view release on metacpan or  search on metacpan

ext/zstd/lib/dictBuilder/zdict.c  view on Meta::CPAN

    for (u=1; u<dictList[0].pos; u++)
        dictSize += dictList[u].length;
    return dictSize;
}


static size_t ZDICT_trainBuffer_legacy(dictItem* dictList, U32 dictListSize,
                            const void* const buffer, size_t bufferSize,   /* buffer must end with noisy guard band */
                            const size_t* fileSizes, unsigned nbFiles,
                            unsigned minRatio, U32 notificationLevel)
{
    int* const suffix0 = (int*)malloc((bufferSize+2)*sizeof(*suffix0));
    int* const suffix = suffix0+1;
    U32* reverseSuffix = (U32*)malloc((bufferSize)*sizeof(*reverseSuffix));
    BYTE* doneMarks = (BYTE*)malloc((bufferSize+16)*sizeof(*doneMarks));   /* +16 for overflow security */
    U32* filePos = (U32*)malloc(nbFiles * sizeof(*filePos));
    size_t result = 0;
    clock_t displayClock = 0;
    clock_t const refreshRate = CLOCKS_PER_SEC * 3 / 10;

#   undef  DISPLAYUPDATE
#   define DISPLAYUPDATE(l, ...) if (notificationLevel>=l) { \
            if (ZDICT_clockSpan(displayClock) > refreshRate)  \
            { displayClock = clock(); DISPLAY(__VA_ARGS__); \
            if (notificationLevel>=4) fflush(stderr); } }

    /* init */
    DISPLAYLEVEL(2, "\r%70s\r", "");   /* clean display line */
    if (!suffix0 || !reverseSuffix || !doneMarks || !filePos) {
        result = ERROR(memory_allocation);
        goto _cleanup;
    }
    if (minRatio < MINRATIO) minRatio = MINRATIO;
    memset(doneMarks, 0, bufferSize+16);

    /* limit sample set size (divsufsort limitation)*/
    if (bufferSize > ZDICT_MAX_SAMPLES_SIZE) DISPLAYLEVEL(3, "sample set too large : reduced to %u MB ...\n", (unsigned)(ZDICT_MAX_SAMPLES_SIZE>>20));
    while (bufferSize > ZDICT_MAX_SAMPLES_SIZE) bufferSize -= fileSizes[--nbFiles];

    /* sort */
    DISPLAYLEVEL(2, "sorting %u files of total size %u MB ...\n", nbFiles, (unsigned)(bufferSize>>20));
    {   int const divSuftSortResult = divsufsort((const unsigned char*)buffer, suffix, (int)bufferSize, 0);
        if (divSuftSortResult != 0) { result = ERROR(GENERIC); goto _cleanup; }
    }
    suffix[bufferSize] = (int)bufferSize;   /* leads into noise */
    suffix0[0] = (int)bufferSize;           /* leads into noise */
    /* build reverse suffix sort */
    {   size_t pos;
        for (pos=0; pos < bufferSize; pos++)
            reverseSuffix[suffix[pos]] = (U32)pos;
        /* note filePos tracks borders between samples.
           It's not used at this stage, but planned to become useful in a later update */
        filePos[0] = 0;
        for (pos=1; pos<nbFiles; pos++)
            filePos[pos] = (U32)(filePos[pos-1] + fileSizes[pos-1]);
    }

    DISPLAYLEVEL(2, "finding patterns ... \n");
    DISPLAYLEVEL(3, "minimum ratio : %u \n", minRatio);

    {   U32 cursor; for (cursor=0; cursor < bufferSize; ) {
            dictItem solution;
            if (doneMarks[cursor]) { cursor++; continue; }
            solution = ZDICT_analyzePos(doneMarks, suffix, reverseSuffix[cursor], buffer, minRatio, notificationLevel);
            if (solution.length==0) { cursor++; continue; }
            ZDICT_insertDictItem(dictList, dictListSize, solution, buffer);
            cursor += solution.length;
            DISPLAYUPDATE(2, "\r%4.2f %% \r", (double)cursor / (double)bufferSize * 100.0);
    }   }

_cleanup:
    free(suffix0);
    free(reverseSuffix);
    free(doneMarks);
    free(filePos);
    return result;
}


static void ZDICT_fillNoise(void* buffer, size_t length)
{
    unsigned const prime1 = 2654435761U;
    unsigned const prime2 = 2246822519U;
    unsigned acc = prime1;
    size_t p=0;
    for (p=0; p<length; p++) {
        acc *= prime2;
        ((unsigned char*)buffer)[p] = (unsigned char)(acc >> 21);
    }
}


typedef struct
{
    ZSTD_CDict* dict;    /* dictionary */
    ZSTD_CCtx* zc;     /* working context */
    void* workPlace;   /* must be ZSTD_BLOCKSIZE_MAX allocated */
} EStats_ress_t;

#define MAXREPOFFSET 1024

static void ZDICT_countEStats(EStats_ress_t esr, const ZSTD_parameters* params,
                              unsigned* countLit, unsigned* offsetcodeCount, unsigned* matchlengthCount, unsigned* litlengthCount, U32* repOffsets,
                              const void* src, size_t srcSize,
                              U32 notificationLevel)
{
    size_t const blockSizeMax = MIN (ZSTD_BLOCKSIZE_MAX, 1 << params->cParams.windowLog);
    size_t cSize;

    if (srcSize > blockSizeMax) srcSize = blockSizeMax;   /* protection vs large samples */
    {   size_t const errorCode = ZSTD_compressBegin_usingCDict_deprecated(esr.zc, esr.dict);
        if (ZSTD_isError(errorCode)) { DISPLAYLEVEL(1, "warning : ZSTD_compressBegin_usingCDict failed \n"); return; }

    }
    cSize = ZSTD_compressBlock_deprecated(esr.zc, esr.workPlace, ZSTD_BLOCKSIZE_MAX, src, srcSize);
    if (ZSTD_isError(cSize)) { DISPLAYLEVEL(3, "warning : could not compress sample size %u \n", (unsigned)srcSize); return; }

    if (cSize) {  /* if == 0; block is not compressible */
        const seqStore_t* const seqStorePtr = ZSTD_getSeqStore(esr.zc);

        /* literals stats */
        {   const BYTE* bytePtr;
            for(bytePtr = seqStorePtr->litStart; bytePtr < seqStorePtr->lit; bytePtr++)
                countLit[*bytePtr]++;
        }

        /* seqStats */
        {   U32 const nbSeq = (U32)(seqStorePtr->sequences - seqStorePtr->sequencesStart);