size_t GetSizeOfHeadData(bool heapPageData)
{
    if (heapPageData) {
        return SizeOfHeapPageHeaderData;
    } else {
        return SizeOfPageHeaderData;
    }
}

// maybe some itemid is not valid
uint16 HeapPageCalcRealRowCnt (char *buf) {
    HeapPageHeaderData *page = (HeapPageHeaderData *)buf;
    uint16 cnt = 0;
    uint16 i;
    uint16 row_cnt = (page->pd_lower - GetPageHeaderSize(page)) / sizeof(ItemIdData);

    for (i = 0; i < row_cnt; i++) {
        if (ItemIdIsNormal(GET_ITEMID_BY_IDX(buf, i))) {
            cnt++;
        }
    }
    return cnt;
}

void DecompressDeconvertRows(char *buf, char *aux_buf, int16 *real_order, uint16 max_row_len, uint16 real_row_cnt) {
    errno_t ret;
    HeapPageHeaderData *page = (HeapPageHeaderData *)buf;
    uint16 row_cnt = real_row_cnt;
    uint32 total_size = page->pd_special - page->pd_upper;
    char *copy_begin = buf + page->pd_upper;
    char *row;
    uint16 i, j, k, cur, up, row_size;

    ret = memset_sp(aux_buf, BLCKSZ, 0, BLCKSZ);
    securec_check(ret, "", "");

    k = 0;
    for (i = 0; i < max_row_len; i++) {
        for (j = 0; j < row_cnt; j++) {
            up = (j == (row_cnt - 1)) ? page->pd_special : GET_ITEMID_BY_IDX(buf, (real_order[j + 1]))->lp_off;
            cur = GET_ITEMID_BY_IDX(buf, (real_order[j]))->lp_off;
            row_size = up - cur;
            row = aux_buf + cur;
            if (i < row_size) {
                row[i] = copy_begin[k++];  // this part is reshaped
            }
        }
    }

    if (k != total_size) {
        printf("ERROR!!! pg_deconvert_rows error...!!!\n");
        ASSERT(0);
        return;
    }

    // cp aux_buf to page_buf
    ret = memcpy_sp(copy_begin, total_size, aux_buf + page->pd_upper, total_size);
    securec_check(ret, "", "");
    return ;
}

// 1: as tuple_offset order, that means asc order.
// 2: store all itemid's idx.
// 3:maybe some itemid is not in order.
void CompressConvertItemRealOrder(char *buf, int16 *real_order, uint16 real_row_cnt) {
    HeapPageHeaderData *page = (HeapPageHeaderData *)buf;
    uint16 row_cnt = (page->pd_lower - GetPageHeaderSize(page)) / sizeof(ItemIdData);
    ItemIdData *begin = (ItemIdData *)(buf + GetPageHeaderSize(page));
    int16 *link_order = real_order + real_row_cnt;

    int16 i, head, curr, prev;
    int16 end = -1; // invalid index

    head = end;
    // very likely to seems that itemids stored by desc order, and ignore invalid itemid
    for (i = 0; i < row_cnt; i++) {
        if (!ItemIdIsNormal(begin + i)) {
            continue;
        }

        if (head == end) {  // set the head idx, insert the first
            link_order[i] = end;
            head = i;
            continue;
        }

        if ((begin + i)->lp_off < (begin + head)->lp_off) {
            link_order[i] = head; // update the head idx
            head = i;
            continue;
        }

        prev = head;
        curr = link_order[head];
        while ((curr != end) && ((begin + i)->lp_off > (begin + curr)->lp_off)) {
            prev = curr;
            curr = link_order[curr];
        }

        link_order[prev] = i;
        link_order[i] = curr;
    }

    // arrange the link to array
    curr = head;
    for (i = 0; i < real_row_cnt; i++) {
        real_order[i] = curr;
        curr = link_order[curr];
    }

    if (curr != end) {
        printf("ERROR!!! pre_convert_real_order error...!!!\n");
        ASSERT(0);
        return;
    }

}

int DecompressPage(const char* src, char* dst, uint8 algorithm)
{
    if (PageIs8BXidHeapVersion(src)) {
        return TemplateDecompressPage<true>(src, dst, algorithm);
    } else {
        return TemplateDecompressPage<false>(src, dst, algorithm);
    }
}

void cprs_diff_deconvert_rows(char *buf, uint32 offset, uint16 min_row_len, uint16 real_row_cnt) {
    uint16 row_cnt = real_row_cnt;
    uint32 common_size = min_row_len;
    uint8 *copy_begin = (uint8 *)(buf + offset);
    uint16 i, j;

    for (i = 0; i < common_size; i++) {
        for (j = 1; j < row_cnt; j++) {
            copy_begin[i  * row_cnt + j] += copy_begin[i  * row_cnt + (j - 1)];
        }
    }
    return ;
}

// to find all row size are diffs in MIN_DIFF_SIZE byts.
bool CompressConvertCheck(char *buf, int16 **real_order, uint16 *max_row_len, uint16 *min_row_len, uint16 *real_row_cnt) {
    HeapPageHeaderData *page = (HeapPageHeaderData *)buf;
    uint16 row_cnt = (page->pd_lower - GetPageHeaderSize(page)) / sizeof(ItemIdData);
    int16 i, row_size;
    ItemIdData *ptr = NULL;
    uint16 up = page->pd_special;
    uint16 min_size = GS_INVALID_ID16;
    uint16 max_size = 0;
    errno_t ret;
    if (page->pd_lower < GetPageHeaderSize(page) || (page->pd_lower >  page->pd_upper)) {
        return false;
    }

    uint16 normal_row_cnt = HeapPageCalcRealRowCnt(buf);
    if (normal_row_cnt < MIN_CONVERT_CNT) {  // no need convert
        return false;
    }

    // to store the real tuple order.
    /*
            --------------------------|--------------------------
            xxxxxxxxxxxxxxxxxxxxxxxxxx|xxxxxxxxxxxxxxxxxxxxxxxxxx
            --------------------------|--------------------------
    */
    // the first part is real array order, and the second part is link.
    *real_order = (int16 *)malloc(sizeof(uint16) * row_cnt * 2);
    if (*real_order == NULL) {
        printf("zfunc compress file");
        return false;
    }
    ret = memset_sp(*real_order, sizeof(uint16) * row_cnt * 2, 0, sizeof(uint16) * row_cnt * 2);
    securec_check(ret, "", "");

    // order the ItemIds by tuple_offset order.
    CompressConvertItemRealOrder(buf, *real_order, normal_row_cnt);

    // do the check,  to check all size of tuples.
    for (i = normal_row_cnt - 1; i >= 0; i--) {
        ptr = GET_ITEMID_BY_IDX(buf, ((*real_order)[i]));

        row_size = up - ptr->lp_off;
        if (row_size < MIN_CONVERT_CNT * 2) {
            return false;
        }

        min_size = (row_size < min_size) ? row_size : min_size;
        max_size = (row_size > max_size) ? row_size : max_size;

        if ((max_size - min_size) > MIN_DIFF_SIZE) {  // no need convert
            return false;
        }
        up = ptr->lp_off;
    }

    // get the min row common size.
    *max_row_len = max_size;
    *min_row_len = min_size;
    *real_row_cnt = normal_row_cnt;
    return true;
}

void DecompressDeconvertItemIds(char *buf, char *aux_buf) {
    errno_t ret;
    HeapPageHeaderData *page = (HeapPageHeaderData *)buf;
    uint16 row_cnt = (page->pd_lower - GetPageHeaderSize(page)) / sizeof(ItemIdData);
    uint32 total_size = row_cnt * sizeof(ItemIdData);
    char *copy_begin = buf + GetPageHeaderSize(page);
    uint16 i, j, k;

    // clear aux_buf
    ret = memset_sp(aux_buf, BLCKSZ, 0, BLCKSZ);
    securec_check(ret, "", "");

    k = 0;
    for (i = 0; i < sizeof(ItemIdData); i++) {
        for (j = 0; j < row_cnt; j++) {
            aux_buf[j  * sizeof(ItemIdData) + i] = copy_begin[k++];
        }
    }

    // cp aux_buf to page_buf
    ret = memcpy_sp(copy_begin, total_size, aux_buf, total_size);
    securec_check(ret, "", "");
    return ;
}


void DecompressDeconvertOnePage(char *buf, char *aux_buf, bool diff_convert) {
    uint16 max_row_len = 0;
    uint16 min_row_len = 0;
    int16 *real_order = NULL; // itemids are not in order sometimes. we must find the real
    uint16 real_row_cnt = 0;

    if (diff_convert) {
        cprs_diff_deconvert_rows(buf, GetPageHeaderSize(buf), sizeof(ItemIdData),
                                 (((HeapPageHeaderData *)buf)->pd_lower - GetPageHeaderSize(buf)) / sizeof(ItemIdData));
    }

    // =======firstly, arrange the itemids.
    DecompressDeconvertItemIds(buf, aux_buf);

    if (!CompressConvertCheck(buf, &real_order, &max_row_len, &min_row_len, &real_row_cnt)) {
        if (real_order != NULL) {
            free(real_order);
        }
        ASSERT(0);
        return ;
    }

    // =======and last, the tuples
    if (diff_convert) {
        cprs_diff_deconvert_rows(buf, ((HeapPageHeaderData *)buf)->pd_upper, min_row_len, real_row_cnt);
    }
    DecompressDeconvertRows(buf, aux_buf, real_order, max_row_len, real_row_cnt);

    if (real_order != NULL) {
        free(real_order);
    }
    return ;
}


void DecompressPageDeconvert(char *src, bool diff_convert)
{
    char *aux_buf = NULL;
    errno_t rc;

    aux_buf = (char *)malloc(BLCKSZ);
    if (aux_buf == NULL) {
        // add log
        return;
    }
    rc = memset_s(aux_buf, BLCKSZ, 0, BLCKSZ);
    securec_check(rc, "", "");

    // do convert
    DecompressDeconvertOnePage(src, aux_buf, diff_convert);

    if (aux_buf != NULL) {
        free(aux_buf);
    }
}


/**
 * DecompressPage() -- Decompress one compressed page.
 *  return size of decompressed page which should be BLCKSZ or
 *         -1 for decompress error
 *         -2 for unrecognized compression algorithm
 *
 * 		note:The size of dst must be greater than or equal to BLCKSZ.
 */
template<bool heapPageData>
int TemplateDecompressPage(const char* src, char* dst, uint8 algorithm)
{
    int decompressed_size;
    char* data;
    uint32 size;
    bool byte_convert, diff_convert;
    size_t sizeOfPageHeaderData = GetSizeOfHeadData(heapPageData);
    int rc = memcpy_s(dst, sizeOfPageHeaderData, src, sizeOfPageHeaderData);
    securec_check(rc, "", "");

    if (heapPageData) {
        data = ((HeapPageCompressData*) src)->data;
        size = ((HeapPageCompressData*) src)->size;
        byte_convert = ((HeapPageCompressData*) src)->byte_convert;
        diff_convert = ((HeapPageCompressData*) src)->diff_convert;
    } else {
        data = ((PageCompressData*) src)->data;
        size = ((PageCompressData*) src)->size;
        byte_convert = ((PageCompressData*) src)->byte_convert;
        diff_convert = ((PageCompressData*) src)->diff_convert;
    }

    switch (algorithm) {
        case COMPRESS_ALGORITHM_PGLZ:
            decompressed_size = lz_decompress(
                data, size, dst + sizeOfPageHeaderData, BLCKSZ - sizeOfPageHeaderData, false);
            break;
        case COMPRESS_ALGORITHM_ZSTD:
            decompressed_size =
                ZSTD_decompress(dst + sizeOfPageHeaderData, BLCKSZ - sizeOfPageHeaderData, data, size);

            if (ZSTD_isError(decompressed_size)) {
                return -1;
            }

            break;

        default:
            return COMPRESS_UNSUPPORTED_ERROR;
            break;
    }

    if (byte_convert) {
        // deconvert dst
        DecompressPageDeconvert(dst, diff_convert);
    }

    return sizeOfPageHeaderData + decompressed_size;
}

// pg_lz
/* ----------
 * pg_lzcompress.c -
 *
 *		This is an implementation of LZ compression for PostgreSQL.
 *		It uses a simple history table and generates 2-3 byte tags
 *		capable of backward copy information for 3-273 bytes with
 *		a max offset of 4095.
 *
 *		Entry routines:
 *
 *			bool
 *			pglz_compress(const char *source, int32 slen, PGLZ_Header *dest,
 *						  const PGLZ_Strategy *strategy);
 *
 *				source is the input data to be compressed.
 *
 *				slen is the length of the input data.
 *
 *				dest is the output area for the compressed result.
 *					It must be at least as big as PGLZ_MAX_OUTPUT(slen).
 *
 *				strategy is a pointer to some information controlling
 *					the compression algorithm. If NULL, the compiled
 *					in default strategy is used.
 *
 *				The return value is TRUE if compression succeeded,
 *				FALSE if not; in the latter case the contents of dest
 *				are undefined.
 *
 *			void
 *			pglz_decompress(const PGLZ_Header *source, char *dest)
 *
 *				source is the compressed input.
 *
 *				dest is the area where the uncompressed data will be
 *					written to. It is the callers responsibility to
 *					provide enough space. The required amount can be
 *					obtained with the macro PGLZ_RAW_SIZE(source).
 *
 *					The data is written to buff exactly as it was handed
 *					to pglz_compress(). No terminating zero byte is added.
 *
 *		The decompression algorithm and internal data format:
 *
 *			PGLZ_Header is defined as
 *
 *				typedef struct PGLZ_Header {
 *					int32		vl_len_;
 *					int32		rawsize;
 *				}
 *
 *			The header is followed by the compressed data itself.
 *
 *			The data representation is easiest explained by describing
 *			the process of decompression.
 *
 *			If VARSIZE(x) == rawsize + sizeof(PGLZ_Header), then the data
 *			is stored uncompressed as plain bytes. Thus, the decompressor
 *			simply copies rawsize bytes from the location after the
 *			header to the destination.
 *
 *			Otherwise the first byte after the header tells what to do
 *			the next 8 times. We call this the control byte.
 *
 *			An unset bit in the control byte means, that one uncompressed
 *			byte follows, which is copied from input to output.
 *
 *			A set bit in the control byte means, that a tag of 2-3 bytes
 *			follows. A tag contains information to copy some bytes, that
 *			are already in the output buffer, to the current location in
 *			the output. Let's call the three tag bytes T1, T2 and T3. The
 *			position of the data to copy is coded as an offset from the
 *			actual output position.
 *
 *			The offset is in the upper nibble of T1 and in T2.
 *			The length is in the lower nibble of T1.
 *
 *			So the 16 bits of a 2 byte tag are coded as
 *
 *				7---T1--0  7---T2--0
 *				OOOO LLLL  OOOO OOOO
 *
 *			This limits the offset to 1-4095 (12 bits) and the length
 *			to 3-18 (4 bits) because 3 is always added to it. To emit
 *			a tag of 2 bytes with a length of 2 only saves one control
 *			bit. But we lose one byte in the possible length of a tag.
 *
 *			In the actual implementation, the 2 byte tag's length is
 *			limited to 3-17, because the value 0xF in the length nibble
 *			has special meaning. It means, that the next following
 *			byte (T3) has to be added to the length value of 18. That
 *			makes total limits of 1-4095 for offset and 3-273 for length.
 *
 *			Now that we have successfully decoded a tag. We simply copy
 *			the output that occurred <offset> bytes back to the current
 *			output location in the specified <length>. Thus, a
 *			sequence of 200 spaces (think about bpchar fields) could be
 *			coded in 4 bytes. One literal space and a three byte tag to
 *			copy 199 bytes with a -1 offset. Whow - that's a compression
 *			rate of 98%! Well, the implementation needs to save the
 *			original data size too, so we need another 4 bytes for it
 *			and end up with a total compression rate of 96%, what's still
 *			worth a Whow.
 *
 *		The compression algorithm
 *
 *			The following uses numbers used in the default strategy.
 *
 *			The compressor works best for attributes of a size between
 *			1K and 1M. For smaller items there's not that much chance of
 *			redundancy in the character sequence (except for large areas
 *			of identical bytes like trailing spaces) and for bigger ones
 *			our 4K maximum look-back distance is too small.
 *
 *			The compressor creates a table for 8192 lists of positions.
 *			For each input position (except the last 3), a hash key is
 *			built from the 4 next input bytes and the position remembered
 *			in the appropriate list. Thus, the table points to linked
 *			lists of likely to be at least in the first 4 characters
 *			matching strings. This is done on the fly while the input
 *			is compressed into the output area.  Table entries are only
 *			kept for the last 4096 input positions, since we cannot use
 *			back-pointers larger than that anyway.
 *
 *			For each byte in the input, it's hash key (built from this
 *			byte and the next 3) is used to find the appropriate list
 *			in the table. The lists remember the positions of all bytes
 *			that had the same hash key in the past in increasing backward
 *			offset order. Now for all entries in the used lists, the
 *			match length is computed by comparing the characters from the
 *			entries position with the characters from the actual input
 *			position.
 *
 *			The compressor starts with a so called "good_match" of 128.
 *			It is a "prefer speed against compression ratio" optimizer.
 *			So if the first entry looked at already has 128 or more
 *			matching characters, the lookup stops and that position is
 *			used for the next tag in the output.
 *
 *			For each subsequent entry in the history list, the "good_match"
 *			is lowered by 10%. So the compressor will be more happy with
 *			short matches the farer it has to go back in the history.
 *			Another "speed against ratio" preference characteristic of
 *			the algorithm.
 *
 *			Thus there are 3 stop conditions for the lookup of matches:
 *
 *				- a match >= good_match is found
 *				- there are no more history entries to look at
 *				- the next history entry is already too far back
 *				  to be coded into a tag.
 *
 *			Finally the match algorithm checks that at least a match
 *			of 3 or more bytes has been found, because thats the smallest
 *			amount of copy information to code into a tag. If so, a tag
 *			is omitted and all the input bytes covered by that are just
 *			scanned for the history add's, otherwise a literal character
 *			is omitted and only his history entry added.
 *
 *		Acknowledgements:
 *
 *			Many thanks to Adisak Pochanayon, who's article about SLZ
 *			inspired me to write the PostgreSQL compression this way.
 *
 *			Jan Wieck
 *
 * Copyright (c) 1999-2012, PostgreSQL Global Development Group
 *
 * src/backend/utils/adt/pg_lzcompress.c
 * ----------
 */
#include "postgres.h"
#include "knl/knl_variable.h"

#include <limits.h>

#include "utils/pg_lzcompress.h"

/* ----------
 * The provided standard strategies
 * ----------
 */
static const PGLZ_Strategy strategy_default_data = {
    32,      /* Data chunks less than 32 bytes are not
              * compressed */
    INT_MAX, /* No upper limit on what we'll try to
              * compress */
    25,      /* Require 25% compression rate, or not worth
              * it */
    1024,    /* Give up if no compression in the first 1KB */
    128,     /* Stop history lookup if a match of 128 bytes
              * is found */
    10       /* Lower good match size by 10% at every loop
              * iteration */
};
const PGLZ_Strategy* const PGLZ_strategy_default = &strategy_default_data;

static const PGLZ_Strategy strategy_always_data = {
    0, /* Chunks of any size are compressed */
    INT_MAX,
    0,       /* It's enough to save one single byte */
    INT_MAX, /* Never give up early */
    128,     /* Stop history lookup if a match of 128 bytes
              * is found */
    6        /* Look harder for a good match */
};
const PGLZ_Strategy* const PGLZ_strategy_always = &strategy_always_data;

/* ----------
 * pglz_hist_idx -
 *
 *		Computes the history table slot for the lookup by the next 4
 *		characters in the input.
 *
 * NB: because we use the next 4 characters, we are not guaranteed to
 * find 3-character matches; they very possibly will be in the wrong
 * hash list.  This seems an acceptable tradeoff for spreading out the
 * hash keys more.
 * ----------
 */
#define pglz_hist_idx(_s, _e)                                                                    \
    (((((_e) - (_s)) < 4) ? (int)(_s)[0]                                                         \
                          : (((unsigned char)((_s)[0]) << 9) ^ ((unsigned char)((_s)[1]) << 6) ^ \
                                ((unsigned char)((_s)[2]) << 3) ^ (unsigned char)((_s)[3]))) &   \
        (PGLZ_HISTORY_MASK))

/* ----------
 * pglz_hist_add -
 *
 *		Adds a new entry to the history table.
 *
 * If _recycle is true, then we are recycling a previously used entry,
 * and must first delink it from its old hashcode's linked list.
 *
 * NOTE: beware of multiple evaluations of macro's arguments, and note that
 * _hn and _recycle are modified in the macro.
 * ----------
 */
#define pglz_hist_add(_hs, _he, _hn, _recycle, _s, _e) \
    do {                                               \
        int __hindex = pglz_hist_idx((_s), (_e));      \
        PGLZ_HistEntry** __myhsp = &(_hs)[__hindex];   \
        PGLZ_HistEntry* __myhe = &(_he)[_hn];          \
        if (_recycle) {                                \
            if (__myhe->prev == NULL)                  \
                (_hs)[__myhe->hindex] = __myhe->next;  \
            else                                       \
                __myhe->prev->next = __myhe->next;     \
            if (__myhe->next != NULL)                  \
                __myhe->next->prev = __myhe->prev;     \
        }                                              \
        __myhe->next = *__myhsp;                       \
        __myhe->prev = NULL;                           \
        __myhe->hindex = __hindex;                     \
        __myhe->pos = (_s);                            \
        if (*__myhsp != NULL)                          \
            (*__myhsp)->prev = __myhe;                 \
        *__myhsp = __myhe;                             \
        if (++(_hn) >= PGLZ_HISTORY_SIZE) {            \
            (_hn) = 0;                                 \
            (_recycle) = true;                         \
        }                                              \
    } while (0)

/* ----------
 * pglz_out_ctrl -
 *
 *		Outputs the last and allocates a new control byte if needed.
 * ----------
 */
#define pglz_out_ctrl(__ctrlp, __ctrlb, __ctrl, __buf) \
    do {                                               \
        if ((((unsigned char)(__ctrl)) & 0xff) == 0) { \
            *(__ctrlp) = __ctrlb;                      \
            __ctrlp = (__buf)++;                       \
            __ctrlb = 0;                               \
            __ctrl = 1;                                \
        }                                              \
    } while (0)

/* ----------
 * pglz_out_literal -
 *
 *		Outputs a literal byte to the destination buffer including the
 *		appropriate control bit.
 * ----------
 */
#define pglz_out_literal(_ctrlp, _ctrlb, _ctrl, _buf, _byte) \
    do {                                                     \
        pglz_out_ctrl(_ctrlp, _ctrlb, _ctrl, _buf);          \
        *(_buf)++ = (unsigned char)(_byte);                  \
        (_ctrl) <<= 1;                                       \
    } while (0)

/* ----------
 * pglz_out_tag -
 *
 *		Outputs a backward reference tag of 2-4 bytes (depending on
 *		offset and length) to the destination buffer including the
 *		appropriate control bit.
 * ----------
 */
#define pglz_out_tag(_ctrlp, _ctrlb, _ctrl, _buf, _len, _off)                                \
    do {                                                                                     \
        pglz_out_ctrl(_ctrlp, _ctrlb, _ctrl, _buf);                                          \
        (_ctrlb) |= (_ctrl);                                                                 \
        (_ctrl) <<= 1;                                                                       \
        if ((_len) > 17) {                                                                   \
            (_buf)[0] = (unsigned char)((((uint32)(_off)&0xf00) >> 4) | 0x0f);               \
            (_buf)[1] = (unsigned char)(((uint32)(_off)&0xff));                              \
            (_buf)[2] = (unsigned char)((_len)-18);                                          \
            (_buf) += 3;                                                                     \
        } else {                                                                             \
            (_buf)[0] = (unsigned char)((((uint32)(_off)&0xf00) >> 4) | ((uint32)(_len)-3)); \
            (_buf)[1] = (unsigned char)((uint32)(_off)&0xff);                                \
            (_buf) += 2;                                                                     \
        }                                                                                    \
    } while (0)

#define HIST_START_LEN (sizeof(PGLZ_HistEntry*) * PGLZ_HISTORY_LISTS)
#define HIST_ENTRIES_LEN (sizeof(PGLZ_HistEntry) * PGLZ_HISTORY_SIZE)

#define PGLZ_MAX_HISTORY_LISTS    8192    /* must be power of 2 */
static PGLZ_HistEntry* hist_start[PGLZ_MAX_HISTORY_LISTS];
static PGLZ_HistEntry hist_entries[PGLZ_HISTORY_SIZE + 1];

/* ----------
 * pglz_find_match -
 *
 *		Lookup the history table if the actual input stream matches
 *		another sequence of characters, starting somewhere earlier
 *		in the input buffer.
 * ----------
 */
static inline int pglz_find_match(
    PGLZ_HistEntry** hstart, const char* input, const char* end, int* lenp, int* offp, int good_match, int good_drop)
{
    PGLZ_HistEntry* hent = NULL;
    int32 len = 0;
    int32 off = 0;

    /*
     * Traverse the linked history list until a good enough match is found.
     */
    hent = hstart[pglz_hist_idx(input, end)];
    while (hent != NULL) {
        const char* ip = input;
        const char* hp = hent->pos;
        int32 thisoff;
        int32 thislen;

        /*
         * Stop if the offset does not fit into our tag anymore.
         */
        thisoff = ip - hp;
        if (thisoff >= 0x0fff)
            break;

        /*
         * Determine length of match. A better match must be larger than the
         * best so far. And if we already have a match of 16 or more bytes,
         * it's worth the call overhead to use memcmp() to check if this match
         * is equal for the same size. After that we must fallback to
         * character by character comparison to know the exact position where
         * the diff occurred.
         */
        thislen = 0;
        if (len >= 16) {
            if (memcmp(ip, hp, len) == 0) {
                thislen = len;
                ip += len;
                hp += len;
                while (ip < end && *ip == *hp && thislen < PGLZ_MAX_MATCH) {
                    thislen++;
                    ip++;
                    hp++;
                }
            }
        } else {
            while (ip < end && *ip == *hp && thislen < PGLZ_MAX_MATCH) {
                thislen++;
                ip++;
                hp++;
            }
        }

        /*
         * Remember this match as the best (if it is)
         */
        if (thislen > len) {
            len = thislen;
            off = thisoff;
        }

        /*
         * Advance to the next history entry
         */
        hent = hent->next;

        /*
         * Be happy with lesser good matches the more entries we visited. But
         * no point in doing calculation if we're at end of list.
         */
        if (hent != NULL) {
            if (len >= good_match)
                break;
            good_match -= (good_match * good_drop) / 100;
        }
    }

    /*
     * Return match information only if it results at least in one byte
     * reduction.
     */
    if (len > 2) {
        *lenp = len;
        *offp = off;
        return 1;
    }

    return 0;
}

/* ----------
 * lz_compress -
 *
 *		Compresses source into dest using strategy. Returns the number of
 *		bytes written in buffer dest, or -1 if compression fails.
 * ----------
 */
int32 lz_compress(const char* source, int32 slen, char* dest)
{
    unsigned char* bp = (unsigned char*) dest;
    unsigned char* bstart = bp;
    int hist_next = 0;
    bool hist_recycle = false;
    const char* dp = source;
    const char* dend = source + slen;
    unsigned char ctrl_dummy = 0;
    unsigned char* ctrlp = &ctrl_dummy;
    unsigned char ctrlb = 0;
    unsigned char ctrl = 0;
    bool found_match = false;
    int32 match_len;
    int32 match_off;
    int32 good_match;
    int32 good_drop;
    int32 result_size;
    int32 result_max;
    int32 need_rate;
    errno_t rc;

    const PGLZ_Strategy* strategy = PGLZ_strategy_always;
    /*
     * Our fallback strategy is the default.
     */
    if (strategy == NULL) {
        strategy = PGLZ_strategy_default;
    }

    /*
     * If the strategy forbids compression (at all or if source chunk size out
     * of range), fail.
     */
    if (strategy->match_size_good <= 0 || slen < strategy->min_input_size || slen > strategy->max_input_size) {
        return -1;
    }

    /*
     * Limit the match parameters to the supported range.
     */
    good_match = strategy->match_size_good;
    if (good_match > PGLZ_MAX_MATCH) {
        good_match = PGLZ_MAX_MATCH;
    } else if (good_match < 17) {
        good_match = 17;
    }

    good_drop = strategy->match_size_drop;
    if (good_drop < 0) {
        good_drop = 0;
    } else if (good_drop > 100) {
        good_drop = 100;
    }

    need_rate = strategy->min_comp_rate;
    if (need_rate < 0) {
        need_rate = 0;
    } else if (need_rate > 99) {
        need_rate = 99;
    }

    /*
     * Compute the maximum result size allowed by the strategy, namely the
     * input size minus the minimum wanted compression rate.  This had better
     * be <= slen, else we might overrun the provided output buffer.
     */
    if (slen > (INT_MAX / 100)) {
        /* Approximate to avoid overflow */
        result_max = (slen / 100) * (100 - need_rate);
    } else {
        result_max = (slen * (100 - need_rate)) / 100;
    }

    /*
     * Initialize the history lists to empty.  We do not need to zero the
     * hist_entries[] array; its entries are initialized as they are used.
     */
    rc = memset_s(hist_start, HIST_START_LEN, 0, HIST_START_LEN);
    securec_check(rc, "\0", "\0");

    /*
     * Compress the source directly into the output buffer.
     */
    while (dp < dend) {
        /*
         * If we already exceeded the maximum result size, fail.
         *
         * We check once per loop; since the loop body could emit as many as 4
         * bytes (a control byte and 3-byte tag), PGLZ_MAX_OUTPUT() had better
         * allow 4 slop bytes.
         */
        if (bp - bstart >= result_max) {
            return -1;
        }

        /*
         * If we've emitted more than first_success_by bytes without finding
         * anything compressible at all, fail.  This lets us fall out
         * reasonably quickly when looking at incompressible input (such as
         * pre-compressed data).
         */
        if (!found_match && bp - bstart >= strategy->first_success_by) {
            return -1;
        }

        /*
         * Try to find a match in the history
         */
        if (pglz_find_match(hist_start, dp, dend, &match_len, &match_off, good_match, good_drop)) {
            /*
             * Create the tag and add history entries for all matched
             * characters.
             */
            pglz_out_tag(ctrlp, ctrlb, ctrl, bp, match_len, match_off);
            while (match_len--) {
                pglz_hist_add(
                    hist_start, hist_entries, hist_next, hist_recycle, dp,
                    dend);
                dp++; /* Do not do this ++ in the line above! */
                /* The macro would do it four times - Jan.  */
            }
            found_match = true;
        } else {
            /*
             * No match found. Copy one literal byte.
             */
            pglz_out_literal(ctrlp, ctrlb, ctrl, bp, *dp);
            pglz_hist_add(
                hist_start, hist_entries, hist_next, hist_recycle, dp, dend);
            dp++; /* Do not do this ++ in the line above! */
            /* The macro would do it four times - Jan.  */
        }
    }

    /*
     * Write out the last control byte and check that we haven't overrun the
     * output size allowed by the strategy.
     */
    *ctrlp = ctrlb;
    result_size = bp - bstart;
    if (result_size >= result_max) {
        return -1;
    }

    /* success */
    return result_size;
}

/* ----------
 * pglz_decompress -
 *
 *		Decompresses source into dest. Returns the number of bytes
 *		decompressed in the destination buffer, and *optionally*
 *		checks that both the source and dest buffers have been
 *		fully read and written to, respectively.
 * ----------
 */
int32 lz_decompress(const char* source, int32 slen, char* dest, int32 rawsize, bool check_complete)
{
    const unsigned char* sp;
    const unsigned char* srcend;
    unsigned char* dp;
    unsigned char* destend;
    errno_t rc = 0;

    sp = (const unsigned char*) source;
    srcend = ((const unsigned char*) source) + slen;
    dp = (unsigned char*) dest;
    destend = dp + rawsize;

    while (sp < srcend && dp < destend) {
        /*
         * Read one control byte and process the next 8 items (or as many as
         * remain in the compressed input).
         */
        unsigned char ctrl = *sp++;
        int ctrlc;

        for (ctrlc = 0; ctrlc < 8 && sp < srcend && dp < destend; ctrlc++) {

            if (ctrl & 1) {
                /*
                 * Set control bit means we must read a match tag. The match
                 * is coded with two bytes. First byte uses lower nibble to
                 * code length - 3. Higher nibble contains upper 4 bits of the
                 * offset. The next following byte contains the lower 8 bits
                 * of the offset. If the length is coded as 18, another
                 * extension tag byte tells how much longer the match really
                 * was (0-255).
                 */
                int32 len;
                int32 off;

                len = (sp[0] & 0x0f) + 3;
                off = ((sp[0] & 0xf0) << 4) | sp[1];
                sp += 2;
                if (len == 18) {
                    len += *sp++;
                }

                /*
                 * Now we copy the bytes specified by the tag from OUTPUT to
                 * OUTPUT (copy len bytes from dp - off to dp). The copied
                 * areas could overlap, to preven possible uncertainty, we
                 * copy only non-overlapping regions.
                 */
                len = Min(len, destend - dp);
                while (off < len) {
                    /*---------
                     * When offset is smaller than length - source and
                     * destination regions overlap. memmove() is resolving
                     * this overlap in an incompatible way with pglz. Thus we
                     * resort to memcpy()-ing non-overlapping regions.
                     *
                     * Consider input: 112341234123412341234
                     * At byte 5       here ^ we have match with length 16 and
                     * offset 4.       11234M(len=16, off=4)
                     * We are decoding first period of match and rewrite match
                     *                 112341234M(len=12, off=8)
                     *
                     * The same match is now at position 9, it points to the
                     * same start byte of output, but from another position:
                     * the offset is doubled.
                     *
                     * We iterate through this offset growth until we can
                     * proceed to usual memcpy(). If we would try to decode
                     * the match at byte 5 (len=16, off=4) by memmove() we
                     * would issue memmove(5, 1, 16) which would produce
                     * 112341234XXXXXXXXXXXX, where series of X is 12
                     * undefined bytes, that were at bytes [5:17].
                     * ---------
                     */
                    errno_t rc = memcpy_s(dp, off + 1, dp - off, off);
                    securec_check(rc, "", "");
                    len -= off;
                    dp += off;
                    off += off;
                }
                rc = memcpy_s(dp, len + 1, dp - off, len);
                securec_check(rc, "", "");
                dp += len;
            } else {
                /*
                 * An unset control bit means LITERAL BYTE. So we just copy
                 * one from INPUT to OUTPUT.
                 */
                *dp++ = *sp++;
            }

            /*
             * Advance the control bit
             */
            ctrl >>= 1;
        }
    }

    /*
     * Check we decompressed the right amount. If we are slicing, then we
     * won't necessarily be at the end of the source or dest buffers when we
     * hit a stop, so we don't test them.
     */
    if (check_complete && (dp != destend || sp != srcend)) {
        return -1;
    }

    /*
     * That's it.
     */
    return (char*) dp - dest;
}


int CompressPage(const char* src, char* dst, int dst_size, RelFileCompressOption option)
{
    if (PageIs8BXidHeapVersion(src)) {
        return TemplateCompressPage<true>(src, dst, dst_size, option);
    } else {
        return TemplateCompressPage<false>(src, dst, dst_size, option);
    }
}

void CompressConvertRows(char *buf, char *aux_buf, int16 *real_order, uint16 max_row_len, uint16 real_row_cnt) {
    errno_t ret;
    HeapPageHeaderData *page = (HeapPageHeaderData *)buf;
    uint16 row_cnt = real_row_cnt;
    uint32 total_size = page->pd_special - page->pd_upper;
    char *copy_begin = buf + page->pd_upper;
    char *row;
    uint16 i, j, k, cur, up, row_size;

    ret = memset_sp(aux_buf, BLCKSZ, 0, BLCKSZ);
    securec_check(ret, "", "");

    k = 0;
    for (i = 0; i < max_row_len; i++) {
        for (j = 0; j < row_cnt; j++) {
            up = (j == (row_cnt - 1)) ? page->pd_special : GET_ITEMID_BY_IDX(buf, (real_order[j + 1]))->lp_off;
            cur = GET_ITEMID_BY_IDX(buf, (real_order[j]))->lp_off;
            row_size = up - cur;
            row = buf + cur;
            if (i < row_size) {
                aux_buf[k++] = row[i];  // this part is reshaped
            }
        }
    }

    if (k != total_size) {
        printf("ERROR!!! convert_rows_2 error...!!!\n");
        ASSERT(0);
        return;
    }

    // cp aux_buf to page_buf
    ret = memcpy_sp(copy_begin, total_size, aux_buf, total_size);
    securec_check(ret, "", "");
    return ;
}

void CompressConvertItemIds(char *buf, char *aux_buf) {
    errno_t ret;
    HeapPageHeaderData *page = (HeapPageHeaderData *)buf;
    uint16 row_cnt = (page->pd_lower - GetPageHeaderSize(page)) / sizeof(ItemIdData);
    uint32 total_size = row_cnt * sizeof(ItemIdData);
    char *copy_begin = buf + GetPageHeaderSize(page);
    uint16 i, j, k;

    // clear aux_buf
    ret = memset_sp(aux_buf, BLCKSZ, 0, BLCKSZ);
    securec_check(ret, "", "");

    k = 0;
    for (i = 0; i < row_cnt; i++) {
        for (j = 0; j < sizeof(ItemIdData); j++) {
            aux_buf[j  * row_cnt + i] = copy_begin[k++];
        }
    }

    // cp aux_buf to page_buf
    ret = memcpy_sp(copy_begin, total_size, aux_buf, total_size);
    securec_check(ret, "", "");
    return ;
}

void cprs_diff_convert_rows(char *buf, uint32 offset,uint16 min_row_len, uint16 real_row_cnt) {
    uint16 row_cnt = real_row_cnt;
    uint32 common_size = min_row_len;
    uint8 *copy_begin = (uint8 *)(buf + offset);
    uint16 i, j;

    for (i = 0; i < common_size; i++) {
        for (j = row_cnt - 1; j > 0; j--) {
            copy_begin[i  * row_cnt + j] -= copy_begin[i  * row_cnt + (j - 1)];
        }
    }
    return ;
}

bool CompressConvertOnePage(char *buf, char *aux_buf, bool diff_convert) {
    uint16 max_row_len = 0;
    uint16 min_row_len = 0;
    int16 *real_order = NULL; // itemids are not in order sometimes. we must find the real
    uint16 real_row_cnt = 0;
    if (!CompressConvertCheck(buf, &real_order, &max_row_len, &min_row_len, &real_row_cnt)) {
        if (real_order != NULL) {
            free(real_order);
        }
        return false;
    }

    CompressConvertRows(buf, aux_buf, real_order, max_row_len, real_row_cnt);
    CompressConvertItemIds(buf, aux_buf);

    if (diff_convert) {
        cprs_diff_convert_rows(buf, ((HeapPageHeaderData *)buf)->pd_upper, min_row_len, real_row_cnt);
        cprs_diff_convert_rows(buf, GetPageHeaderSize(buf), sizeof(ItemIdData),
                               (((HeapPageHeaderData *)buf)->pd_lower - GetPageHeaderSize(buf)) / sizeof(ItemIdData));
    }

    if (real_order != NULL) {
        free(real_order);
    }
    return true;
}

void CompressPagePrepareConvert(char *src, bool diff_convert, bool *real_ByteConvert)
{
    char *aux_buf = NULL;
    errno_t rc;

    aux_buf = (char *)malloc(BLCKSZ);
    if (aux_buf == NULL) {
        // add log
        return;
    }
    rc = memset_sp(aux_buf, BLCKSZ, 0, BLCKSZ);
    securec_check(rc, "", "");

    // do convert
    *real_ByteConvert = false;
    if (CompressConvertOnePage(src, aux_buf, diff_convert)) {
        *real_ByteConvert = true;
    }

    if (aux_buf != NULL) {
        free(aux_buf);
    }
}


/**
 * CompressPage() -- Compress one page.
 *
 *		Only the parts other than the page header will be compressed. The
 *		compressed data is rounded by chunck_size, The insufficient part is
 *		filled with zero.  Compression needs to be able to save at least one
 *		chunk of space, otherwise it fail.
 *		This function returen the size of compressed data or
 *		    -1 for compression fail
 *		    COMPRESS_UNSUPPORTED_ERROR for unrecognized compression algorithm
 */
template<bool heapPageData>
int TemplateCompressPage(const char* src, char* dst, int dst_size, RelFileCompressOption option)
{
    int compressed_size;
    int8 level = option.compressLevelSymbol ? option.compressLevel : -option.compressLevel;
    size_t sizeOfHeaderData = GetSizeOfHeadData(heapPageData);
    char *src_copy = NULL;
    bool real_ByteConvert = false;
    errno_t rc;
    char* data;

    if (option.byteConvert) {
        // copy and maybe change it
        src_copy = (char *)malloc(BLCKSZ);
        if (src_copy == NULL) {
            // add log
            return -1;
        }
        rc = memcpy_s(src_copy, BLCKSZ, src, BLCKSZ);
        securec_check(rc, "", "");
        CompressPagePrepareConvert(src_copy, option.diffConvert, &real_ByteConvert);  /* preprocess convert src */
    }

    if (heapPageData) {
        data = ((HeapPageCompressData*)dst)->data;
    } else {
        data = ((PageCompressData*)dst)->data;
    }

    switch (option.compressAlgorithm) {
        case COMPRESS_ALGORITHM_PGLZ:
            if (real_ByteConvert) {
                compressed_size = lz_compress(src_copy + sizeOfHeaderData, BLCKSZ - sizeOfHeaderData, data);
            } else {
                compressed_size = lz_compress(src + sizeOfHeaderData, BLCKSZ - sizeOfHeaderData, data);
            }
            break;
        case COMPRESS_ALGORITHM_ZSTD: {
            if (level == 0 || level < MIN_ZSTD_COMPRESSION_LEVEL || level > MAX_ZSTD_COMPRESSION_LEVEL) {
                level = DEFAULT_ZSTD_COMPRESSION_LEVEL;
            }

            if (real_ByteConvert) {
                compressed_size = ZSTD_compress(data, dst_size, src_copy + sizeOfHeaderData, BLCKSZ - sizeOfHeaderData, level);
            } else {
                compressed_size = ZSTD_compress(data, dst_size, src + sizeOfHeaderData, BLCKSZ - sizeOfHeaderData, level);
            }

            if (ZSTD_isError(compressed_size)) {
                if (src_copy != NULL) {
                    free(src_copy);
                }
                return -1;
            }
            break;
        }
        default:
            if (src_copy != NULL) {
                free(src_copy);
            }
            return COMPRESS_UNSUPPORTED_ERROR;
    }

    if (compressed_size < 0) {
        if (src_copy != NULL) {
            free(src_copy);
        }
        return -1;
    }

    if (heapPageData) {
        HeapPageCompressData* pcdptr = ((HeapPageCompressData*)dst);
        rc = memcpy_s(pcdptr->page_header, sizeOfHeaderData, src, sizeOfHeaderData);
        securec_check(rc, "", "");
        pcdptr->size = compressed_size;
        pcdptr->byte_convert = real_ByteConvert;
        pcdptr->diff_convert = option.diffConvert;
    } else {
        PageCompressData* pcdptr = ((PageCompressData*)dst);
        rc = memcpy_s(pcdptr->page_header, sizeOfHeaderData, src, sizeOfHeaderData);
        securec_check(rc, "", "");
        pcdptr->size = compressed_size;
        pcdptr->byte_convert = real_ByteConvert;
        pcdptr->diff_convert = option.diffConvert;
    }

    if (src_copy != NULL) {
        free(src_copy);
    }
    return SIZE_OF_PAGE_COMPRESS_DATA_HEADER_DATA(heapPageData) + compressed_size;
}

/**
 * CompressPageBufferBound()
 * -- Get the destination buffer boundary to compress one page.
 * Return needed destination buffer size for compress one page or
 *     -1 for unrecognized compression algorithm
 */
int CompressPageBufferBound(const char* page, uint8 algorithm)
{
    switch (algorithm) {
        case COMPRESS_ALGORITHM_PGLZ:
            return BLCKSZ + 4;
        case COMPRESS_ALGORITHM_ZSTD:
            return ZSTD_compressBound(BLCKSZ - GetPageHeaderSize(page));
        default:
            return -1;
    }
}