Header And Logo
|
#include "postgres.h"#include "trgm.h"#include "access/skey.h"
Go to the source code of this file.
Data Structures | |
| struct | gtrgm_consistent_cache |
| struct | CACHESIGN |
| struct | SPLITCOST |
Defines | |
| #define | GETENTRY(vec, pos) ((TRGM *) DatumGetPointer((vec)->vector[(pos)].key)) |
| #define | WISH_F(a, b, c) (double)( -(double)(((a)-(b))*((a)-(b))*((a)-(b)))*(c) ) |
Functions | |
| PG_FUNCTION_INFO_V1 (gtrgm_in) | |
| Datum | gtrgm_in (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_out) | |
| Datum | gtrgm_out (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_compress) | |
| Datum | gtrgm_compress (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_decompress) | |
| Datum | gtrgm_decompress (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_consistent) | |
| Datum | gtrgm_consistent (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_distance) | |
| Datum | gtrgm_distance (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_union) | |
| Datum | gtrgm_union (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_same) | |
| Datum | gtrgm_same (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_penalty) | |
| Datum | gtrgm_penalty (PG_FUNCTION_ARGS) |
| PG_FUNCTION_INFO_V1 (gtrgm_picksplit) | |
| Datum | gtrgm_picksplit (PG_FUNCTION_ARGS) |
| static void | makesign (BITVECP sign, TRGM *a) |
| static int32 | cnt_sml_sign_common (TRGM *qtrg, BITVECP sign) |
| static int32 | unionkey (BITVECP sbase, TRGM *add) |
| static int32 | sizebitvec (BITVECP sign) |
| static int | hemdistsign (BITVECP a, BITVECP b) |
| static int | hemdist (TRGM *a, TRGM *b) |
| static void | fillcache (CACHESIGN *item, TRGM *key) |
| static int | comparecost (const void *a, const void *b) |
| static int | hemdistcache (CACHESIGN *a, CACHESIGN *b) |
Variables | |
| static const uint8 | number_of_ones [256] |
| #define GETENTRY | ( | vec, | ||
| pos | ||||
| ) | ((TRGM *) DatumGetPointer((vec)->vector[(pos)].key)) |
Definition at line 27 of file trgm_gist.c.
Referenced by gtrgm_picksplit(), and gtrgm_union().
Definition at line 743 of file trgm_gist.c.
Referenced by gtrgm_picksplit().
Definition at line 181 of file trgm_gist.c.
References ARRNELEM, CPTRGM, GETARR, GETBIT, and HASHVAL.
Referenced by gtrgm_consistent(), and gtrgm_distance().
| static int comparecost | ( | const void * | a, | |
| const void * | b | |||
| ) | [static] |
Definition at line 751 of file trgm_gist.c.
Referenced by gtrgm_picksplit().
Definition at line 732 of file trgm_gist.c.
References CACHESIGN::allistrue, GETSIGN, ISALLTRUE, ISARRKEY, makesign(), and CACHESIGN::sign.
Referenced by gtrgm_picksplit().
| Datum gtrgm_compress | ( | PG_FUNCTION_ARGS | ) |
Definition at line 113 of file trgm_gist.c.
References ALLISTRUE, CALCGTSIZE, DatumGetPointer, DatumGetTextP, FALSE, TRGM::flag, generate_trgm(), GETSIGN, gistentryinit, i, ISALLTRUE, ISSIGNKEY, GISTENTRY::key, GISTENTRY::leafkey, LOOPBYTE, GISTENTRY::offset, GISTENTRY::page, palloc(), PG_GETARG_POINTER, PG_RETURN_POINTER, PointerGetDatum, GISTENTRY::rel, SET_VARSIZE, sign, SIGNKEY, val, VARDATA, VARHDRSZ, and VARSIZE.
{
GISTENTRY *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
GISTENTRY *retval = entry;
if (entry->leafkey)
{ /* trgm */
TRGM *res;
text *val = DatumGetTextP(entry->key);
res = generate_trgm(VARDATA(val), VARSIZE(val) - VARHDRSZ);
retval = (GISTENTRY *) palloc(sizeof(GISTENTRY));
gistentryinit(*retval, PointerGetDatum(res),
entry->rel, entry->page,
entry->offset, FALSE);
}
else if (ISSIGNKEY(DatumGetPointer(entry->key)) &&
!ISALLTRUE(DatumGetPointer(entry->key)))
{
int32 i,
len;
TRGM *res;
BITVECP sign = GETSIGN(DatumGetPointer(entry->key));
LOOPBYTE
{
if ((sign[i] & 0xff) != 0xff)
PG_RETURN_POINTER(retval);
}
len = CALCGTSIZE(SIGNKEY | ALLISTRUE, 0);
res = (TRGM *) palloc(len);
SET_VARSIZE(res, len);
res->flag = SIGNKEY | ALLISTRUE;
retval = (GISTENTRY *) palloc(sizeof(GISTENTRY));
gistentryinit(*retval, PointerGetDatum(res),
entry->rel, entry->page,
entry->offset, FALSE);
}
PG_RETURN_POINTER(retval);
}
| Datum gtrgm_consistent | ( | PG_FUNCTION_ARGS | ) |
Definition at line 199 of file trgm_gist.c.
References ARRNELEM, cnt_sml(), cnt_sml_sign_common(), CPTRGM, createTrgmNFA(), DatumGetPointer, elog, ERROR, generate_trgm(), generate_wildcard_trgm(), GETARR, GETBIT, GETSIGN, GIST_LEAF, gtrgm_consistent_cache::graph, HASHVAL, ILikeStrategyNumber, ISALLTRUE, GISTENTRY::key, LikeStrategyNumber, MAXALIGN, memcmp(), MemoryContextAlloc(), NULL, palloc(), pfree(), PG_GET_COLLATION, PG_GETARG_POINTER, PG_GETARG_TEXT_P, PG_GETARG_UINT16, PG_RETURN_BOOL, gtrgm_consistent_cache::query, RegExpICaseStrategyNumber, RegExpStrategyNumber, sign, SimilarityStrategyNumber, gtrgm_consistent_cache::strategy, trgm_contained_by(), trgm_limit, trgm_presence_map(), gtrgm_consistent_cache::trigrams, trigramsMatchGraph(), VARDATA, VARHDRSZ, and VARSIZE.
{
GISTENTRY *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
text *query = PG_GETARG_TEXT_P(1);
StrategyNumber strategy = (StrategyNumber) PG_GETARG_UINT16(2);
/* Oid subtype = PG_GETARG_OID(3); */
bool *recheck = (bool *) PG_GETARG_POINTER(4);
TRGM *key = (TRGM *) DatumGetPointer(entry->key);
TRGM *qtrg;
bool res;
Size querysize = VARSIZE(query);
gtrgm_consistent_cache *cache;
/*
* We keep the extracted trigrams in cache, because trigram extraction is
* relatively CPU-expensive. When trying to reuse a cached value, check
* strategy number not just query itself, because trigram extraction
* depends on strategy.
*
* The cached structure is a single palloc chunk containing the
* gtrgm_consistent_cache header, then the input query (starting at a
* MAXALIGN boundary), then the TRGM value (also starting at a MAXALIGN
* boundary). However we don't try to include the regex graph (if any) in
* that struct. (XXX currently, this approach can leak regex graphs
* across index rescans. Not clear if that's worth fixing.)
*/
cache = (gtrgm_consistent_cache *) fcinfo->flinfo->fn_extra;
if (cache == NULL ||
cache->strategy != strategy ||
VARSIZE(cache->query) != querysize ||
memcmp((char *) cache->query, (char *) query, querysize) != 0)
{
gtrgm_consistent_cache *newcache;
TrgmPackedGraph *graph = NULL;
Size qtrgsize;
switch (strategy)
{
case SimilarityStrategyNumber:
qtrg = generate_trgm(VARDATA(query),
querysize - VARHDRSZ);
break;
case ILikeStrategyNumber:
#ifndef IGNORECASE
elog(ERROR, "cannot handle ~~* with case-sensitive trigrams");
#endif
/* FALL THRU */
case LikeStrategyNumber:
qtrg = generate_wildcard_trgm(VARDATA(query),
querysize - VARHDRSZ);
break;
case RegExpICaseStrategyNumber:
#ifndef IGNORECASE
elog(ERROR, "cannot handle ~* with case-sensitive trigrams");
#endif
/* FALL THRU */
case RegExpStrategyNumber:
qtrg = createTrgmNFA(query, PG_GET_COLLATION(),
&graph, fcinfo->flinfo->fn_mcxt);
/* just in case an empty array is returned ... */
if (qtrg && ARRNELEM(qtrg) <= 0)
{
pfree(qtrg);
qtrg = NULL;
}
break;
default:
elog(ERROR, "unrecognized strategy number: %d", strategy);
qtrg = NULL; /* keep compiler quiet */
break;
}
qtrgsize = qtrg ? VARSIZE(qtrg) : 0;
newcache = (gtrgm_consistent_cache *)
MemoryContextAlloc(fcinfo->flinfo->fn_mcxt,
MAXALIGN(sizeof(gtrgm_consistent_cache)) +
MAXALIGN(querysize) +
qtrgsize);
newcache->strategy = strategy;
newcache->query = (text *)
((char *) newcache + MAXALIGN(sizeof(gtrgm_consistent_cache)));
memcpy((char *) newcache->query, (char *) query, querysize);
if (qtrg)
{
newcache->trigrams = (TRGM *)
((char *) newcache->query + MAXALIGN(querysize));
memcpy((char *) newcache->trigrams, (char *) qtrg, qtrgsize);
/* release qtrg in case it was made in fn_mcxt */
pfree(qtrg);
}
else
newcache->trigrams = NULL;
newcache->graph = graph;
if (cache)
pfree(cache);
fcinfo->flinfo->fn_extra = (void *) newcache;
cache = newcache;
}
qtrg = cache->trigrams;
switch (strategy)
{
case SimilarityStrategyNumber:
/* Similarity search is exact */
*recheck = false;
if (GIST_LEAF(entry))
{ /* all leafs contains orig trgm */
float4 tmpsml = cnt_sml(key, qtrg);
/* strange bug at freebsd 5.2.1 and gcc 3.3.3 */
res = (*(int *) &tmpsml == *(int *) &trgm_limit || tmpsml > trgm_limit) ? true : false;
}
else if (ISALLTRUE(key))
{ /* non-leaf contains signature */
res = true;
}
else
{ /* non-leaf contains signature */
int32 count = cnt_sml_sign_common(qtrg, GETSIGN(key));
int32 len = ARRNELEM(qtrg);
if (len == 0)
res = false;
else
res = (((((float8) count) / ((float8) len))) >= trgm_limit) ? true : false;
}
break;
case ILikeStrategyNumber:
#ifndef IGNORECASE
elog(ERROR, "cannot handle ~~* with case-sensitive trigrams");
#endif
/* FALL THRU */
case LikeStrategyNumber:
/* Wildcard search is inexact */
*recheck = true;
/*
* Check if all the extracted trigrams can be present in child
* nodes.
*/
if (GIST_LEAF(entry))
{ /* all leafs contains orig trgm */
res = trgm_contained_by(qtrg, key);
}
else if (ISALLTRUE(key))
{ /* non-leaf contains signature */
res = true;
}
else
{ /* non-leaf contains signature */
int32 k,
tmp = 0,
len = ARRNELEM(qtrg);
trgm *ptr = GETARR(qtrg);
BITVECP sign = GETSIGN(key);
res = true;
for (k = 0; k < len; k++)
{
CPTRGM(((char *) &tmp), ptr + k);
if (!GETBIT(sign, HASHVAL(tmp)))
{
res = false;
break;
}
}
}
break;
case RegExpICaseStrategyNumber:
#ifndef IGNORECASE
elog(ERROR, "cannot handle ~* with case-sensitive trigrams");
#endif
/* FALL THRU */
case RegExpStrategyNumber:
/* Regexp search is inexact */
*recheck = true;
/* Check regex match as much as we can with available info */
if (qtrg)
{
if (GIST_LEAF(entry))
{ /* all leafs contains orig trgm */
bool *check;
check = trgm_presence_map(qtrg, key);
res = trigramsMatchGraph(cache->graph, check);
pfree(check);
}
else if (ISALLTRUE(key))
{ /* non-leaf contains signature */
res = true;
}
else
{ /* non-leaf contains signature */
int32 k,
tmp = 0,
len = ARRNELEM(qtrg);
trgm *ptr = GETARR(qtrg);
BITVECP sign = GETSIGN(key);
bool *check;
/*
* GETBIT() tests may give false positives, due to limited
* size of the sign array. But since trigramsMatchGraph()
* implements a monotone boolean function, false positives
* in the check array can't lead to false negative answer.
* So we can apply trigramsMatchGraph despite uncertainty,
* and that usefully improves the quality of the search.
*/
check = (bool *) palloc(len * sizeof(bool));
for (k = 0; k < len; k++)
{
CPTRGM(((char *) &tmp), ptr + k);
check[k] = GETBIT(sign, HASHVAL(tmp));
}
res = trigramsMatchGraph(cache->graph, check);
pfree(check);
}
}
else
{
/* trigram-free query must be rechecked everywhere */
res = true;
}
break;
default:
elog(ERROR, "unrecognized strategy number: %d", strategy);
res = false; /* keep compiler quiet */
break;
}
PG_RETURN_BOOL(res);
}
| Datum gtrgm_decompress | ( | PG_FUNCTION_ARGS | ) |
Definition at line 157 of file trgm_gist.c.
References DatumGetPointer, DatumGetTextP, gistentryinit, GISTENTRY::key, GISTENTRY::leafkey, GISTENTRY::offset, GISTENTRY::page, palloc(), PG_GETARG_POINTER, PG_RETURN_POINTER, PointerGetDatum, and GISTENTRY::rel.
{
GISTENTRY *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
GISTENTRY *retval;
text *key;
key = DatumGetTextP(entry->key);
if (key != (text *) DatumGetPointer(entry->key))
{
/* need to pass back the decompressed item */
retval = palloc(sizeof(GISTENTRY));
gistentryinit(*retval, PointerGetDatum(key),
entry->rel, entry->page, entry->offset, entry->leafkey);
PG_RETURN_POINTER(retval);
}
else
{
/* we can return the entry as-is */
PG_RETURN_POINTER(entry);
}
}
| Datum gtrgm_distance | ( | PG_FUNCTION_ARGS | ) |
Definition at line 440 of file trgm_gist.c.
References ARRNELEM, cnt_sml(), cnt_sml_sign_common(), DatumGetPointer, DistanceStrategyNumber, elog, ERROR, generate_trgm(), GETSIGN, GIST_LEAF, ISALLTRUE, GISTENTRY::key, MAXALIGN, memcmp(), MemoryContextAlloc(), NULL, pfree(), PG_GETARG_POINTER, PG_GETARG_TEXT_P, PG_GETARG_UINT16, PG_RETURN_FLOAT8, VARDATA, VARHDRSZ, and VARSIZE.
{
GISTENTRY *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
text *query = PG_GETARG_TEXT_P(1);
StrategyNumber strategy = (StrategyNumber) PG_GETARG_UINT16(2);
/* Oid subtype = PG_GETARG_OID(3); */
TRGM *key = (TRGM *) DatumGetPointer(entry->key);
TRGM *qtrg;
float8 res;
Size querysize = VARSIZE(query);
char *cache = (char *) fcinfo->flinfo->fn_extra;
/*
* Cache the generated trigrams across multiple calls with the same query.
*/
if (cache == NULL ||
VARSIZE(cache) != querysize ||
memcmp(cache, query, querysize) != 0)
{
char *newcache;
qtrg = generate_trgm(VARDATA(query), querysize - VARHDRSZ);
newcache = MemoryContextAlloc(fcinfo->flinfo->fn_mcxt,
MAXALIGN(querysize) +
VARSIZE(qtrg));
memcpy(newcache, query, querysize);
memcpy(newcache + MAXALIGN(querysize), qtrg, VARSIZE(qtrg));
if (cache)
pfree(cache);
fcinfo->flinfo->fn_extra = newcache;
cache = newcache;
}
qtrg = (TRGM *) (cache + MAXALIGN(querysize));
switch (strategy)
{
case DistanceStrategyNumber:
if (GIST_LEAF(entry))
{ /* all leafs contains orig trgm */
res = 1.0 - cnt_sml(key, qtrg);
}
else if (ISALLTRUE(key))
{ /* all leafs contains orig trgm */
res = 0.0;
}
else
{ /* non-leaf contains signature */
int32 count = cnt_sml_sign_common(qtrg, GETSIGN(key));
int32 len = ARRNELEM(qtrg);
res = (len == 0) ? -1.0 : 1.0 - ((float8) count) / ((float8) len);
}
break;
default:
elog(ERROR, "unrecognized strategy number: %d", strategy);
res = 0; /* keep compiler quiet */
break;
}
PG_RETURN_FLOAT8(res);
}
| Datum gtrgm_in | ( | PG_FUNCTION_ARGS | ) |
Definition at line 82 of file trgm_gist.c.
References elog, ERROR, and PG_RETURN_DATUM.
{
elog(ERROR, "not implemented");
PG_RETURN_DATUM(0);
}
| Datum gtrgm_out | ( | PG_FUNCTION_ARGS | ) |
Definition at line 89 of file trgm_gist.c.
References elog, ERROR, and PG_RETURN_DATUM.
{
elog(ERROR, "not implemented");
PG_RETURN_DATUM(0);
}
| Datum gtrgm_penalty | ( | PG_FUNCTION_ARGS | ) |
Definition at line 671 of file trgm_gist.c.
References DatumGetPointer, GETSIGN, hemdist(), hemdistsign(), ISALLTRUE, ISARRKEY, GISTENTRY::key, makesign(), MAXALIGN, memcmp(), MemoryContextAlloc(), NULL, pfree(), PG_GETARG_POINTER, PG_RETURN_POINTER, SIGLENBIT, sign, sizebitvec(), and VARSIZE.
{
GISTENTRY *origentry = (GISTENTRY *) PG_GETARG_POINTER(0); /* always ISSIGNKEY */
GISTENTRY *newentry = (GISTENTRY *) PG_GETARG_POINTER(1);
float *penalty = (float *) PG_GETARG_POINTER(2);
TRGM *origval = (TRGM *) DatumGetPointer(origentry->key);
TRGM *newval = (TRGM *) DatumGetPointer(newentry->key);
BITVECP orig = GETSIGN(origval);
*penalty = 0.0;
if (ISARRKEY(newval))
{
char *cache = (char *) fcinfo->flinfo->fn_extra;
TRGM *cachedVal = (TRGM *) (cache + MAXALIGN(sizeof(BITVEC)));
Size newvalsize = VARSIZE(newval);
BITVECP sign;
/*
* Cache the sign data across multiple calls with the same newval.
*/
if (cache == NULL ||
VARSIZE(cachedVal) != newvalsize ||
memcmp(cachedVal, newval, newvalsize) != 0)
{
char *newcache;
newcache = MemoryContextAlloc(fcinfo->flinfo->fn_mcxt,
MAXALIGN(sizeof(BITVEC)) +
newvalsize);
makesign((BITVECP) newcache, newval);
cachedVal = (TRGM *) (newcache + MAXALIGN(sizeof(BITVEC)));
memcpy(cachedVal, newval, newvalsize);
if (cache)
pfree(cache);
fcinfo->flinfo->fn_extra = newcache;
cache = newcache;
}
sign = (BITVECP) cache;
if (ISALLTRUE(origval))
*penalty = ((float) (SIGLENBIT - sizebitvec(sign))) / (float) (SIGLENBIT + 1);
else
*penalty = hemdistsign(sign, orig);
}
else
*penalty = hemdist(origval, newval);
PG_RETURN_POINTER(penalty);
}
| Datum gtrgm_picksplit | ( | PG_FUNCTION_ARGS | ) |
Definition at line 777 of file trgm_gist.c.
References ALLISTRUE, CALCGTSIZE, comparecost(), SPLITCOST::cost, fillcache(), FirstOffsetNumber, TRGM::flag, GETENTRY, GETSIGN, hemdistcache(), hemdistsign(), i, ISALLTRUE, LOOPBYTE, MemSet, GistEntryVector::n, OffsetNumberNext, palloc(), PG_GETARG_POINTER, PG_RETURN_POINTER, PointerGetDatum, SPLITCOST::pos, qsort, SET_VARSIZE, SIGLENBIT, CACHESIGN::sign, sign, SIGNKEY, sizebitvec(), GIST_SPLITVEC::spl_ldatum, GIST_SPLITVEC::spl_left, GIST_SPLITVEC::spl_nleft, GIST_SPLITVEC::spl_nright, GIST_SPLITVEC::spl_rdatum, GIST_SPLITVEC::spl_right, and WISH_F.
{
GistEntryVector *entryvec = (GistEntryVector *) PG_GETARG_POINTER(0);
OffsetNumber maxoff = entryvec->n - 2;
GIST_SPLITVEC *v = (GIST_SPLITVEC *) PG_GETARG_POINTER(1);
OffsetNumber k,
j;
TRGM *datum_l,
*datum_r;
BITVECP union_l,
union_r;
int32 size_alpha,
size_beta;
int32 size_waste,
waste = -1;
int32 nbytes;
OffsetNumber seed_1 = 0,
seed_2 = 0;
OffsetNumber *left,
*right;
BITVECP ptr;
int i;
CACHESIGN *cache;
SPLITCOST *costvector;
/* cache the sign data for each existing item */
cache = (CACHESIGN *) palloc(sizeof(CACHESIGN) * (maxoff + 2));
for (k = FirstOffsetNumber; k <= maxoff; k = OffsetNumberNext(k))
fillcache(&cache[k], GETENTRY(entryvec, k));
/* now find the two furthest-apart items */
for (k = FirstOffsetNumber; k < maxoff; k = OffsetNumberNext(k))
{
for (j = OffsetNumberNext(k); j <= maxoff; j = OffsetNumberNext(j))
{
size_waste = hemdistcache(&(cache[j]), &(cache[k]));
if (size_waste > waste)
{
waste = size_waste;
seed_1 = k;
seed_2 = j;
}
}
}
/* just in case we didn't make a selection ... */
if (seed_1 == 0 || seed_2 == 0)
{
seed_1 = 1;
seed_2 = 2;
}
/* initialize the result vectors */
nbytes = (maxoff + 2) * sizeof(OffsetNumber);
v->spl_left = left = (OffsetNumber *) palloc(nbytes);
v->spl_right = right = (OffsetNumber *) palloc(nbytes);
v->spl_nleft = 0;
v->spl_nright = 0;
/* form initial .. */
if (cache[seed_1].allistrue)
{
datum_l = (TRGM *) palloc(CALCGTSIZE(SIGNKEY | ALLISTRUE, 0));
SET_VARSIZE(datum_l, CALCGTSIZE(SIGNKEY | ALLISTRUE, 0));
datum_l->flag = SIGNKEY | ALLISTRUE;
}
else
{
datum_l = (TRGM *) palloc(CALCGTSIZE(SIGNKEY, 0));
SET_VARSIZE(datum_l, CALCGTSIZE(SIGNKEY, 0));
datum_l->flag = SIGNKEY;
memcpy((void *) GETSIGN(datum_l), (void *) cache[seed_1].sign, sizeof(BITVEC));
}
if (cache[seed_2].allistrue)
{
datum_r = (TRGM *) palloc(CALCGTSIZE(SIGNKEY | ALLISTRUE, 0));
SET_VARSIZE(datum_r, CALCGTSIZE(SIGNKEY | ALLISTRUE, 0));
datum_r->flag = SIGNKEY | ALLISTRUE;
}
else
{
datum_r = (TRGM *) palloc(CALCGTSIZE(SIGNKEY, 0));
SET_VARSIZE(datum_r, CALCGTSIZE(SIGNKEY, 0));
datum_r->flag = SIGNKEY;
memcpy((void *) GETSIGN(datum_r), (void *) cache[seed_2].sign, sizeof(BITVEC));
}
union_l = GETSIGN(datum_l);
union_r = GETSIGN(datum_r);
maxoff = OffsetNumberNext(maxoff);
fillcache(&cache[maxoff], GETENTRY(entryvec, maxoff));
/* sort before ... */
costvector = (SPLITCOST *) palloc(sizeof(SPLITCOST) * maxoff);
for (j = FirstOffsetNumber; j <= maxoff; j = OffsetNumberNext(j))
{
costvector[j - 1].pos = j;
size_alpha = hemdistcache(&(cache[seed_1]), &(cache[j]));
size_beta = hemdistcache(&(cache[seed_2]), &(cache[j]));
costvector[j - 1].cost = abs(size_alpha - size_beta);
}
qsort((void *) costvector, maxoff, sizeof(SPLITCOST), comparecost);
for (k = 0; k < maxoff; k++)
{
j = costvector[k].pos;
if (j == seed_1)
{
*left++ = j;
v->spl_nleft++;
continue;
}
else if (j == seed_2)
{
*right++ = j;
v->spl_nright++;
continue;
}
if (ISALLTRUE(datum_l) || cache[j].allistrue)
{
if (ISALLTRUE(datum_l) && cache[j].allistrue)
size_alpha = 0;
else
size_alpha = SIGLENBIT - sizebitvec(
(cache[j].allistrue) ? GETSIGN(datum_l) : GETSIGN(cache[j].sign)
);
}
else
size_alpha = hemdistsign(cache[j].sign, GETSIGN(datum_l));
if (ISALLTRUE(datum_r) || cache[j].allistrue)
{
if (ISALLTRUE(datum_r) && cache[j].allistrue)
size_beta = 0;
else
size_beta = SIGLENBIT - sizebitvec(
(cache[j].allistrue) ? GETSIGN(datum_r) : GETSIGN(cache[j].sign)
);
}
else
size_beta = hemdistsign(cache[j].sign, GETSIGN(datum_r));
if (size_alpha < size_beta + WISH_F(v->spl_nleft, v->spl_nright, 0.1))
{
if (ISALLTRUE(datum_l) || cache[j].allistrue)
{
if (!ISALLTRUE(datum_l))
MemSet((void *) GETSIGN(datum_l), 0xff, sizeof(BITVEC));
}
else
{
ptr = cache[j].sign;
LOOPBYTE
union_l[i] |= ptr[i];
}
*left++ = j;
v->spl_nleft++;
}
else
{
if (ISALLTRUE(datum_r) || cache[j].allistrue)
{
if (!ISALLTRUE(datum_r))
MemSet((void *) GETSIGN(datum_r), 0xff, sizeof(BITVEC));
}
else
{
ptr = cache[j].sign;
LOOPBYTE
union_r[i] |= ptr[i];
}
*right++ = j;
v->spl_nright++;
}
}
*right = *left = FirstOffsetNumber;
v->spl_ldatum = PointerGetDatum(datum_l);
v->spl_rdatum = PointerGetDatum(datum_r);
PG_RETURN_POINTER(v);
}
| Datum gtrgm_same | ( | PG_FUNCTION_ARGS | ) |
Definition at line 571 of file trgm_gist.c.
References ARRNELEM, CMPTRGM, GETARR, GETSIGN, i, ISALLTRUE, ISSIGNKEY, LOOPBYTE, PG_GETARG_POINTER, and PG_RETURN_POINTER.
{
TRGM *a = (TRGM *) PG_GETARG_POINTER(0);
TRGM *b = (TRGM *) PG_GETARG_POINTER(1);
bool *result = (bool *) PG_GETARG_POINTER(2);
if (ISSIGNKEY(a))
{ /* then b also ISSIGNKEY */
if (ISALLTRUE(a) && ISALLTRUE(b))
*result = true;
else if (ISALLTRUE(a))
*result = false;
else if (ISALLTRUE(b))
*result = false;
else
{
int32 i;
BITVECP sa = GETSIGN(a),
sb = GETSIGN(b);
*result = true;
LOOPBYTE
{
if (sa[i] != sb[i])
{
*result = false;
break;
}
}
}
}
else
{ /* a and b ISARRKEY */
int32 lena = ARRNELEM(a),
lenb = ARRNELEM(b);
if (lena != lenb)
*result = false;
else
{
trgm *ptra = GETARR(a),
*ptrb = GETARR(b);
int32 i;
*result = true;
for (i = 0; i < lena; i++)
if (CMPTRGM(ptra + i, ptrb + i))
{
*result = false;
break;
}
}
}
PG_RETURN_POINTER(result);
}
| Datum gtrgm_union | ( | PG_FUNCTION_ARGS | ) |
Definition at line 538 of file trgm_gist.c.
References CALCGTSIZE, TRGM::flag, flag(), GETENTRY, GETSIGN, i, ISALLTRUE, MemSet, GistEntryVector::n, palloc(), PG_GETARG_POINTER, PG_RETURN_POINTER, SET_VARSIZE, and unionkey().
{
GistEntryVector *entryvec = (GistEntryVector *) PG_GETARG_POINTER(0);
int32 len = entryvec->n;
int *size = (int *) PG_GETARG_POINTER(1);
BITVEC base;
int32 i;
int32 flag = 0;
TRGM *result;
MemSet((void *) base, 0, sizeof(BITVEC));
for (i = 0; i < len; i++)
{
if (unionkey(base, GETENTRY(entryvec, i)))
{
flag = ALLISTRUE;
break;
}
}
flag |= SIGNKEY;
len = CALCGTSIZE(flag, 0);
result = (TRGM *) palloc(len);
SET_VARSIZE(result, len);
result->flag = flag;
if (!ISALLTRUE(result))
memcpy((void *) GETSIGN(result), (void *) base, sizeof(BITVEC));
*size = len;
PG_RETURN_POINTER(result);
}
Definition at line 655 of file trgm_gist.c.
References GETSIGN, hemdistsign(), ISALLTRUE, SIGLENBIT, and sizebitvec().
Referenced by gtrgm_penalty().
{
if (ISALLTRUE(a))
{
if (ISALLTRUE(b))
return 0;
else
return SIGLENBIT - sizebitvec(GETSIGN(b));
}
else if (ISALLTRUE(b))
return SIGLENBIT - sizebitvec(GETSIGN(a));
return hemdistsign(GETSIGN(a), GETSIGN(b));
}
Definition at line 761 of file trgm_gist.c.
References CACHESIGN::allistrue, hemdistsign(), SIGLENBIT, CACHESIGN::sign, and sizebitvec().
Referenced by gtrgm_picksplit().
{
if (a->allistrue)
{
if (b->allistrue)
return 0;
else
return SIGLENBIT - sizebitvec(b->sign);
}
else if (b->allistrue)
return SIGLENBIT - sizebitvec(a->sign);
return hemdistsign(a->sign, b->sign);
}
Definition at line 640 of file trgm_gist.c.
References i, LOOPBYTE, and number_of_ones.
Referenced by gtrgm_penalty(), gtrgm_picksplit(), hemdist(), and hemdistcache().
{
int i,
diff,
dist = 0;
LOOPBYTE
{
diff = (unsigned char) (a[i] ^ b[i]);
dist += number_of_ones[diff];
}
return dist;
}
Definition at line 96 of file trgm_gist.c.
References ARRNELEM, CPTRGM, GETARR, HASH, MemSet, SETBIT, and SIGLENBIT.
Referenced by fillcache(), and gtrgm_penalty().
| PG_FUNCTION_INFO_V1 | ( | gtrgm_compress | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_consistent | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_decompress | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_union | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_distance | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_out | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_penalty | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_in | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_picksplit | ) |
| PG_FUNCTION_INFO_V1 | ( | gtrgm_same | ) |
Definition at line 629 of file trgm_gist.c.
References i, LOOPBYTE, and number_of_ones.
Referenced by gtrgm_penalty(), gtrgm_picksplit(), hemdist(), and hemdistcache().
const uint8 number_of_ones[256] [static] |
{
0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
4, 5, 5, 6, 5, 6, 6, 7, 5, 6, 6, 7, 6, 7, 7, 8
}
Definition at line 61 of file trgm_gist.c.
Referenced by hemdistsign(), and sizebitvec().
1.7.1