/*
* Make array of trigrams without sorting and removing duplicate items.
*
* trg: where to return the array of trigrams.
* str: source string, of length slen bytes.
* bounds: where to return bounds of trigrams (if needed).
*
* Returns length of the generated array.
*/
static int
generate_trgm_only(trgm *trg, char *str, int slen, TrgmBound *bounds)
{
trgm *tptr;
char *buf;
int charlen,
bytelen;
char *bword,
*eword;
if (slen + LPADDING + RPADDING < 3 || slen == 0)
return 0;
tptr = trg;
/* Allocate a buffer for case-folded, blank-padded words */
buf = (char *) palloc(slen * pg_database_encoding_max_length() + 4);
if (LPADDING > 0)
{
*buf = ' ';
if (LPADDING > 1)
*(buf + 1) = ' ';
}
eword = str;
while ((bword = find_word(eword, slen - (eword - str), &eword, &charlen)) != NULL)
{
#ifdef IGNORECASE
bword = lowerstr_with_len(bword, eword - bword);
bytelen = strlen(bword);
#else
bytelen = eword - bword;
#endif
memcpy(buf + LPADDING, bword, bytelen);
#ifdef IGNORECASE
pfree(bword);
#endif
buf[LPADDING + bytelen] = ' ';
buf[LPADDING + bytelen + 1] = ' ';
/* Calculate trigrams marking their bounds if needed */
if (bounds)
bounds[tptr - trg] |= TRGM_BOUND_LEFT;
tptr = make_trigrams(tptr, buf, bytelen + LPADDING + RPADDING,
charlen + LPADDING + RPADDING);
if (bounds)
bounds[tptr - trg - 1] |= TRGM_BOUND_RIGHT;
}
pfree(buf);
return tptr - trg;
}
/*
* Generates trigrams for wildcard search string.
*
* Returns array of trigrams that must occur in any string that matches the
* wildcard string. For example, given pattern "a%bcd%" the trigrams
* " a", "bcd" would be extracted.
*/
TRGM *
generate_wildcard_trgm(const char *str, int slen)
{
TRGM *trg;
char *buf,
*buf2;
trgm *tptr;
int len,
charlen,
bytelen;
const char *eword;
trg = (TRGM *) palloc(TRGMHDRSIZE + sizeof(trgm) * (slen / 2 + 1) *3);
trg->flag = ARRKEY;
SET_VARSIZE(trg, TRGMHDRSIZE);
if (slen + LPADDING + RPADDING < 3 || slen == 0)
return trg;
tptr = GETARR(trg);
buf = palloc(sizeof(char) * (slen + 4));
/*
* Extract trigrams from each substring extracted by get_wildcard_part.
*/
eword = str;
while ((eword = get_wildcard_part(eword, slen - (eword - str),
buf, &bytelen, &charlen)) != NULL)
{
#ifdef IGNORECASE
buf2 = lowerstr_with_len(buf, bytelen);
bytelen = strlen(buf2);
#else
buf2 = buf;
#endif
/*
* count trigrams
*/
tptr = make_trigrams(tptr, buf2, bytelen, charlen);
#ifdef IGNORECASE
pfree(buf2);
#endif
}
pfree(buf);
if ((len = tptr - GETARR(trg)) == 0)
return trg;
/*
* Make trigrams unique.
*/
if (len > 0)
{
qsort((void *) GETARR(trg), len, sizeof(trgm), comp_trgm);
len = unique_array(GETARR(trg), len);
}
SET_VARSIZE(trg, CALCGTSIZE(ARRKEY, len));
return trg;
}
TRGM *
generate_trgm(char *str, int slen)
{
TRGM *trg;
char *buf;
trgm *tptr;
int len,
charlen,
bytelen;
char *bword,
*eword;
trg = (TRGM *) palloc(TRGMHDRSIZE + sizeof(trgm) * (slen / 2 + 1) *3);
trg->flag = ARRKEY;
SET_VARSIZE(trg, TRGMHDRSIZE);
if (slen + LPADDING + RPADDING < 3 || slen == 0)
return trg;
tptr = GETARR(trg);
buf = palloc(sizeof(char) * (slen + 4));
if (LPADDING > 0)
{
*buf = ' ';
if (LPADDING > 1)
*(buf + 1) = ' ';
}
eword = str;
while ((bword = find_word(eword, slen - (eword - str), &eword, &charlen)) != NULL)
{
#ifdef IGNORECASE
bword = lowerstr_with_len(bword, eword - bword);
bytelen = strlen(bword);
#else
bytelen = eword - bword;
#endif
memcpy(buf + LPADDING, bword, bytelen);
#ifdef IGNORECASE
pfree(bword);
#endif
buf[LPADDING + bytelen] = ' ';
buf[LPADDING + bytelen + 1] = ' ';
/*
* count trigrams
*/
tptr = make_trigrams(tptr, buf, bytelen + LPADDING + RPADDING,
charlen + LPADDING + RPADDING);
}
pfree(buf);
if ((len = tptr - GETARR(trg)) == 0)
return trg;
if (len > 0)
{
qsort((void *) GETARR(trg), len, sizeof(trgm), comp_trgm);
len = unique_array(GETARR(trg), len);
}
SET_VARSIZE(trg, CALCGTSIZE(ARRKEY, len));
return trg;
}
/*
* lowerstr --- fold null-terminated string to lower case
*
* Returned string is palloc'd
*/
char *
lowerstr(const char *str)
{
return lowerstr_with_len(str, strlen(str));
}
/*
* Generates trigrams for wildcard search string.
*
* Returns array of trigrams that must occur in any string that matches the
* wildcard string. For example, given pattern "a%bcd%" the trigrams
* " a", "bcd" would be extracted.
*/
TRGM *
generate_wildcard_trgm(const char *str, int slen)
{
TRGM *trg;
char *buf,
*buf2;
trgm *tptr;
int len,
charlen,
bytelen;
const char *eword;
/*
* Guard against possible overflow in the palloc requests below. (We
* don't worry about the additive constants, since palloc can detect
* requests that are a little above MaxAllocSize --- we just need to
* prevent integer overflow in the multiplications.)
*/
if ((Size) (slen / 2) >= (MaxAllocSize / (sizeof(trgm) * 3)) ||
(Size) slen >= (MaxAllocSize / pg_database_encoding_max_length()))
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("out of memory")));
trg = (TRGM *) palloc(TRGMHDRSIZE + sizeof(trgm) * (slen / 2 + 1) *3);
trg->flag = ARRKEY;
SET_VARSIZE(trg, TRGMHDRSIZE);
if (slen + LPADDING + RPADDING < 3 || slen == 0)
return trg;
tptr = GETARR(trg);
/* Allocate a buffer for blank-padded, but not yet case-folded, words */
buf = palloc(sizeof(char) * (slen + 4));
/*
* Extract trigrams from each substring extracted by get_wildcard_part.
*/
eword = str;
while ((eword = get_wildcard_part(eword, slen - (eword - str),
buf, &bytelen, &charlen)) != NULL)
{
#ifdef IGNORECASE
buf2 = lowerstr_with_len(buf, bytelen);
bytelen = strlen(buf2);
#else
buf2 = buf;
#endif
/*
* count trigrams
*/
tptr = make_trigrams(tptr, buf2, bytelen, charlen);
#ifdef IGNORECASE
pfree(buf2);
#endif
}
pfree(buf);
if ((len = tptr - GETARR(trg)) == 0)
return trg;
/*
* Make trigrams unique.
*/
if (len > 1)
{
qsort((void *) GETARR(trg), len, sizeof(trgm), comp_trgm);
len = unique_array(GETARR(trg), len);
}
SET_VARSIZE(trg, CALCGTSIZE(ARRKEY, len));
return trg;
}
TRGM *
generate_trgm(char *str, int slen)
{
TRGM *trg;
char *buf;
trgm *tptr;
int len,
charlen,
bytelen;
char *bword,
*eword;
/*
* Guard against possible overflow in the palloc requests below. (We
* don't worry about the additive constants, since palloc can detect
* requests that are a little above MaxAllocSize --- we just need to
* prevent integer overflow in the multiplications.)
*/
if ((Size) (slen / 2) >= (MaxAllocSize / (sizeof(trgm) * 3)) ||
(Size) slen >= (MaxAllocSize / pg_database_encoding_max_length()))
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("out of memory")));
trg = (TRGM *) palloc(TRGMHDRSIZE + sizeof(trgm) * (slen / 2 + 1) *3);
trg->flag = ARRKEY;
SET_VARSIZE(trg, TRGMHDRSIZE);
if (slen + LPADDING + RPADDING < 3 || slen == 0)
return trg;
tptr = GETARR(trg);
/* Allocate a buffer for case-folded, blank-padded words */
buf = (char *) palloc(slen * pg_database_encoding_max_length() + 4);
if (LPADDING > 0)
{
*buf = ' ';
if (LPADDING > 1)
*(buf + 1) = ' ';
}
eword = str;
while ((bword = find_word(eword, slen - (eword - str), &eword, &charlen)) != NULL)
{
#ifdef IGNORECASE
bword = lowerstr_with_len(bword, eword - bword);
bytelen = strlen(bword);
#else
bytelen = eword - bword;
#endif
memcpy(buf + LPADDING, bword, bytelen);
#ifdef IGNORECASE
pfree(bword);
#endif
buf[LPADDING + bytelen] = ' ';
buf[LPADDING + bytelen + 1] = ' ';
/*
* count trigrams
*/
tptr = make_trigrams(tptr, buf, bytelen + LPADDING + RPADDING,
charlen + LPADDING + RPADDING);
}
pfree(buf);
if ((len = tptr - GETARR(trg)) == 0)
return trg;
/*
* Make trigrams unique.
*/
if (len > 1)
{
qsort((void *) GETARR(trg), len, sizeof(trgm), comp_trgm);
len = unique_array(GETARR(trg), len);
}
SET_VARSIZE(trg, CALCGTSIZE(ARRKEY, len));
return trg;
}
