict);
    }

    while (srcSize >= ZSTD_frameHeaderSize_prefix) {
        U32 magicNumber;

#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
        if (ZSTD_isLegacy(src, srcSize)) {
            size_t decodedSize;
            size_t const frameSize = ZSTD_findFrameCompressedSizeLegacy(src, srcSize);

            continue;
        }
#endif

        magicNumber = MEM_readLE32(src);
        DEBUGLOG(4, "reading magic number %08X (expecting %08X)",
                    (U32)magicNumber, (U32)ZSTD_MAGICNUMBER);
        if (magicNumber != ZSTD_MAGICNUMBER) {
            if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
                size_t skippableSize;
                if (srcSize < ZSTD_skippableHeaderSize)
                    return ERROR(srcSize_wrong);
                skippableSize = MEM_readLE32((const BYTE*)src + ZSTD_frameIdSize)
                              + ZSTD_skippableHeaderSize;
                if (srcSize < skippableSize) return ERROR(srcSize_wrong);

                src = (const BYTE *)src + skippableSize;
                srcSize -= skippableSize;
                continue;
            }
            return ERROR(prefix_unknown);
        }

        if (ddict) {
            /* we were called from ZSTD_decompress_usingDDict */
            CHECK_F(ZSTD_decompressBegin_usingDDict(dctx, ddict));
        } else {

        }
        ZSTD_checkContinuity(dctx, dst);

        {   const size_t res = ZSTD_decompressFrame(dctx, dst, dstCapacity,
                                                    &src, &srcSize);
            if (ZSTD_isError(res)) return res;
            /* no need to bound check, ZSTD_decompressFrame already has */
            dst = (BYTE*)dst + res;
            dstCapacity -= res;
        }
    }  /* while (srcSize >= ZSTD_frameHeaderSize_prefix) */

    if (srcSize) return ERROR(srcSize_wrong); /* input not entirely consumed */

    return (BYTE*)dst - (BYTE*)dststart;

    regenSize = ZSTD_decompressDCtx(dctx, dst, dstCapacity, src, srcSize);
    ZSTD_freeDCtx(dctx);
    return regenSize;
#else   /* stack mode */
    ZSTD_DCtx dctx;
    return ZSTD_decompressDCtx(&dctx, dst, dstCapacity, src, srcSize);
#endif
}

    switch (dctx->stage)
    {
    case ZSTDds_getFrameHeaderSize :
        assert(src != NULL);
        if (dctx->format == ZSTD_f_zstd1) {  /* allows header */
            assert(srcSize >= ZSTD_frameIdSize);  /* to read skippable magic number */
            if ((MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {        /* skippable frame */
                memcpy(dctx->headerBuffer, src, srcSize);
                dctx->expected = ZSTD_skippableHeaderSize - srcSize;  /* remaining to load to get full skippable frame header */
                dctx->stage = ZSTDds_decodeSkippableHeader;
                return 0;

    case ZSTDds_decodeSkippableHeader:
        assert(src != NULL);
        assert(srcSize <= ZSTD_skippableHeaderSize);
        memcpy(dctx->headerBuffer + (ZSTD_skippableHeaderSize - srcSize), src, srcSize);   /* complete skippable header */
        dctx->expected = MEM_readLE32(dctx->headerBuffer + ZSTD_frameIdSize);   /* note : dctx->expected can grow seriously large, beyond local buffer size */
        dctx->stage = ZSTDds_skipFrame;
        return 0;

    case ZSTDds_skipFrame:
        dctx->expected = 0;

static size_t ZSTD_refDictContent(ZSTD_DCtx* dctx, const void* dict, size_t dictSize)
{
    dctx->dictEnd = dctx->previousDstEnd;
    dctx->vBase = (const char*)dict - ((const char*)(dctx->previousDstEnd) - (const char*)(dctx->base));
    dctx->base = dict;
    dctx->previousDstEnd = (const char*)dict + dictSize;
    return 0;
}

/* ZSTD_loadEntropy() :
 * dict : must point at beginning of a valid zstd dictionary
 * @return : size of entropy tables read */
static size_t ZSTD_loadEntropy(ZSTD_entropyDTables_t* entropy, const void* const dict, size_t const dictSize)
{
    const BYTE* dictPtr = (const BYTE*)dict;
    const BYTE* const dictEnd = dictPtr + dictSize;

    if (dictSize <= 8) return ERROR(dictionary_corrupted);
    dictPtr += 8;   /* skip header = magic + dictID */


    {   size_t const hSize = HUF_readDTableX4_wksp(
            entropy->hufTable, dictPtr, dictEnd - dictPtr,
            entropy->workspace, sizeof(entropy->workspace));
        if (HUF_isError(hSize)) return ERROR(dictionary_corrupted);
        dictPtr += hSize;
    }

    {   short offcodeNCount[MaxOff+1];
        U32 offcodeMaxValue = MaxOff, offcodeLog;
        size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, dictEnd-dictPtr);
        if (FSE_isError(offcodeHeaderSize)) return ERROR(dictionary_corrupted);
        if (offcodeMaxValue > MaxOff) return ERROR(dictionary_corrupted);
        if (offcodeLog > OffFSELog) return ERROR(dictionary_corrupted);
        ZSTD_buildFSETable(entropy->OFTable,
                            offcodeNCount, offcodeMaxValue,
                            OF_base, OF_bits,
                            offcodeLog);
        dictPtr += offcodeHeaderSize;
    }

        unsigned matchlengthMaxValue = MaxML, matchlengthLog;
        size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, dictEnd-dictPtr);
        if (FSE_isError(matchlengthHeaderSize)) return ERROR(dictionary_corrupted);
        if (matchlengthMaxValue > MaxML) return ERROR(dictionary_corrupted);
        if (matchlengthLog > MLFSELog) return ERROR(dictionary_corrupted);
        ZSTD_buildFSETable(entropy->MLTable,
                            matchlengthNCount, matchlengthMaxValue,
                            ML_base, ML_bits,
                            matchlengthLog);
        dictPtr += matchlengthHeaderSize;
    }

        unsigned litlengthMaxValue = MaxLL, litlengthLog;
        size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, dictEnd-dictPtr);
        if (FSE_isError(litlengthHeaderSize)) return ERROR(dictionary_corrupted);
        if (litlengthMaxValue > MaxLL) return ERROR(dictionary_corrupted);
        if (litlengthLog > LLFSELog) return ERROR(dictionary_corrupted);
        ZSTD_buildFSETable(entropy->LLTable,
                            litlengthNCount, litlengthMaxValue,
                            LL_base, LL_bits,
                            litlengthLog);
        dictPtr += litlengthHeaderSize;
    }

    if (dictSize < 8) return ZSTD_refDictContent(dctx, dict, dictSize);
    {   U32 const magic = MEM_readLE32(dict);
        if (magic != ZSTD_MAGIC_DICTIONARY) {
            return ZSTD_refDictContent(dctx, dict, dictSize);   /* pure content mode */
    }   }
    dctx->dictID = MEM_readLE32((const char*)dict + ZSTD_frameIdSize);

    /* load entropy tables */
    {   size_t const eSize = ZSTD_loadEntropy(&dctx->entropy, dict, dictSize);
        if (ZSTD_isError(eSize)) return ERROR(dictionary_corrupted);
        dict = (const char*)dict + eSize;

    /* reference dictionary content */
    return ZSTD_refDictContent(dctx, dict, dictSize);
}

/* Note : this function cannot fail */
size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx)
{
    assert(dctx != NULL);
    dctx->expected = ZSTD_startingInputLength(dctx->format);  /* dctx->format must be properly set */
    dctx->stage = ZSTDds_getFrameHeaderSize;
    dctx->decodedSize = 0;
    dctx->previousDstEnd = NULL;
    dctx->base = NULL;
    dctx->vBase = NULL;
    dctx->dictEnd = NULL;
    dctx->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001);  /* cover both little and big endian */
    dctx->litEntropy = dctx->fseEntropy = 0;
    dctx->dictID = 0;
    ZSTD_STATIC_ASSERT(sizeof(dctx->entropy.rep) == sizeof(repStartValue));

    ZSTD_customMem cMem;
};  /* typedef'd to ZSTD_DDict within "zstd.h" */

static const void* ZSTD_DDictDictContent(const ZSTD_DDict* ddict)
{
    return ddict->dictContent;
}

static size_t ZSTD_DDictDictSize(const ZSTD_DDict* ddict)
{
    return ddict->dictSize;
}

size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dstDCtx, const ZSTD_DDict* ddict)
{
    CHECK_F( ZSTD_decompressBegin(dstDCtx) );
    if (ddict) {   /* support begin on NULL */
        dstDCtx->dictID = ddict->dictID;
        dstDCtx->base = ddict->dictContent;
        dstDCtx->vBase = ddict->dictContent;
        dstDCtx->dictEnd = (const BYTE*)ddict->dictContent + ddict->dictSize;
        dstDCtx->previousDstEnd = dstDCtx->dictEnd;
        if (ddict->entropyPresent) {
            dstDCtx->litEntropy = 1;
            dstDCtx->fseEntropy = 1;
            dstDCtx->LLTptr = ddict->entropy.LLTable;
            dstDCtx->MLTptr = ddict->entropy.MLTable;
            dstDCtx->OFTptr = ddict->entropy.OFTable;
            dstDCtx->HUFptr = ddict->entropy.hufTable;
            dstDCtx->entropy.rep[0] = ddict->entropy.rep[0];
            dstDCtx->entropy.rep[1] = ddict->entropy.rep[1];
            dstDCtx->entropy.rep[2] = ddict->entropy.rep[2];
        } else {
            dstDCtx->litEntropy = 0;
            dstDCtx->fseEntropy = 0;
        }
    }
    return 0;
}

static size_t ZSTD_loadEntropy_inDDict(ZSTD_DDict* ddict, ZSTD_dictContentType_e dictContentType)
{
    ddict->dictID = 0;
    ddict->entropyPresent = 0;
    if (dictContentType == ZSTD_dct_rawContent) return 0;

            if (dictContentType == ZSTD_dct_fullDict)
                return ERROR(dictionary_corrupted);   /* only accept specified dictionaries */
            return 0;   /* pure content mode */
        }
    }
    ddict->dictID = MEM_readLE32((const char*)ddict->dictContent + ZSTD_frameIdSize);

    /* load entropy tables */
    CHECK_E( ZSTD_loadEntropy(&ddict->entropy, ddict->dictContent, ddict->dictSize), dictionary_corrupted );
    ddict->entropyPresent = 1;
    return 0;
}

                                      ZSTD_dictContentType_e dictContentType)
{
    if ((dictLoadMethod == ZSTD_dlm_byRef) || (!dict) || (!dictSize)) {
        ddict->dictBuffer = NULL;
        ddict->dictContent = dict;
    } else {
        void* const internalBuffer = ZSTD_malloc(dictSize, ddict->cMem);
        ddict->dictBuffer = internalBuffer;
        ddict->dictContent = internalBuffer;
        if (!internalBuffer) return ERROR(memory_allocation);

                                      ZSTD_customMem customMem)
{
    if (!customMem.customAlloc ^ !customMem.customFree) return NULL;

    {   ZSTD_DDict* const ddict = (ZSTD_DDict*) ZSTD_malloc(sizeof(ZSTD_DDict), customMem);
        if (!ddict) return NULL;
        ddict->cMem = customMem;

        if (ZSTD_isError( ZSTD_initDDict_internal(ddict, dict, dictSize, dictLoadMethod, dictContentType) )) {
            ZSTD_freeDDict(ddict);
            return NULL;
        }

        return ddict;
    }
}

/*! ZSTD_createDDict() :

    return ZSTD_createDDict_advanced(dictBuffer, dictSize, ZSTD_dlm_byRef, ZSTD_dct_auto, allocator);
}


const ZSTD_DDict* ZSTD_initStaticDDict(
                                void* workspace, size_t workspaceSize,
                                const void* dict, size_t dictSize,
                                ZSTD_dictLoadMethod_e dictLoadMethod,
                                ZSTD_dictContentType_e dictContentType)
{
    size_t const neededSpace =
            sizeof(ZSTD_DDict) + (dictLoadMethod == ZSTD_dlm_byRef ? 0 : dictSize);
    ZSTD_DDict* const ddict = (ZSTD_DDict*)workspace;
    assert(workspace != NULL);
    assert(dict != NULL);
    if ((size_t)workspace & 7) return NULL;  /* 8-aligned */
    if (workspaceSize < neededSpace) return NULL;
    if (dictLoadMethod == ZSTD_dlm_byCopy) {
        memcpy(ddict+1, dict, dictSize);  /* local copy */
        dict = ddict+1;
    }
    if (ZSTD_isError( ZSTD_initDDict_internal(ddict, dict, dictSize, ZSTD_dlm_byRef, dictContentType) ))
        return NULL;
    return ddict;
}

 *  It can still be loaded, but as a content-only dictionary. */
unsigned ZSTD_getDictID_fromDict(const void* dict, size_t dictSize)
{
    if (dictSize < 8) return 0;
    if (MEM_readLE32(dict) != ZSTD_MAGIC_DICTIONARY) return 0;
    return MEM_readLE32((const char*)dict + ZSTD_frameIdSize);
}

/*! ZSTD_getDictID_fromDDict() :
 *  Provides the dictID of the dictionary loaded into `ddict`.
 *  If @return == 0, the dictionary is not conformant to Zstandard specification, or empty.

{
    return ZSTD_freeDCtx(zds);
}


/* *** Initialization *** */

size_t ZSTD_DStreamInSize(void)  { return ZSTD_BLOCKSIZE_MAX + ZSTD_blockHeaderSize; }
size_t ZSTD_DStreamOutSize(void) { return ZSTD_BLOCKSIZE_MAX; }

size_t ZSTD_DCtx_loadDictionary_advanced(ZSTD_DCtx* dctx, const void* dict, size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod, ZSTD_dictContentType_e dictContentType)
{
    if (dctx->streamStage != zdss_init) return ERROR(stage_wrong);
    ZSTD_freeDDict(dctx->ddictLocal);
    if (dict && dictSize >= 8) {
        dctx->ddictLocal = ZSTD_createDDict_advanced(dict, dictSize, dictLoadMethod, dictContentType, dctx->customMem);

 * this function cannot fail */
size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize)
{
    DEBUGLOG(4, "ZSTD_initDStream_usingDict");
    zds->streamStage = zdss_init;
    CHECK_F( ZSTD_DCtx_loadDictionary(zds, dict, dictSize) );
    return ZSTD_frameHeaderSize_prefix;
}

/* note : this variant can't fail */
size_t ZSTD_initDStream(ZSTD_DStream* zds)
{
    DEBUGLOG(4, "ZSTD_initDStream");
    return ZSTD_initDStream_usingDict(zds, NULL, 0);
}

size_t ZSTD_DCtx_refDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict)
{
    if (dctx->streamStage != zdss_init) return ERROR(stage_wrong);
    dctx->ddict = ddict;
    return 0;
}

/* ZSTD_initDStream_usingDDict() :
 * ddict will just be referenced, and must outlive decompression session
 * this function cannot fail */

        case DStream_p_maxWindowSize :
            DEBUGLOG(4, "setting maxWindowSize = %u KB", paramValue >> 10);
            dctx->maxWindowSize = paramValue ? paramValue : (U32)(-1);
            break;
    }
    return 0;
}

size_t ZSTD_DCtx_setMaxWindowSize(ZSTD_DCtx* dctx, size_t maxWindowSize)
{

                {   size_t const hint = ZSTD_decompressLegacyStream(zds->legacyContext, zds->legacyVersion, output, input);
                    if (hint==0) zds->streamStage = zdss_init;
                    return hint;
            }   }
#endif
            {   size_t const hSize = ZSTD_getFrameHeader_internal(&zds->fParams, zds->headerBuffer, zds->lhSize, zds->format);
                DEBUGLOG(5, "header size : %u", (U32)hSize);
                if (ZSTD_isError(hSize)) {
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
                    U32 const legacyVersion = ZSTD_isLegacy(istart, iend-istart);
                    if (legacyVersion) {

            /* Consume header (see ZSTDds_decodeFrameHeader) */
            DEBUGLOG(4, "Consume header");
            CHECK_F(ZSTD_decompressBegin_usingDDict(zds, zds->ddict));

            if ((MEM_readLE32(zds->headerBuffer) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {  /* skippable frame */
                zds->expected = MEM_readLE32(zds->headerBuffer + ZSTD_frameIdSize);
                zds->stage = ZSTDds_skipFrame;
            } else {
                CHECK_F(ZSTD_decodeFrameHeader(zds, zds->headerBuffer, zds->lhSize));
                zds->expected = ZSTD_blockHeaderSize;
                zds->stage = ZSTDds_decodeBlockHeader;

        default: return ERROR(GENERIC);   /* impossible */
    }   }

    /* result */
    input->pos += (size_t)(ip-istart);
    output->pos += (size_t)(op-ostart);
    {   size_t nextSrcSizeHint = ZSTD_nextSrcSizeToDecompress(zds);
        if (!nextSrcSizeHint) {   /* frame fully decoded */
            if (zds->outEnd == zds->outStart) {  /* output fully flushed */
                if (zds->hostageByte) {
                    if (input->pos >= input->size) {

40121:73fef626dae3

 */
#ifndef ZSTD_MAXWINDOWSIZE_DEFAULT
#  define ZSTD_MAXWINDOWSIZE_DEFAULT (((U32)1 << ZSTD_WINDOWLOG_DEFAULTMAX) + 1)
#endif

/*!
 *  NO_FORWARD_PROGRESS_MAX :
 *  maximum allowed nb of calls to ZSTD_decompressStream() and ZSTD_decompress_generic()
 *  without any forward progress
 *  (defined as: no byte read from input, and no byte flushed to output)
 *  before triggering an error.
 */
#ifndef ZSTD_NO_FORWARD_PROGRESS_MAX
#  define ZSTD_NO_FORWARD_PROGRESS_MAX 16
#endif


/*-*******************************************************
*  Dependencies
*********************************************************/
#include <string.h>      /* memcpy, memmove, memset */
#include "compiler.h"    /* prefetch */
#include "cpu.h"         /* bmi2 */
#include "mem.h"         /* low level memory routines */
#define FSE_STATIC_LINKING_ONLY
#include "fse.h"
#define HUF_STATIC_LINKING_ONLY
#include "huf.h"
#include "zstd_internal.h"

#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
#  include "zstd_legacy.h"
#endif

static const void* ZSTD_DDictDictContent(const ZSTD_DDict* ddict);
static size_t ZSTD_DDictDictSize(const ZSTD_DDict* ddict);


/*-*************************************
*  Errors
***************************************/

} ZSTD_seqSymbol;

#define SEQSYMBOL_TABLE_SIZE(log)   (1 + (1 << (log)))

typedef struct {
    ZSTD_seqSymbol LLTable[SEQSYMBOL_TABLE_SIZE(LLFSELog)];    /* Note : Space reserved for FSE Tables */
    ZSTD_seqSymbol OFTable[SEQSYMBOL_TABLE_SIZE(OffFSELog)];   /* is also used as temporary workspace while building hufTable during DDict creation */
    ZSTD_seqSymbol MLTable[SEQSYMBOL_TABLE_SIZE(MLFSELog)];    /* and therefore must be at least HUF_DECOMPRESS_WORKSPACE_SIZE large */
    HUF_DTable hufTable[HUF_DTABLE_SIZE(HufLog)];  /* can accommodate HUF_decompress4X */
    U32 rep[ZSTD_REP_NUM];
} ZSTD_entropyDTables_t;

struct ZSTD_DCtx_s
{
    const ZSTD_seqSymbol* LLTptr;
    const ZSTD_seqSymbol* MLTptr;
    const ZSTD_seqSymbol* OFTptr;
    const HUF_DTable* HUFptr;
    ZSTD_entropyDTables_t entropy;
    U32 workspace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];   /* space needed when building huffman tables */
    const void* previousDstEnd;   /* detect continuity */
    const void* prefixStart;      /* start of current segment */
    const void* virtualStart;     /* virtual start of previous segment if it was just before current one */
    const void* dictEnd;          /* end of previous segment */
    size_t expected;
    ZSTD_frameHeader fParams;
    U64 decodedSize;
    blockType_e bType;            /* used in ZSTD_decompressContinue(), store blockType between block header decoding and block decompression stages */
    ZSTD_dStage stage;
    U32 litEntropy;
    U32 fseEntropy;
    XXH64_state_t xxhState;
    size_t headerSize;
    ZSTD_format_e format;
    const BYTE* litPtr;
    ZSTD_customMem customMem;
    size_t litSize;
    size_t rleSize;
    size_t staticSize;
    int bmi2;                     /* == 1 if the CPU supports BMI2 and 0 otherwise. CPU support is determined dynamically once per context lifetime. */

    /* dictionary */
    ZSTD_DDict* ddictLocal;
    const ZSTD_DDict* ddict;     /* set by ZSTD_initDStream_usingDDict(), or ZSTD_DCtx_refDDict() */
    U32 dictID;
    int ddictIsCold;             /* if == 1 : dictionary is "new" for working context, and presumed "cold" (not in cpu cache) */

    /* streaming */
    ZSTD_dStreamStage streamStage;
    char*  inBuff;
    size_t inBuffSize;
    size_t inPos;
    size_t maxWindowSize;

    size_t lhSize;
    void* legacyContext;
    U32 previousLegacyVersion;
    U32 legacyVersion;
    U32 hostageByte;
    int noForwardProgress;

    /* workspace */
    BYTE litBuffer[ZSTD_BLOCKSIZE_MAX + WILDCOPY_OVERLENGTH];
    BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX];
};  /* typedef'd to ZSTD_DCtx within "zstd.h" */

static size_t ZSTD_startingInputLength(ZSTD_format_e format)
{
    size_t const startingInputLength = (format==ZSTD_f_zstd1_magicless) ?
                    ZSTD_frameHeaderSize_prefix - ZSTD_FRAMEIDSIZE :
                    ZSTD_frameHeaderSize_prefix;
    ZSTD_STATIC_ASSERT(ZSTD_FRAMEHEADERSIZE_PREFIX >= ZSTD_FRAMEIDSIZE);
    /* only supports formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless */
    assert( (format == ZSTD_f_zstd1) || (format == ZSTD_f_zstd1_magicless) );
    return startingInputLength;

    dctx->format = ZSTD_f_zstd1;  /* ZSTD_decompressBegin() invokes ZSTD_startingInputLength() with argument dctx->format */
    dctx->staticSize  = 0;
    dctx->maxWindowSize = ZSTD_MAXWINDOWSIZE_DEFAULT;
    dctx->ddict       = NULL;
    dctx->ddictLocal  = NULL;
    dctx->dictEnd     = NULL;
    dctx->ddictIsCold = 0;
    dctx->inBuff      = NULL;
    dctx->inBuffSize  = 0;
    dctx->outBuffSize = 0;
    dctx->streamStage = zdss_init;
    dctx->legacyContext = NULL;
    dctx->previousLegacyVersion = 0;
    dctx->noForwardProgress = 0;
    dctx->bmi2 = ZSTD_cpuid_bmi2(ZSTD_cpuid());
}

ZSTD_DCtx* ZSTD_initStaticDCtx(void *workspace, size_t workspaceSize)
{

    if (!customMem.customAlloc ^ !customMem.customFree) return NULL;

    {   ZSTD_DCtx* const dctx = (ZSTD_DCtx*)ZSTD_malloc(sizeof(*dctx), customMem);
        if (!dctx) return NULL;
        dctx->customMem = customMem;
        ZSTD_initDCtx_internal(dctx);
        return dctx;
    }
}

 *  Note : Frame Identifier is 4 bytes. If `size < 4`, @return will always be 0.
 *  Note 2 : Legacy Frame Identifiers are considered valid only if Legacy Support is enabled.
 *  Note 3 : Skippable Frame Identifiers are considered valid. */
unsigned ZSTD_isFrame(const void* buffer, size_t size)
{
    if (size < ZSTD_FRAMEIDSIZE) return 0;
    {   U32 const magic = MEM_readLE32(buffer);
        if (magic == ZSTD_MAGICNUMBER) return 1;
        if ((magic & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) return 1;
    }
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)

    }
}

/** ZSTD_frameHeaderSize() :
 *  srcSize must be >= ZSTD_frameHeaderSize_prefix.
 * @return : size of the Frame Header,
 *           or an error code (if srcSize is too small) */
size_t ZSTD_frameHeaderSize(const void* src, size_t srcSize)
{
    return ZSTD_frameHeaderSize_internal(src, srcSize, ZSTD_f_zstd1);
}


/** ZSTD_getFrameHeader_advanced() :
 *  decode Frame Header, or require larger `srcSize`.
 *  note : only works for formats ZSTD_f_zstd1 and ZSTD_f_zstd1_magicless
 * @return : 0, `zfhPtr` is correctly filled,
 *          >0, `srcSize` is too small, value is wanted `srcSize` amount,
 *           or an error code, which can be tested using ZSTD_isError() */
size_t ZSTD_getFrameHeader_advanced(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize, ZSTD_format_e format)
{
    const BYTE* ip = (const BYTE*)src;
    size_t const minInputSize = ZSTD_startingInputLength(format);

    memset(zfhPtr, 0, sizeof(*zfhPtr));   /* not strictly necessary, but static analyzer do not understand that zfhPtr is only going to be read only if return value is zero, since they are 2 different signals */
    if (srcSize < minInputSize) return minInputSize;
    if (src==NULL) return ERROR(GENERIC);   /* invalid parameter */

    if ( (format != ZSTD_f_zstd1_magicless)
      && (MEM_readLE32(src) != ZSTD_MAGICNUMBER) ) {
        if ((MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
            /* skippable frame */
            if (srcSize < ZSTD_skippableHeaderSize)
                return ZSTD_skippableHeaderSize; /* magic number + frame length */
            memset(zfhPtr, 0, sizeof(*zfhPtr));
            zfhPtr->frameContentSize = MEM_readLE32((const char *)src + ZSTD_FRAMEIDSIZE);
            zfhPtr->frameType = ZSTD_skippableFrame;
            return 0;
        }
        return ERROR(prefix_unknown);
    }

 * @return : 0, `zfhPtr` is correctly filled,
 *          >0, `srcSize` is too small, value is wanted `srcSize` amount,
 *           or an error code, which can be tested using ZSTD_isError() */
size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize)
{
    return ZSTD_getFrameHeader_advanced(zfhPtr, src, srcSize, ZSTD_f_zstd1);
}


/** ZSTD_getFrameContentSize() :
 *  compatible with legacy mode

        if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
            size_t skippableSize;
            if (srcSize < ZSTD_skippableHeaderSize)
                return ERROR(srcSize_wrong);
            skippableSize = MEM_readLE32((const BYTE *)src + ZSTD_FRAMEIDSIZE)
                          + ZSTD_skippableHeaderSize;
            if (srcSize < skippableSize) {
                return ZSTD_CONTENTSIZE_ERROR;
            }

/** ZSTD_decodeFrameHeader() :
*   `headerSize` must be the size provided by ZSTD_frameHeaderSize().
*   @return : 0 if success, or an error code, which can be tested using ZSTD_isError() */
static size_t ZSTD_decodeFrameHeader(ZSTD_DCtx* dctx, const void* src, size_t headerSize)
{
    size_t const result = ZSTD_getFrameHeader_advanced(&(dctx->fParams), src, headerSize, dctx->format);
    if (ZSTD_isError(result)) return result;    /* invalid header */
    if (result>0) return ERROR(srcSize_wrong);  /* headerSize too small */
    if (dctx->fParams.dictID && (dctx->dictID != dctx->fParams.dictID))
        return ERROR(dictionary_wrong);
    if (dctx->fParams.checksumFlag) XXH64_reset(&dctx->xxhState, 0);

static size_t ZSTD_copyRawBlock(void* dst, size_t dstCapacity,
                          const void* src, size_t srcSize)
{
    if (dst==NULL) return ERROR(dstSize_tooSmall);
    if (srcSize > dstCapacity) return ERROR(dstSize_tooSmall);
    memcpy(dst, src, srcSize);
    return srcSize;
}

    if (regenSize > dstCapacity) return ERROR(dstSize_tooSmall);
    memset(dst, *(const BYTE*)src, regenSize);
    return regenSize;
}

/* Hidden declaration for fullbench */
size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx* dctx,
                          const void* src, size_t srcSize);
/*! ZSTD_decodeLiteralsBlock() :
 * @return : nb of bytes read from src (< srcSize )
 *  note : symbol not declared but exposed for fullbench */
size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx* dctx,
                          const void* src, size_t srcSize)   /* note : srcSize < BLOCKSIZE */

        switch(litEncType)
        {
        case set_repeat:
            if (dctx->litEntropy==0) return ERROR(dictionary_corrupted);
            /* fall-through */

        case set_compressed:
            if (srcSize < 5) return ERROR(corruption_detected);   /* srcSize >= MIN_CBLOCK_SIZE == 3; here we need up to 5 for case 3 */
            {   size_t lhSize, litSize, litCSize;
                U32 singleStream=0;
                U32 const lhlCode = (istart[0] >> 2) & 3;

                    break;
                }
                if (litSize > ZSTD_BLOCKSIZE_MAX) return ERROR(corruption_detected);
                if (litCSize + lhSize > srcSize) return ERROR(corruption_detected);

                /* prefetch huffman table if cold */
                if (dctx->ddictIsCold && (litSize > 768 /* heuristic */)) {
                    PREFETCH_AREA(dctx->HUFptr, sizeof(dctx->entropy.hufTable));
                }

                if (HUF_isError((litEncType==set_repeat) ?
                                    ( singleStream ?
                                        HUF_decompress1X_usingDTable_bmi2(dctx->litBuffer, litSize, istart+lhSize, litCSize, dctx->HUFptr, dctx->bmi2) :
                                        HUF_decompress4X_usingDTable_bmi2(dctx->litBuffer, litSize, istart+lhSize, litCSize, dctx->HUFptr, dctx->bmi2) ) :
                                    ( singleStream ?
                                        HUF_decompress1X1_DCtx_wksp_bmi2(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart+lhSize, litCSize,
                                                                         dctx->workspace, sizeof(dctx->workspace), dctx->bmi2) :
                                        HUF_decompress4X_hufOnly_wksp_bmi2(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart+lhSize, litCSize,
                                                                           dctx->workspace, sizeof(dctx->workspace), dctx->bmi2))))
                    return ERROR(corruption_detected);

                dctx->litPtr = dctx->litBuffer;
                dctx->litSize = litSize;
                dctx->litEntropy = 1;

 *           or an error code if it fails */
static size_t ZSTD_buildSeqTable(ZSTD_seqSymbol* DTableSpace, const ZSTD_seqSymbol** DTablePtr,
                                 symbolEncodingType_e type, U32 max, U32 maxLog,
                                 const void* src, size_t srcSize,
                                 const U32* baseValue, const U32* nbAdditionalBits,
                                 const ZSTD_seqSymbol* defaultTable, U32 flagRepeatTable,
                                 int ddictIsCold, int nbSeq)
{
    switch(type)
    {
    case set_rle :
        if (!srcSize) return ERROR(srcSize_wrong);

    case set_basic :
        *DTablePtr = defaultTable;
        return 0;
    case set_repeat:
        if (!flagRepeatTable) return ERROR(corruption_detected);
        /* prefetch FSE table if used */
        if (ddictIsCold && (nbSeq > 24 /* heuristic */)) {
            const void* const pStart = *DTablePtr;
            size_t const pSize = sizeof(ZSTD_seqSymbol) * (SEQSYMBOL_TABLE_SIZE(maxLog));
            PREFETCH_AREA(pStart, pSize);
        }
        return 0;
    case set_compressed :
        {   U32 tableLog;
            S16 norm[MaxSeq+1];
            size_t const headerSize = FSE_readNCount(norm, &max, &tableLog, src, srcSize);

                    27, 28, 29,   30,    31,    32,    33,    34,
                    35, 37, 39,   41,    43,    47,    51,    59,
                    67, 83, 99, 0x83, 0x103, 0x203, 0x403, 0x803,
                    0x1003, 0x2003, 0x4003, 0x8003, 0x10003 };

/* Hidden delcaration for fullbench */
size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx* dctx, int* nbSeqPtr,
                             const void* src, size_t srcSize);

size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx* dctx, int* nbSeqPtr,
                             const void* src, size_t srcSize)
{
    const BYTE* const istart = (const BYTE* const)src;
    const BYTE* const iend = istart + srcSize;
    const BYTE* ip = istart;
    int nbSeq;
    DEBUGLOG(5, "ZSTD_decodeSeqHeaders");

    /* check */
    if (srcSize < MIN_SEQUENCES_SIZE) return ERROR(srcSize_wrong);

    /* SeqHead */
    nbSeq = *ip++;
    if (!nbSeq) { *nbSeqPtr=0; return 1; }
    if (nbSeq > 0x7F) {
        if (nbSeq == 0xFF) {
            if (ip+2 > iend) return ERROR(srcSize_wrong);
            nbSeq = MEM_readLE16(ip) + LONGNBSEQ, ip+=2;
        } else {
            if (ip >= iend) return ERROR(srcSize_wrong);
            nbSeq = ((nbSeq-0x80)<<8) + *ip++;
        }
    }
    *nbSeqPtr = nbSeq;

    /* FSE table descriptors */
    if (ip+4 > iend) return ERROR(srcSize_wrong); /* minimum possible size */
    {   symbolEncodingType_e const LLtype = (symbolEncodingType_e)(*ip >> 6);
        symbolEncodingType_e const OFtype = (symbolEncodingType_e)((*ip >> 4) & 3);

        /* Build DTables */
        {   size_t const llhSize = ZSTD_buildSeqTable(dctx->entropy.LLTable, &dctx->LLTptr,
                                                      LLtype, MaxLL, LLFSELog,
                                                      ip, iend-ip,
                                                      LL_base, LL_bits,
                                                      LL_defaultDTable, dctx->fseEntropy,
                                                      dctx->ddictIsCold, nbSeq);
            if (ZSTD_isError(llhSize)) return ERROR(corruption_detected);
            ip += llhSize;
        }

        {   size_t const ofhSize = ZSTD_buildSeqTable(dctx->entropy.OFTable, &dctx->OFTptr,
                                                      OFtype, MaxOff, OffFSELog,
                                                      ip, iend-ip,
                                                      OF_base, OF_bits,
                                                      OF_defaultDTable, dctx->fseEntropy,
                                                      dctx->ddictIsCold, nbSeq);
            if (ZSTD_isError(ofhSize)) return ERROR(corruption_detected);
            ip += ofhSize;
        }

        {   size_t const mlhSize = ZSTD_buildSeqTable(dctx->entropy.MLTable, &dctx->MLTptr,
                                                      MLtype, MaxML, MLFSELog,
                                                      ip, iend-ip,
                                                      ML_base, ML_bits,
                                                      ML_defaultDTable, dctx->fseEntropy,
                                                      dctx->ddictIsCold, nbSeq);
            if (ZSTD_isError(mlhSize)) return ERROR(corruption_detected);
            ip += mlhSize;
        }
    }

    /* prefetch dictionary content */
    if (dctx->ddictIsCold) {
        size_t const dictSize = (const char*)dctx->prefixStart - (const char*)dctx->virtualStart;
        size_t const psmin = MIN(dictSize, (size_t)(64*nbSeq) /* heuristic */ );
        size_t const pSize = MIN(psmin, 128 KB /* protection */ );
        const void* const pStart = (const char*)dctx->dictEnd - pSize;
        PREFETCH_AREA(pStart, pSize);
        dctx->ddictIsCold = 0;
    }

    return ip-istart;
}

HINT_INLINE
size_t ZSTD_execSequence(BYTE* op,
                         BYTE* const oend, seq_t sequence,
                         const BYTE** litPtr, const BYTE* const litLimit,
                         const BYTE* const prefixStart, const BYTE* const virtualStart, const BYTE* const dictEnd)
{
    BYTE* const oLitEnd = op + sequence.litLength;
    size_t const sequenceLength = sequence.litLength + sequence.matchLength;
    BYTE* const oMatchEnd = op + sequenceLength;   /* risk : address space overflow (32-bits) */
    BYTE* const oend_w = oend - WILDCOPY_OVERLENGTH;

    const BYTE* match = oLitEnd - sequence.offset;

    /* check */
    if (oMatchEnd>oend) return ERROR(dstSize_tooSmall); /* last match must start at a minimum distance of WILDCOPY_OVERLENGTH from oend */
    if (iLitEnd > litLimit) return ERROR(corruption_detected);   /* over-read beyond lit buffer */
    if (oLitEnd>oend_w) return ZSTD_execSequenceLast7(op, oend, sequence, litPtr, litLimit, prefixStart, virtualStart, dictEnd);

    /* copy Literals */
    ZSTD_copy8(op, *litPtr);
    if (sequence.litLength > 8)
        ZSTD_wildcopy(op+8, (*litPtr)+8, sequence.litLength - 8);   /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */
    op = oLitEnd;
    *litPtr = iLitEnd;   /* update for next sequence */

    /* copy Match */
    if (sequence.offset > (size_t)(oLitEnd - prefixStart)) {
        /* offset beyond prefix -> go into extDict */
        if (sequence.offset > (size_t)(oLitEnd - virtualStart))
            return ERROR(corruption_detected);
        match = dictEnd + (match - prefixStart);
        if (match + sequence.matchLength <= dictEnd) {
            memmove(oLitEnd, match, sequence.matchLength);
            return sequenceLength;
        }
        /* span extDict & currentPrefixSegment */
        {   size_t const length1 = dictEnd - match;
            memmove(oLitEnd, match, length1);
            op = oLitEnd + length1;
            sequence.matchLength -= length1;
            match = prefixStart;
            if (op > oend_w || sequence.matchLength < MINMATCH) {
              U32 i;
              for (i = 0; i < sequence.matchLength; ++i) op[i] = match[i];
              return sequenceLength;
            }

    BYTE* const ostart = (BYTE* const)dst;
    BYTE* const oend = ostart + maxDstSize;
    BYTE* op = ostart;
    const BYTE* litPtr = dctx->litPtr;
    const BYTE* const litEnd = litPtr + dctx->litSize;
    const BYTE* const prefixStart = (const BYTE*) (dctx->prefixStart);
    const BYTE* const vBase = (const BYTE*) (dctx->virtualStart);
    const BYTE* const dictEnd = (const BYTE*) (dctx->dictEnd);
    DEBUGLOG(5, "ZSTD_decompressSequences_body");

    /* Regen sequences */
    if (nbSeq) {
        seqState_t seqState;
        dctx->fseEntropy = 1;

        ZSTD_initFseState(&seqState.stateML, &seqState.DStream, dctx->MLTptr);

        for ( ; (BIT_reloadDStream(&(seqState.DStream)) <= BIT_DStream_completed) && nbSeq ; ) {
            nbSeq--;
            {   seq_t const sequence = ZSTD_decodeSequence(&seqState, isLongOffset);
                size_t const oneSeqSize = ZSTD_execSequence(op, oend, sequence, &litPtr, litEnd, prefixStart, vBase, dictEnd);
                DEBUGLOG(6, "regenerated sequence size : %u", (U32)oneSeqSize);
                if (ZSTD_isError(oneSeqSize)) return oneSeqSize;
                op += oneSeqSize;
        }   }

        /* check if reached exact end */
        DEBUGLOG(5, "ZSTD_decompressSequences_body: after decode loop, remaining nbSeq : %i", nbSeq);
        if (nbSeq) return ERROR(corruption_detected);
        /* save reps for next block */
        { U32 i; for (i=0; i<ZSTD_REP_NUM; i++) dctx->entropy.rep[i] = (U32)(seqState.prevOffset[i]); }
    }

    BYTE* const ostart = (BYTE* const)dst;
    BYTE* const oend = ostart + maxDstSize;
    BYTE* op = ostart;
    const BYTE* litPtr = dctx->litPtr;
    const BYTE* const litEnd = litPtr + dctx->litSize;
    const BYTE* const prefixStart = (const BYTE*) (dctx->prefixStart);
    const BYTE* const dictStart = (const BYTE*) (dctx->virtualStart);
    const BYTE* const dictEnd = (const BYTE*) (dctx->dictEnd);

    /* Regen sequences */
    if (nbSeq) {
#define STORED_SEQS 4

{   /* blockType == blockCompressed */
    const BYTE* ip = (const BYTE*)src;
    /* isLongOffset must be true if there are long offsets.
     * Offsets are long if they are larger than 2^STREAM_ACCUMULATOR_MIN.
     * We don't expect that to be the case in 64-bit mode.
     * In block mode, window size is not known, so we have to be conservative.
     * (note: but it could be evaluated from current-lowLimit)
     */
    ZSTD_longOffset_e const isLongOffset = (ZSTD_longOffset_e)(MEM_32bits() && (!frame || dctx->fParams.windowSize > (1ULL << STREAM_ACCUMULATOR_MIN)));
    DEBUGLOG(5, "ZSTD_decompressBlock_internal (size : %u)", (U32)srcSize);

    if (srcSize >= ZSTD_BLOCKSIZE_MAX) return ERROR(srcSize_wrong);

static void ZSTD_checkContinuity(ZSTD_DCtx* dctx, const void* dst)
{
    if (dst != dctx->previousDstEnd) {   /* not contiguous */
        dctx->dictEnd = dctx->previousDstEnd;
        dctx->virtualStart = (const char*)dst - ((const char*)(dctx->previousDstEnd) - (const char*)(dctx->prefixStart));
        dctx->prefixStart = dst;
        dctx->previousDstEnd = dst;
    }
}

size_t ZSTD_decompressBlock(ZSTD_DCtx* dctx,

    dctx->previousDstEnd = (const char*)blockStart + blockSize;
    return blockSize;
}


static size_t ZSTD_generateNxBytes(void* dst, size_t dstCapacity, BYTE value, size_t length)
{
    if (length > dstCapacity) return ERROR(dstSize_tooSmall);
    memset(dst, value, length);
    return length;
}

/** ZSTD_findFrameCompressedSize() :
 *  compatible with legacy mode

    if (ZSTD_isLegacy(src, srcSize))
        return ZSTD_findFrameCompressedSizeLegacy(src, srcSize);
#endif
    if ( (srcSize >= ZSTD_skippableHeaderSize)
      && (MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START ) {
        return ZSTD_skippableHeaderSize + MEM_readLE32((const BYTE*)src + ZSTD_FRAMEIDSIZE);
    } else {
        const BYTE* ip = (const BYTE*)src;
        const BYTE* const ipstart = ip;
        size_t remainingSize = srcSize;
        ZSTD_frameHeader zfh;

        }

        if (zfh.checksumFlag) {   /* Final frame content checksum */
            if (remainingSize < 4) return ERROR(srcSize_wrong);
            ip += 4;
        }

        return ip - ipstart;
    }
}

    *srcPtr = ip;
    *srcSizePtr = remainingSize;
    return op-ostart;
}

static size_t ZSTD_decompressMultiFrame(ZSTD_DCtx* dctx,
                                        void* dst, size_t dstCapacity,
                                  const void* src, size_t srcSize,
                                  const void* dict, size_t dictSize,
                                  const ZSTD_DDict* ddict)
{
    void* const dststart = dst;
    int moreThan1Frame = 0;

    DEBUGLOG(5, "ZSTD_decompressMultiFrame");
    assert(dict==NULL || ddict==NULL);  /* either dict or ddict set, not both */

    if (ddict) {
        dict = ZSTD_DDictDictContent(ddict);
        dictSize = ZSTD_DDictDictSize(ddict);
    }

    while (srcSize >= ZSTD_frameHeaderSize_prefix) {

#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT >= 1)
        if (ZSTD_isLegacy(src, srcSize)) {
            size_t decodedSize;
            size_t const frameSize = ZSTD_findFrameCompressedSizeLegacy(src, srcSize);

            continue;
        }
#endif

        {   U32 const magicNumber = MEM_readLE32(src);
            DEBUGLOG(4, "reading magic number %08X (expecting %08X)",
                        (U32)magicNumber, (U32)ZSTD_MAGICNUMBER);
            if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
                size_t skippableSize;
                if (srcSize < ZSTD_skippableHeaderSize)
                    return ERROR(srcSize_wrong);
                skippableSize = MEM_readLE32((const BYTE*)src + ZSTD_FRAMEIDSIZE)
                              + ZSTD_skippableHeaderSize;
                if (srcSize < skippableSize) return ERROR(srcSize_wrong);

                src = (const BYTE *)src + skippableSize;
                srcSize -= skippableSize;
                continue;
        }   }

        if (ddict) {
            /* we were called from ZSTD_decompress_usingDDict */
            CHECK_F(ZSTD_decompressBegin_usingDDict(dctx, ddict));
        } else {

        }
        ZSTD_checkContinuity(dctx, dst);

        {   const size_t res = ZSTD_decompressFrame(dctx, dst, dstCapacity,
                                                    &src, &srcSize);
            if ( (ZSTD_getErrorCode(res) == ZSTD_error_prefix_unknown)
              && (moreThan1Frame==1) ) {
                /* at least one frame successfully completed,
                 * but following bytes are garbage :
                 * it's more likely to be a srcSize error,
                 * specifying more bytes than compressed size of frame(s).
                 * This error message replaces ERROR(prefix_unknown),
                 * which would be confusing, as the first header is actually correct.
                 * Note that one could be unlucky, it might be a corruption error instead,
                 * happening right at the place where we expect zstd magic bytes.
                 * But this is _much_ less likely than a srcSize field error. */
                return ERROR(srcSize_wrong);
            }
            if (ZSTD_isError(res)) return res;
            /* no need to bound check, ZSTD_decompressFrame already has */
            dst = (BYTE*)dst + res;
            dstCapacity -= res;
        }
        moreThan1Frame = 1;
    }  /* while (srcSize >= ZSTD_frameHeaderSize_prefix) */

    if (srcSize) return ERROR(srcSize_wrong); /* input not entirely consumed */

    return (BYTE*)dst - (BYTE*)dststart;

    regenSize = ZSTD_decompressDCtx(dctx, dst, dstCapacity, src, srcSize);
    ZSTD_freeDCtx(dctx);
    return regenSize;
#else   /* stack mode */
    ZSTD_DCtx dctx;
    ZSTD_initDCtx_internal(&dctx);
    return ZSTD_decompressDCtx(&dctx, dst, dstCapacity, src, srcSize);
#endif
}

    switch (dctx->stage)
    {
    case ZSTDds_getFrameHeaderSize :
        assert(src != NULL);
        if (dctx->format == ZSTD_f_zstd1) {  /* allows header */
            assert(srcSize >= ZSTD_FRAMEIDSIZE);  /* to read skippable magic number */
            if ((MEM_readLE32(src) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {        /* skippable frame */
                memcpy(dctx->headerBuffer, src, srcSize);
                dctx->expected = ZSTD_skippableHeaderSize - srcSize;  /* remaining to load to get full skippable frame header */
                dctx->stage = ZSTDds_decodeSkippableHeader;
                return 0;

    case ZSTDds_decodeSkippableHeader:
        assert(src != NULL);
        assert(srcSize <= ZSTD_skippableHeaderSize);
        memcpy(dctx->headerBuffer + (ZSTD_skippableHeaderSize - srcSize), src, srcSize);   /* complete skippable header */
        dctx->expected = MEM_readLE32(dctx->headerBuffer + ZSTD_FRAMEIDSIZE);   /* note : dctx->expected can grow seriously large, beyond local buffer size */
        dctx->stage = ZSTDds_skipFrame;
        return 0;

    case ZSTDds_skipFrame:
        dctx->expected = 0;

static size_t ZSTD_refDictContent(ZSTD_DCtx* dctx, const void* dict, size_t dictSize)
{
    dctx->dictEnd = dctx->previousDstEnd;
    dctx->virtualStart = (const char*)dict - ((const char*)(dctx->previousDstEnd) - (const char*)(dctx->prefixStart));
    dctx->prefixStart = dict;
    dctx->previousDstEnd = (const char*)dict + dictSize;
    return 0;
}

/*! ZSTD_loadEntropy() :
 *  dict : must point at beginning of a valid zstd dictionary.
 * @return : size of entropy tables read */
static size_t ZSTD_loadEntropy(ZSTD_entropyDTables_t* entropy,
                         const void* const dict, size_t const dictSize)
{
    const BYTE* dictPtr = (const BYTE*)dict;
    const BYTE* const dictEnd = dictPtr + dictSize;

    if (dictSize <= 8) return ERROR(dictionary_corrupted);
    assert(MEM_readLE32(dict) == ZSTD_MAGIC_DICTIONARY);   /* dict must be valid */
    dictPtr += 8;   /* skip header = magic + dictID */

    ZSTD_STATIC_ASSERT(offsetof(ZSTD_entropyDTables_t, OFTable) == offsetof(ZSTD_entropyDTables_t, LLTable) + sizeof(entropy->LLTable));
    ZSTD_STATIC_ASSERT(offsetof(ZSTD_entropyDTables_t, MLTable) == offsetof(ZSTD_entropyDTables_t, OFTable) + sizeof(entropy->OFTable));
    ZSTD_STATIC_ASSERT(sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable) >= HUF_DECOMPRESS_WORKSPACE_SIZE);
    {   void* const workspace = &entropy->LLTable;   /* use fse tables as temporary workspace; implies fse tables are grouped together */
        size_t const workspaceSize = sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable);
        size_t const hSize = HUF_readDTableX2_wksp(entropy->hufTable,
                                                dictPtr, dictEnd - dictPtr,
                                                workspace, workspaceSize);
        if (HUF_isError(hSize)) return ERROR(dictionary_corrupted);
        dictPtr += hSize;
    }

    {   short offcodeNCount[MaxOff+1];
        U32 offcodeMaxValue = MaxOff, offcodeLog;
        size_t const offcodeHeaderSize = FSE_readNCount(offcodeNCount, &offcodeMaxValue, &offcodeLog, dictPtr, dictEnd-dictPtr);
        if (FSE_isError(offcodeHeaderSize)) return ERROR(dictionary_corrupted);
        if (offcodeMaxValue > MaxOff) return ERROR(dictionary_corrupted);
        if (offcodeLog > OffFSELog) return ERROR(dictionary_corrupted);
        ZSTD_buildFSETable( entropy->OFTable,
                            offcodeNCount, offcodeMaxValue,
                            OF_base, OF_bits,
                            offcodeLog);
        dictPtr += offcodeHeaderSize;
    }

        unsigned matchlengthMaxValue = MaxML, matchlengthLog;
        size_t const matchlengthHeaderSize = FSE_readNCount(matchlengthNCount, &matchlengthMaxValue, &matchlengthLog, dictPtr, dictEnd-dictPtr);
        if (FSE_isError(matchlengthHeaderSize)) return ERROR(dictionary_corrupted);
        if (matchlengthMaxValue > MaxML) return ERROR(dictionary_corrupted);
        if (matchlengthLog > MLFSELog) return ERROR(dictionary_corrupted);
        ZSTD_buildFSETable( entropy->MLTable,
                            matchlengthNCount, matchlengthMaxValue,
                            ML_base, ML_bits,
                            matchlengthLog);
        dictPtr += matchlengthHeaderSize;
    }

        unsigned litlengthMaxValue = MaxLL, litlengthLog;
        size_t const litlengthHeaderSize = FSE_readNCount(litlengthNCount, &litlengthMaxValue, &litlengthLog, dictPtr, dictEnd-dictPtr);
        if (FSE_isError(litlengthHeaderSize)) return ERROR(dictionary_corrupted);
        if (litlengthMaxValue > MaxLL) return ERROR(dictionary_corrupted);
        if (litlengthLog > LLFSELog) return ERROR(dictionary_corrupted);
        ZSTD_buildFSETable( entropy->LLTable,
                            litlengthNCount, litlengthMaxValue,
                            LL_base, LL_bits,
                            litlengthLog);
        dictPtr += litlengthHeaderSize;
    }

    if (dictSize < 8) return ZSTD_refDictContent(dctx, dict, dictSize);
    {   U32 const magic = MEM_readLE32(dict);
        if (magic != ZSTD_MAGIC_DICTIONARY) {
            return ZSTD_refDictContent(dctx, dict, dictSize);   /* pure content mode */
    }   }
    dctx->dictID = MEM_readLE32((const char*)dict + ZSTD_FRAMEIDSIZE);

    /* load entropy tables */
    {   size_t const eSize = ZSTD_loadEntropy(&dctx->entropy, dict, dictSize);
        if (ZSTD_isError(eSize)) return ERROR(dictionary_corrupted);
        dict = (const char*)dict + eSize;

    /* reference dictionary content */
    return ZSTD_refDictContent(dctx, dict, dictSize);
}

size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx)
{
    assert(dctx != NULL);
    dctx->expected = ZSTD_startingInputLength(dctx->format);  /* dctx->format must be properly set */
    dctx->stage = ZSTDds_getFrameHeaderSize;
    dctx->decodedSize = 0;
    dctx->previousDstEnd = NULL;
    dctx->prefixStart = NULL;
    dctx->virtualStart = NULL;
    dctx->dictEnd = NULL;
    dctx->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001);  /* cover both little and big endian */
    dctx->litEntropy = dctx->fseEntropy = 0;
    dctx->dictID = 0;
    ZSTD_STATIC_ASSERT(sizeof(dctx->entropy.rep) == sizeof(repStartValue));

    ZSTD_customMem cMem;
};  /* typedef'd to ZSTD_DDict within "zstd.h" */

static const void* ZSTD_DDictDictContent(const ZSTD_DDict* ddict)
{
    assert(ddict != NULL);
    return ddict->dictContent;
}

static size_t ZSTD_DDictDictSize(const ZSTD_DDict* ddict)
{
    assert(ddict != NULL);
    return ddict->dictSize;
}

size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict)
{
    DEBUGLOG(4, "ZSTD_decompressBegin_usingDDict");
    assert(dctx != NULL);
    if (ddict) {
        dctx->ddictIsCold = (dctx->dictEnd != (const char*)ddict->dictContent + ddict->dictSize);
        DEBUGLOG(4, "DDict is %s",
                    dctx->ddictIsCold ? "~cold~" : "hot!");
    }
    CHECK_F( ZSTD_decompressBegin(dctx) );
    if (ddict) {   /* NULL ddict is equivalent to no dictionary */
        dctx->dictID = ddict->dictID;
        dctx->prefixStart = ddict->dictContent;
        dctx->virtualStart = ddict->dictContent;
        dctx->dictEnd = (const BYTE*)ddict->dictContent + ddict->dictSize;
        dctx->previousDstEnd = dctx->dictEnd;
        if (ddict->entropyPresent) {
            dctx->litEntropy = 1;
            dctx->fseEntropy = 1;
            dctx->LLTptr = ddict->entropy.LLTable;
            dctx->MLTptr = ddict->entropy.MLTable;
            dctx->OFTptr = ddict->entropy.OFTable;
            dctx->HUFptr = ddict->entropy.hufTable;
            dctx->entropy.rep[0] = ddict->entropy.rep[0];
            dctx->entropy.rep[1] = ddict->entropy.rep[1];
            dctx->entropy.rep[2] = ddict->entropy.rep[2];
        } else {
            dctx->litEntropy = 0;
            dctx->fseEntropy = 0;
        }
    }
    return 0;
}

static size_t
ZSTD_loadEntropy_inDDict(ZSTD_DDict* ddict,
                         ZSTD_dictContentType_e dictContentType)
{
    ddict->dictID = 0;
    ddict->entropyPresent = 0;
    if (dictContentType == ZSTD_dct_rawContent) return 0;

            if (dictContentType == ZSTD_dct_fullDict)
                return ERROR(dictionary_corrupted);   /* only accept specified dictionaries */
            return 0;   /* pure content mode */
        }
    }
    ddict->dictID = MEM_readLE32((const char*)ddict->dictContent + ZSTD_FRAMEIDSIZE);

    /* load entropy tables */
    CHECK_E( ZSTD_loadEntropy(&ddict->entropy,
                              ddict->dictContent, ddict->dictSize),
             dictionary_corrupted );
    ddict->entropyPresent = 1;
    return 0;
}

                                      ZSTD_dictContentType_e dictContentType)
{
    if ((dictLoadMethod == ZSTD_dlm_byRef) || (!dict) || (!dictSize)) {
        ddict->dictBuffer = NULL;
        ddict->dictContent = dict;
        if (!dict) dictSize = 0;
    } else {
        void* const internalBuffer = ZSTD_malloc(dictSize, ddict->cMem);
        ddict->dictBuffer = internalBuffer;
        ddict->dictContent = internalBuffer;
        if (!internalBuffer) return ERROR(memory_allocation);

                                      ZSTD_customMem customMem)
{
    if (!customMem.customAlloc ^ !customMem.customFree) return NULL;

    {   ZSTD_DDict* const ddict = (ZSTD_DDict*) ZSTD_malloc(sizeof(ZSTD_DDict), customMem);
        if (ddict == NULL) return NULL;
        ddict->cMem = customMem;
        {   size_t const initResult = ZSTD_initDDict_internal(ddict,
                                            dict, dictSize,
                                            dictLoadMethod, dictContentType);
            if (ZSTD_isError(initResult)) {
                ZSTD_freeDDict(ddict);
                return NULL;
        }   }
        return ddict;
    }
}

/*! ZSTD_createDDict() :

    return ZSTD_createDDict_advanced(dictBuffer, dictSize, ZSTD_dlm_byRef, ZSTD_dct_auto, allocator);
}


const ZSTD_DDict* ZSTD_initStaticDDict(
                                void* sBuffer, size_t sBufferSize,
                                const void* dict, size_t dictSize,
                                ZSTD_dictLoadMethod_e dictLoadMethod,
                                ZSTD_dictContentType_e dictContentType)
{
    size_t const neededSpace = sizeof(ZSTD_DDict)
                             + (dictLoadMethod == ZSTD_dlm_byRef ? 0 : dictSize);
    ZSTD_DDict* const ddict = (ZSTD_DDict*)sBuffer;
    assert(sBuffer != NULL);
    assert(dict != NULL);
    if ((size_t)sBuffer & 7) return NULL;   /* 8-aligned */
    if (sBufferSize < neededSpace) return NULL;
    if (dictLoadMethod == ZSTD_dlm_byCopy) {
        memcpy(ddict+1, dict, dictSize);  /* local copy */
        dict = ddict+1;
    }
    if (ZSTD_isError( ZSTD_initDDict_internal(ddict,
                                              dict, dictSize,
                                              ZSTD_dlm_byRef, dictContentType) ))
        return NULL;
    return ddict;
}

 *  It can still be loaded, but as a content-only dictionary. */
unsigned ZSTD_getDictID_fromDict(const void* dict, size_t dictSize)
{
    if (dictSize < 8) return 0;
    if (MEM_readLE32(dict) != ZSTD_MAGIC_DICTIONARY) return 0;
    return MEM_readLE32((const char*)dict + ZSTD_FRAMEIDSIZE);
}

/*! ZSTD_getDictID_fromDDict() :
 *  Provides the dictID of the dictionary loaded into `ddict`.
 *  If @return == 0, the dictionary is not conformant to Zstandard specification, or empty.

{
    return ZSTD_freeDCtx(zds);
}


/* ***  Initialization  *** */

size_t ZSTD_DStreamInSize(void)  { return ZSTD_BLOCKSIZE_MAX + ZSTD_blockHeaderSize; }
size_t ZSTD_DStreamOutSize(void) { return ZSTD_BLOCKSIZE_MAX; }

size_t ZSTD_DCtx_loadDictionary_advanced(ZSTD_DCtx* dctx,
                                   const void* dict, size_t dictSize,
                                         ZSTD_dictLoadMethod_e dictLoadMethod,
                                         ZSTD_dictContentType_e dictContentType)
{
    if (dctx->streamStage != zdss_init) return ERROR(stage_wrong);
    ZSTD_freeDDict(dctx->ddictLocal);
    if (dict && dictSize >= 8) {
        dctx->ddictLocal = ZSTD_createDDict_advanced(dict, dictSize, dictLoadMethod, dictContentType, dctx->customMem);

 * this function cannot fail */
size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize)
{
    DEBUGLOG(4, "ZSTD_initDStream_usingDict");
    zds->streamStage = zdss_init;
    zds->noForwardProgress = 0;
    CHECK_F( ZSTD_DCtx_loadDictionary(zds, dict, dictSize) );
    return ZSTD_frameHeaderSize_prefix;
}

/* note : this variant can't fail */
size_t ZSTD_initDStream(ZSTD_DStream* zds)
{
    DEBUGLOG(4, "ZSTD_initDStream");
    return ZSTD_initDStream_usingDict(zds, NULL, 0);
}

/* ZSTD_initDStream_usingDDict() :
 * ddict will just be referenced, and must outlive decompression session
 * this function cannot fail */

        case DStream_p_maxWindowSize :
            DEBUGLOG(4, "setting maxWindowSize = %u KB", paramValue >> 10);
            dctx->maxWindowSize = paramValue ? paramValue : (U32)(-1);
            break;
    }
    return 0;
}

size_t ZSTD_DCtx_refDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict)
{
    if (dctx->streamStage != zdss_init) return ERROR(stage_wrong);
    dctx->ddict = ddict;
    return 0;
}

size_t ZSTD_DCtx_setMaxWindowSize(ZSTD_DCtx* dctx, size_t maxWindowSize)
{

                {   size_t const hint = ZSTD_decompressLegacyStream(zds->legacyContext, zds->legacyVersion, output, input);
                    if (hint==0) zds->streamStage = zdss_init;
                    return hint;
            }   }
#endif
            {   size_t const hSize = ZSTD_getFrameHeader_advanced(&zds->fParams, zds->headerBuffer, zds->lhSize, zds->format);
                DEBUGLOG(5, "header size : %u", (U32)hSize);
                if (ZSTD_isError(hSize)) {
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
                    U32 const legacyVersion = ZSTD_isLegacy(istart, iend-istart);
                    if (legacyVersion) {

            /* Consume header (see ZSTDds_decodeFrameHeader) */
            DEBUGLOG(4, "Consume header");
            CHECK_F(ZSTD_decompressBegin_usingDDict(zds, zds->ddict));

            if ((MEM_readLE32(zds->headerBuffer) & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {  /* skippable frame */
                zds->expected = MEM_readLE32(zds->headerBuffer + ZSTD_FRAMEIDSIZE);
                zds->stage = ZSTDds_skipFrame;
            } else {
                CHECK_F(ZSTD_decodeFrameHeader(zds, zds->headerBuffer, zds->lhSize));
                zds->expected = ZSTD_blockHeaderSize;
                zds->stage = ZSTDds_decodeBlockHeader;

        default: return ERROR(GENERIC);   /* impossible */
    }   }

    /* result */
    input->pos = (size_t)(ip - (const char*)(input->src));
    output->pos = (size_t)(op - (char*)(output->dst));
    if ((ip==istart) && (op==ostart)) {  /* no forward progress */
        zds->noForwardProgress ++;
        if (zds->noForwardProgress >= ZSTD_NO_FORWARD_PROGRESS_MAX) {
            if (op==oend) return ERROR(dstSize_tooSmall);
            if (ip==iend) return ERROR(srcSize_wrong);
            assert(0);
        }
    } else {
        zds->noForwardProgress = 0;
    }
    {   size_t nextSrcSizeHint = ZSTD_nextSrcSizeToDecompress(zds);
        if (!nextSrcSizeHint) {   /* frame fully decoded */
            if (zds->outEnd == zds->outStart) {  /* output fully flushed */
                if (zds->hostageByte) {
                    if (input->pos >= input->size) {