PostgreSQL Source Code  git master
selfuncs.c
Go to the documentation of this file.
1 /*-------------------------------------------------------------------------
2  *
3  * selfuncs.c
4  * Selectivity functions and index cost estimation functions for
5  * standard operators and index access methods.
6  *
7  * Selectivity routines are registered in the pg_operator catalog
8  * in the "oprrest" and "oprjoin" attributes.
9  *
10  * Index cost functions are located via the index AM's API struct,
11  * which is obtained from the handler function registered in pg_am.
12  *
13  * Portions Copyright (c) 1996-2021, PostgreSQL Global Development Group
14  * Portions Copyright (c) 1994, Regents of the University of California
15  *
16  *
17  * IDENTIFICATION
18  * src/backend/utils/adt/selfuncs.c
19  *
20  *-------------------------------------------------------------------------
21  */
22 
23 /*----------
24  * Operator selectivity estimation functions are called to estimate the
25  * selectivity of WHERE clauses whose top-level operator is their operator.
26  * We divide the problem into two cases:
27  * Restriction clause estimation: the clause involves vars of just
28  * one relation.
29  * Join clause estimation: the clause involves vars of multiple rels.
30  * Join selectivity estimation is far more difficult and usually less accurate
31  * than restriction estimation.
32  *
33  * When dealing with the inner scan of a nestloop join, we consider the
34  * join's joinclauses as restriction clauses for the inner relation, and
35  * treat vars of the outer relation as parameters (a/k/a constants of unknown
36  * values). So, restriction estimators need to be able to accept an argument
37  * telling which relation is to be treated as the variable.
38  *
39  * The call convention for a restriction estimator (oprrest function) is
40  *
41  * Selectivity oprrest (PlannerInfo *root,
42  * Oid operator,
43  * List *args,
44  * int varRelid);
45  *
46  * root: general information about the query (rtable and RelOptInfo lists
47  * are particularly important for the estimator).
48  * operator: OID of the specific operator in question.
49  * args: argument list from the operator clause.
50  * varRelid: if not zero, the relid (rtable index) of the relation to
51  * be treated as the variable relation. May be zero if the args list
52  * is known to contain vars of only one relation.
53  *
54  * This is represented at the SQL level (in pg_proc) as
55  *
56  * float8 oprrest (internal, oid, internal, int4);
57  *
58  * The result is a selectivity, that is, a fraction (0 to 1) of the rows
59  * of the relation that are expected to produce a TRUE result for the
60  * given operator.
61  *
62  * The call convention for a join estimator (oprjoin function) is similar
63  * except that varRelid is not needed, and instead join information is
64  * supplied:
65  *
66  * Selectivity oprjoin (PlannerInfo *root,
67  * Oid operator,
68  * List *args,
69  * JoinType jointype,
70  * SpecialJoinInfo *sjinfo);
71  *
72  * float8 oprjoin (internal, oid, internal, int2, internal);
73  *
74  * (Before Postgres 8.4, join estimators had only the first four of these
75  * parameters. That signature is still allowed, but deprecated.) The
76  * relationship between jointype and sjinfo is explained in the comments for
77  * clause_selectivity() --- the short version is that jointype is usually
78  * best ignored in favor of examining sjinfo.
79  *
80  * Join selectivity for regular inner and outer joins is defined as the
81  * fraction (0 to 1) of the cross product of the relations that is expected
82  * to produce a TRUE result for the given operator. For both semi and anti
83  * joins, however, the selectivity is defined as the fraction of the left-hand
84  * side relation's rows that are expected to have a match (ie, at least one
85  * row with a TRUE result) in the right-hand side.
86  *
87  * For both oprrest and oprjoin functions, the operator's input collation OID
88  * (if any) is passed using the standard fmgr mechanism, so that the estimator
89  * function can fetch it with PG_GET_COLLATION(). Note, however, that all
90  * statistics in pg_statistic are currently built using the relevant column's
91  * collation.
92  *----------
93  */
94 
95 #include "postgres.h"
96 
97 #include <ctype.h>
98 #include <math.h>
99 
100 #include "access/brin.h"
101 #include "access/brin_page.h"
102 #include "access/gin.h"
103 #include "access/table.h"
104 #include "access/tableam.h"
105 #include "access/visibilitymap.h"
106 #include "catalog/pg_am.h"
107 #include "catalog/pg_collation.h"
108 #include "catalog/pg_operator.h"
109 #include "catalog/pg_statistic.h"
110 #include "catalog/pg_statistic_ext.h"
111 #include "executor/nodeAgg.h"
112 #include "miscadmin.h"
113 #include "nodes/makefuncs.h"
114 #include "nodes/nodeFuncs.h"
115 #include "optimizer/clauses.h"
116 #include "optimizer/cost.h"
117 #include "optimizer/optimizer.h"
118 #include "optimizer/pathnode.h"
119 #include "optimizer/paths.h"
120 #include "optimizer/plancat.h"
121 #include "parser/parse_clause.h"
122 #include "parser/parsetree.h"
123 #include "statistics/statistics.h"
124 #include "storage/bufmgr.h"
125 #include "utils/acl.h"
126 #include "utils/builtins.h"
127 #include "utils/date.h"
128 #include "utils/datum.h"
129 #include "utils/fmgroids.h"
130 #include "utils/index_selfuncs.h"
131 #include "utils/lsyscache.h"
132 #include "utils/memutils.h"
133 #include "utils/pg_locale.h"
134 #include "utils/rel.h"
135 #include "utils/selfuncs.h"
136 #include "utils/snapmgr.h"
137 #include "utils/spccache.h"
138 #include "utils/syscache.h"
139 #include "utils/timestamp.h"
140 #include "utils/typcache.h"
141 
142 
143 /* Hooks for plugins to get control when we ask for stats */
144 get_relation_stats_hook_type get_relation_stats_hook = NULL;
145 get_index_stats_hook_type get_index_stats_hook = NULL;
146 
147 static double eqsel_internal(PG_FUNCTION_ARGS, bool negate);
148 static double eqjoinsel_inner(Oid opfuncoid, Oid collation,
149  VariableStatData *vardata1, VariableStatData *vardata2,
150  double nd1, double nd2,
151  bool isdefault1, bool isdefault2,
152  AttStatsSlot *sslot1, AttStatsSlot *sslot2,
153  Form_pg_statistic stats1, Form_pg_statistic stats2,
154  bool have_mcvs1, bool have_mcvs2);
155 static double eqjoinsel_semi(Oid opfuncoid, Oid collation,
156  VariableStatData *vardata1, VariableStatData *vardata2,
157  double nd1, double nd2,
158  bool isdefault1, bool isdefault2,
159  AttStatsSlot *sslot1, AttStatsSlot *sslot2,
160  Form_pg_statistic stats1, Form_pg_statistic stats2,
161  bool have_mcvs1, bool have_mcvs2,
162  RelOptInfo *inner_rel);
163 static bool estimate_multivariate_ndistinct(PlannerInfo *root,
164  RelOptInfo *rel, List **varinfos, double *ndistinct);
165 static bool convert_to_scalar(Datum value, Oid valuetypid, Oid collid,
166  double *scaledvalue,
167  Datum lobound, Datum hibound, Oid boundstypid,
168  double *scaledlobound, double *scaledhibound);
169 static double convert_numeric_to_scalar(Datum value, Oid typid, bool *failure);
170 static void convert_string_to_scalar(char *value,
171  double *scaledvalue,
172  char *lobound,
173  double *scaledlobound,
174  char *hibound,
175  double *scaledhibound);
176 static void convert_bytea_to_scalar(Datum value,
177  double *scaledvalue,
178  Datum lobound,
179  double *scaledlobound,
180  Datum hibound,
181  double *scaledhibound);
182 static double convert_one_string_to_scalar(char *value,
183  int rangelo, int rangehi);
184 static double convert_one_bytea_to_scalar(unsigned char *value, int valuelen,
185  int rangelo, int rangehi);
186 static char *convert_string_datum(Datum value, Oid typid, Oid collid,
187  bool *failure);
188 static double convert_timevalue_to_scalar(Datum value, Oid typid,
189  bool *failure);
190 static void examine_simple_variable(PlannerInfo *root, Var *var,
191  VariableStatData *vardata);
192 static bool get_variable_range(PlannerInfo *root, VariableStatData *vardata,
193  Oid sortop, Oid collation,
194  Datum *min, Datum *max);
195 static void get_stats_slot_range(AttStatsSlot *sslot,
196  Oid opfuncoid, FmgrInfo *opproc,
197  Oid collation, int16 typLen, bool typByVal,
198  Datum *min, Datum *max, bool *p_have_data);
199 static bool get_actual_variable_range(PlannerInfo *root,
200  VariableStatData *vardata,
201  Oid sortop, Oid collation,
202  Datum *min, Datum *max);
203 static bool get_actual_variable_endpoint(Relation heapRel,
204  Relation indexRel,
205  ScanDirection indexscandir,
206  ScanKey scankeys,
207  int16 typLen,
208  bool typByVal,
209  TupleTableSlot *tableslot,
210  MemoryContext outercontext,
211  Datum *endpointDatum);
212 static RelOptInfo *find_join_input_rel(PlannerInfo *root, Relids relids);
213 
214 
215 /*
216  * eqsel - Selectivity of "=" for any data types.
217  *
218  * Note: this routine is also used to estimate selectivity for some
219  * operators that are not "=" but have comparable selectivity behavior,
220  * such as "~=" (geometric approximate-match). Even for "=", we must
221  * keep in mind that the left and right datatypes may differ.
222  */
223 Datum
224 eqsel(PG_FUNCTION_ARGS)
225 {
226  PG_RETURN_FLOAT8((float8) eqsel_internal(fcinfo, false));
227 }
228 
229 /*
230  * Common code for eqsel() and neqsel()
231  */
232 static double
233 eqsel_internal(PG_FUNCTION_ARGS, bool negate)
234 {
235  PlannerInfo *root = (PlannerInfo *) PG_GETARG_POINTER(0);
236  Oid operator = PG_GETARG_OID(1);
237  List *args = (List *) PG_GETARG_POINTER(2);
238  int varRelid = PG_GETARG_INT32(3);
239  Oid collation = PG_GET_COLLATION();
240  VariableStatData vardata;
241  Node *other;
242  bool varonleft;
243  double selec;
244 
245  /*
246  * When asked about <>, we do the estimation using the corresponding =
247  * operator, then convert to <> via "1.0 - eq_selectivity - nullfrac".
248  */
249  if (negate)
250  {
251  operator = get_negator(operator);
252  if (!OidIsValid(operator))
253  {
254  /* Use default selectivity (should we raise an error instead?) */
255  return 1.0 - DEFAULT_EQ_SEL;
256  }
257  }
258 
259  /*
260  * If expression is not variable = something or something = variable, then
261  * punt and return a default estimate.
262  */
263  if (!get_restriction_variable(root, args, varRelid,
264  &vardata, &other, &varonleft))
265  return negate ? (1.0 - DEFAULT_EQ_SEL) : DEFAULT_EQ_SEL;
266 
267  /*
268  * We can do a lot better if the something is a constant. (Note: the
269  * Const might result from estimation rather than being a simple constant
270  * in the query.)
271  */
272  if (IsA(other, Const))
273  selec = var_eq_const(&vardata, operator, collation,
274  ((Const *) other)->constvalue,
275  ((Const *) other)->constisnull,
276  varonleft, negate);
277  else
278  selec = var_eq_non_const(&vardata, operator, collation, other,
279  varonleft, negate);
280 
281  ReleaseVariableStats(vardata);
282 
283  return selec;
284 }
285 
286 /*
287  * var_eq_const --- eqsel for var = const case
288  *
289  * This is exported so that some other estimation functions can use it.
290  */
291 double
292 var_eq_const(VariableStatData *vardata, Oid operator, Oid collation,
293  Datum constval, bool constisnull,
294  bool varonleft, bool negate)
295 {
296  double selec;
297  double nullfrac = 0.0;
298  bool isdefault;
299  Oid opfuncoid;
300 
301  /*
302  * If the constant is NULL, assume operator is strict and return zero, ie,
303  * operator will never return TRUE. (It's zero even for a negator op.)
304  */
305  if (constisnull)
306  return 0.0;
307 
308  /*
309  * Grab the nullfrac for use below. Note we allow use of nullfrac
310  * regardless of security check.
311  */
312  if (HeapTupleIsValid(vardata->statsTuple))
313  {
314  Form_pg_statistic stats;
315 
316  stats = (Form_pg_statistic) GETSTRUCT(vardata->statsTuple);
317  nullfrac = stats->stanullfrac;
318  }
319 
320  /*
321  * If we matched the var to a unique index or DISTINCT clause, assume
322  * there is exactly one match regardless of anything else. (This is
323  * slightly bogus, since the index or clause's equality operator might be
324  * different from ours, but it's much more likely to be right than
325  * ignoring the information.)
326  */
327  if (vardata->isunique && vardata->rel && vardata->rel->tuples >= 1.0)
328  {
329  selec = 1.0 / vardata->rel->tuples;
330  }
331  else if (HeapTupleIsValid(vardata->statsTuple) &&
332  statistic_proc_security_check(vardata,
333  (opfuncoid = get_opcode(operator))))
334  {
335  AttStatsSlot sslot;
336  bool match = false;
337  int i;
338 
339  /*
340  * Is the constant "=" to any of the column's most common values?
341  * (Although the given operator may not really be "=", we will assume
342  * that seeing whether it returns TRUE is an appropriate test. If you
343  * don't like this, maybe you shouldn't be using eqsel for your
344  * operator...)
345  */
346  if (get_attstatsslot(&sslot, vardata->statsTuple,
347  STATISTIC_KIND_MCV, InvalidOid,
348  ATTSTATSSLOT_VALUES | ATTSTATSSLOT_NUMBERS))
349  {
350  LOCAL_FCINFO(fcinfo, 2);
351  FmgrInfo eqproc;
352 
353  fmgr_info(opfuncoid, &eqproc);
354 
355  /*
356  * Save a few cycles by setting up the fcinfo struct just once.
357  * Using FunctionCallInvoke directly also avoids failure if the
358  * eqproc returns NULL, though really equality functions should
359  * never do that.
360  */
361  InitFunctionCallInfoData(*fcinfo, &eqproc, 2, collation,
362  NULL, NULL);
363  fcinfo->args[0].isnull = false;
364  fcinfo->args[1].isnull = false;
365  /* be careful to apply operator right way 'round */
366  if (varonleft)
367  fcinfo->args[1].value = constval;
368  else
369  fcinfo->args[0].value = constval;
370 
371  for (i = 0; i < sslot.nvalues; i++)
372  {
373  Datum fresult;
374 
375  if (varonleft)
376  fcinfo->args[0].value = sslot.values[i];
377  else
378  fcinfo->args[1].value = sslot.values[i];
379  fcinfo->isnull = false;
380  fresult = FunctionCallInvoke(fcinfo);
381  if (!fcinfo->isnull && DatumGetBool(fresult))
382  {
383  match = true;
384  break;
385  }
386  }
387  }
388  else
389  {
390  /* no most-common-value info available */
391  i = 0; /* keep compiler quiet */
392  }
393 
394  if (match)
395  {
396  /*
397  * Constant is "=" to this common value. We know selectivity
398  * exactly (or as exactly as ANALYZE could calculate it, anyway).
399  */
400  selec = sslot.numbers[i];
401  }
402  else
403  {
404  /*
405  * Comparison is against a constant that is neither NULL nor any
406  * of the common values. Its selectivity cannot be more than
407  * this:
408  */
409  double sumcommon = 0.0;
410  double otherdistinct;
411 
412  for (i = 0; i < sslot.nnumbers; i++)
413  sumcommon += sslot.numbers[i];
414  selec = 1.0 - sumcommon - nullfrac;
415  CLAMP_PROBABILITY(selec);
416 
417  /*
418  * and in fact it's probably a good deal less. We approximate that
419  * all the not-common values share this remaining fraction
420  * equally, so we divide by the number of other distinct values.
421  */
422  otherdistinct = get_variable_numdistinct(vardata, &isdefault) -
423  sslot.nnumbers;
424  if (otherdistinct > 1)
425  selec /= otherdistinct;
426 
427  /*
428  * Another cross-check: selectivity shouldn't be estimated as more
429  * than the least common "most common value".
430  */
431  if (sslot.nnumbers > 0 && selec > sslot.numbers[sslot.nnumbers - 1])
432  selec = sslot.numbers[sslot.nnumbers - 1];
433  }
434 
435  free_attstatsslot(&sslot);
436  }
437  else
438  {
439  /*
440  * No ANALYZE stats available, so make a guess using estimated number
441  * of distinct values and assuming they are equally common. (The guess
442  * is unlikely to be very good, but we do know a few special cases.)
443  */
444  selec = 1.0 / get_variable_numdistinct(vardata, &isdefault);
445  }
446 
447  /* now adjust if we wanted <> rather than = */
448  if (negate)
449  selec = 1.0 - selec - nullfrac;
450 
451  /* result should be in range, but make sure... */
452  CLAMP_PROBABILITY(selec);
453 
454  return selec;
455 }
456 
457 /*
458  * var_eq_non_const --- eqsel for var = something-other-than-const case
459  *
460  * This is exported so that some other estimation functions can use it.
461  */
462 double
463 var_eq_non_const(VariableStatData *vardata, Oid operator, Oid collation,
464  Node *other,
465  bool varonleft, bool negate)
466 {
467  double selec;
468  double nullfrac = 0.0;
469  bool isdefault;
470 
471  /*
472  * Grab the nullfrac for use below.
473  */
474  if (HeapTupleIsValid(vardata->statsTuple))
475  {
476  Form_pg_statistic stats;
477 
478  stats = (Form_pg_statistic) GETSTRUCT(vardata->statsTuple);
479  nullfrac = stats->stanullfrac;
480  }
481 
482  /*
483  * If we matched the var to a unique index or DISTINCT clause, assume
484  * there is exactly one match regardless of anything else. (This is
485  * slightly bogus, since the index or clause's equality operator might be
486  * different from ours, but it's much more likely to be right than
487  * ignoring the information.)
488  */
489  if (vardata->isunique && vardata->rel && vardata->rel->tuples >= 1.0)
490  {
491  selec = 1.0 / vardata->rel->tuples;
492  }
493  else if (HeapTupleIsValid(vardata->statsTuple))
494  {
495  double ndistinct;
496  AttStatsSlot sslot;
497 
498  /*
499  * Search is for a value that we do not know a priori, but we will
500  * assume it is not NULL. Estimate the selectivity as non-null
501  * fraction divided by number of distinct values, so that we get a
502  * result averaged over all possible values whether common or
503  * uncommon. (Essentially, we are assuming that the not-yet-known
504  * comparison value is equally likely to be any of the possible
505  * values, regardless of their frequency in the table. Is that a good
506  * idea?)
507  */
508  selec = 1.0 - nullfrac;
509  ndistinct = get_variable_numdistinct(vardata, &isdefault);
510  if (ndistinct > 1)
511  selec /= ndistinct;
512 
513  /*
514  * Cross-check: selectivity should never be estimated as more than the
515  * most common value's.
516  */
517  if (get_attstatsslot(&sslot, vardata->statsTuple,
518  STATISTIC_KIND_MCV, InvalidOid,
519  ATTSTATSSLOT_NUMBERS))
520  {
521  if (sslot.nnumbers > 0 && selec > sslot.numbers[0])
522  selec = sslot.numbers[0];
523  free_attstatsslot(&sslot);
524  }
525  }
526  else
527  {
528  /*
529  * No ANALYZE stats available, so make a guess using estimated number
530  * of distinct values and assuming they are equally common. (The guess
531  * is unlikely to be very good, but we do know a few special cases.)
532  */
533  selec = 1.0 / get_variable_numdistinct(vardata, &isdefault);
534  }
535 
536  /* now adjust if we wanted <> rather than = */
537  if (negate)
538  selec = 1.0 - selec - nullfrac;
539 
540  /* result should be in range, but make sure... */
541  CLAMP_PROBABILITY(selec);
542 
543  return selec;
544 }
545 
546 /*
547  * neqsel - Selectivity of "!=" for any data types.
548  *
549  * This routine is also used for some operators that are not "!="
550  * but have comparable selectivity behavior. See above comments
551  * for eqsel().
552  */
553 Datum
554 neqsel(PG_FUNCTION_ARGS)
555 {
556  PG_RETURN_FLOAT8((float8) eqsel_internal(fcinfo, true));
557 }
558 
559 /*
560  * scalarineqsel - Selectivity of "<", "<=", ">", ">=" for scalars.
561  *
562  * This is the guts of scalarltsel/scalarlesel/scalargtsel/scalargesel.
563  * The isgt and iseq flags distinguish which of the four cases apply.
564  *
565  * The caller has commuted the clause, if necessary, so that we can treat
566  * the variable as being on the left. The caller must also make sure that
567  * the other side of the clause is a non-null Const, and dissect that into
568  * a value and datatype. (This definition simplifies some callers that
569  * want to estimate against a computed value instead of a Const node.)
570  *
571  * This routine works for any datatype (or pair of datatypes) known to
572  * convert_to_scalar(). If it is applied to some other datatype,
573  * it will return an approximate estimate based on assuming that the constant
574  * value falls in the middle of the bin identified by binary search.
575  */
576 static double
577 scalarineqsel(PlannerInfo *root, Oid operator, bool isgt, bool iseq,
578  Oid collation,
579  VariableStatData *vardata, Datum constval, Oid consttype)
580 {
581  Form_pg_statistic stats;
582  FmgrInfo opproc;
583  double mcv_selec,
584  hist_selec,
585  sumcommon;
586  double selec;
587 
588  if (!HeapTupleIsValid(vardata->statsTuple))
589  {
590  /*
591  * No stats are available. Typically this means we have to fall back
592  * on the default estimate; but if the variable is CTID then we can
593  * make an estimate based on comparing the constant to the table size.
594  */
595  if (vardata->var && IsA(vardata->var, Var) &&
596  ((Var *) vardata->var)->varattno == SelfItemPointerAttributeNumber)
597  {
598  ItemPointer itemptr;
599  double block;
600  double density;
601 
602  /*
603  * If the relation's empty, we're going to include all of it.
604  * (This is mostly to avoid divide-by-zero below.)
605  */
606  if (vardata->rel->pages == 0)
607  return 1.0;
608 
609  itemptr = (ItemPointer) DatumGetPointer(constval);
610  block = ItemPointerGetBlockNumberNoCheck(itemptr);
611 
612  /*
613  * Determine the average number of tuples per page (density).
614  *
615  * Since the last page will, on average, be only half full, we can
616  * estimate it to have half as many tuples as earlier pages. So
617  * give it half the weight of a regular page.
618  */
619  density = vardata->rel->tuples / (vardata->rel->pages - 0.5);
620 
621  /* If target is the last page, use half the density. */
622  if (block >= vardata->rel->pages - 1)
623  density *= 0.5;
624 
625  /*
626  * Using the average tuples per page, calculate how far into the
627  * page the itemptr is likely to be and adjust block accordingly,
628  * by adding that fraction of a whole block (but never more than a
629  * whole block, no matter how high the itemptr's offset is). Here
630  * we are ignoring the possibility of dead-tuple line pointers,
631  * which is fairly bogus, but we lack the info to do better.
632  */
633  if (density > 0.0)
634  {
635  OffsetNumber offset = ItemPointerGetOffsetNumberNoCheck(itemptr);
636 
637  block += Min(offset / density, 1.0);
638  }
639 
640  /*
641  * Convert relative block number to selectivity. Again, the last
642  * page has only half weight.
643  */
644  selec = block / (vardata->rel->pages - 0.5);
645 
646  /*
647  * The calculation so far gave us a selectivity for the "<=" case.
648  * We'll have one fewer tuple for "<" and one additional tuple for
649  * ">=", the latter of which we'll reverse the selectivity for
650  * below, so we can simply subtract one tuple for both cases. The
651  * cases that need this adjustment can be identified by iseq being
652  * equal to isgt.
653  */
654  if (iseq == isgt && vardata->rel->tuples >= 1.0)
655  selec -= (1.0 / vardata->rel->tuples);
656 
657  /* Finally, reverse the selectivity for the ">", ">=" cases. */
658  if (isgt)
659  selec = 1.0 - selec;
660 
661  CLAMP_PROBABILITY(selec);
662  return selec;
663  }
664 
665  /* no stats available, so default result */
666  return DEFAULT_INEQ_SEL;
667  }
668  stats = (Form_pg_statistic) GETSTRUCT(vardata->statsTuple);
669 
670  fmgr_info(get_opcode(operator), &opproc);
671 
672  /*
673  * If we have most-common-values info, add up the fractions of the MCV
674  * entries that satisfy MCV OP CONST. These fractions contribute directly
675  * to the result selectivity. Also add up the total fraction represented
676  * by MCV entries.
677  */
678  mcv_selec = mcv_selectivity(vardata, &opproc, collation, constval, true,
679  &sumcommon);
680 
681  /*
682  * If there is a histogram, determine which bin the constant falls in, and
683  * compute the resulting contribution to selectivity.
684  */
685  hist_selec = ineq_histogram_selectivity(root, vardata,
686  operator, &opproc, isgt, iseq,
687  collation,
688  constval, consttype);
689 
690  /*
691  * Now merge the results from the MCV and histogram calculations,
692  * realizing that the histogram covers only the non-null values that are
693  * not listed in MCV.
694  */
695  selec = 1.0 - stats->stanullfrac - sumcommon;
696 
697  if (hist_selec >= 0.0)
698  selec *= hist_selec;
699  else
700  {
701  /*
702  * If no histogram but there are values not accounted for by MCV,
703  * arbitrarily assume half of them will match.
704  */
705  selec *= 0.5;
706  }
707 
708  selec += mcv_selec;
709 
710  /* result should be in range, but make sure... */
711  CLAMP_PROBABILITY(selec);
712 
713  return selec;
714 }
715 
716 /*
717  * mcv_selectivity - Examine the MCV list for selectivity estimates
718  *
719  * Determine the fraction of the variable's MCV population that satisfies
720  * the predicate (VAR OP CONST), or (CONST OP VAR) if !varonleft. Also
721  * compute the fraction of the total column population represented by the MCV
722  * list. This code will work for any boolean-returning predicate operator.
723  *
724  * The function result is the MCV selectivity, and the fraction of the
725  * total population is returned into *sumcommonp. Zeroes are returned
726  * if there is no MCV list.
727  */
728 double
729 mcv_selectivity(VariableStatData *vardata, FmgrInfo *opproc, Oid collation,
730  Datum constval, bool varonleft,
731  double *sumcommonp)
732 {
733  double mcv_selec,
734  sumcommon;
735  AttStatsSlot sslot;
736  int i;
737 
738  mcv_selec = 0.0;
739  sumcommon = 0.0;
740 
741  if (HeapTupleIsValid(vardata->statsTuple) &&
742  statistic_proc_security_check(vardata, opproc->fn_oid) &&
743  get_attstatsslot(&sslot, vardata->statsTuple,
744  STATISTIC_KIND_MCV, InvalidOid,
745  ATTSTATSSLOT_VALUES | ATTSTATSSLOT_NUMBERS))
746  {
747  LOCAL_FCINFO(fcinfo, 2);
748 
749  /*
750  * We invoke the opproc "by hand" so that we won't fail on NULL
751  * results. Such cases won't arise for normal comparison functions,
752  * but generic_restriction_selectivity could perhaps be used with
753  * operators that can return NULL. A small side benefit is to not
754  * need to re-initialize the fcinfo struct from scratch each time.
755  */
756  InitFunctionCallInfoData(*fcinfo, opproc, 2, collation,
757  NULL, NULL);
758  fcinfo->args[0].isnull = false;
759  fcinfo->args[1].isnull = false;
760  /* be careful to apply operator right way 'round */
761  if (varonleft)
762  fcinfo->args[1].value = constval;
763  else
764  fcinfo->args[0].value = constval;
765 
766  for (i = 0; i < sslot.nvalues; i++)
767  {
768  Datum fresult;
769 
770  if (varonleft)
771  fcinfo->args[0].value = sslot.values[i];
772  else
773  fcinfo->args[1].value = sslot.values[i];
774  fcinfo->isnull = false;
775  fresult = FunctionCallInvoke(fcinfo);
776  if (!fcinfo->isnull && DatumGetBool(fresult))
777  mcv_selec += sslot.numbers[i];
778  sumcommon += sslot.numbers[i];
779  }
780  free_attstatsslot(&sslot);
781  }
782 
783  *sumcommonp = sumcommon;
784  return mcv_selec;
785 }
786 
787 /*
788  * histogram_selectivity - Examine the histogram for selectivity estimates
789  *
790  * Determine the fraction of the variable's histogram entries that satisfy
791  * the predicate (VAR OP CONST), or (CONST OP VAR) if !varonleft.
792  *
793  * This code will work for any boolean-returning predicate operator, whether
794  * or not it has anything to do with the histogram sort operator. We are
795  * essentially using the histogram just as a representative sample. However,
796  * small histograms are unlikely to be all that representative, so the caller
797  * should be prepared to fall back on some other estimation approach when the
798  * histogram is missing or very small. It may also be prudent to combine this
799  * approach with another one when the histogram is small.
800  *
801  * If the actual histogram size is not at least min_hist_size, we won't bother
802  * to do the calculation at all. Also, if the n_skip parameter is > 0, we
803  * ignore the first and last n_skip histogram elements, on the grounds that
804  * they are outliers and hence not very representative. Typical values for
805  * these parameters are 10 and 1.
806  *
807  * The function result is the selectivity, or -1 if there is no histogram
808  * or it's smaller than min_hist_size.
809  *
810  * The output parameter *hist_size receives the actual histogram size,
811  * or zero if no histogram. Callers may use this number to decide how
812  * much faith to put in the function result.
813  *
814  * Note that the result disregards both the most-common-values (if any) and
815  * null entries. The caller is expected to combine this result with
816  * statistics for those portions of the column population. It may also be
817  * prudent to clamp the result range, ie, disbelieve exact 0 or 1 outputs.
818  */
819 double
820 histogram_selectivity(VariableStatData *vardata,
821  FmgrInfo *opproc, Oid collation,
822  Datum constval, bool varonleft,
823  int min_hist_size, int n_skip,
824  int *hist_size)
825 {
826  double result;
827  AttStatsSlot sslot;
828 
829  /* check sanity of parameters */
830  Assert(n_skip >= 0);
831  Assert(min_hist_size > 2 * n_skip);
832 
833  if (HeapTupleIsValid(vardata->statsTuple) &&
834  statistic_proc_security_check(vardata, opproc->fn_oid) &&
835  get_attstatsslot(&sslot, vardata->statsTuple,
836  STATISTIC_KIND_HISTOGRAM, InvalidOid,
837  ATTSTATSSLOT_VALUES))
838  {
839  *hist_size = sslot.nvalues;
840  if (sslot.nvalues >= min_hist_size)
841  {
842  LOCAL_FCINFO(fcinfo, 2);
843  int nmatch = 0;
844  int i;
845 
846  /*
847  * We invoke the opproc "by hand" so that we won't fail on NULL
848  * results. Such cases won't arise for normal comparison
849  * functions, but generic_restriction_selectivity could perhaps be
850  * used with operators that can return NULL. A small side benefit
851  * is to not need to re-initialize the fcinfo struct from scratch
852  * each time.
853  */
854  InitFunctionCallInfoData(*fcinfo, opproc, 2, collation,
855  NULL, NULL);
856  fcinfo->args[0].isnull = false;
857  fcinfo->args[1].isnull = false;
858  /* be careful to apply operator right way 'round */
859  if (varonleft)
860  fcinfo->args[1].value = constval;
861  else
862  fcinfo->args[0].value = constval;
863 
864  for (i = n_skip; i < sslot.nvalues - n_skip; i++)
865  {
866  Datum fresult;
867 
868  if (varonleft)
869  fcinfo->args[0].value = sslot.values[i];
870  else
871  fcinfo->args[1].value = sslot.values[i];
872  fcinfo->isnull = false;
873  fresult = FunctionCallInvoke(fcinfo);
874  if (!fcinfo->isnull && DatumGetBool(fresult))
875  nmatch++;
876  }
877  result = ((double) nmatch) / ((double) (sslot.nvalues - 2 * n_skip));
878  }
879  else
880  result = -1;
881  free_attstatsslot(&sslot);
882  }
883  else
884  {
885  *hist_size = 0;
886  result = -1;
887  }
888 
889  return result;
890 }
891 
892 /*
893  * generic_restriction_selectivity - Selectivity for almost anything
894  *
895  * This function estimates selectivity for operators that we don't have any
896  * special knowledge about, but are on data types that we collect standard
897  * MCV and/or histogram statistics for. (Additional assumptions are that
898  * the operator is strict and immutable, or at least stable.)
899  *
900  * If we have "VAR OP CONST" or "CONST OP VAR", selectivity is estimated by
901  * applying the operator to each element of the column's MCV and/or histogram
902  * stats, and merging the results using the assumption that the histogram is
903  * a reasonable random sample of the column's non-MCV population. Note that
904  * if the operator's semantics are related to the histogram ordering, this
905  * might not be such a great assumption; other functions such as
906  * scalarineqsel() are probably a better match in such cases.
907  *
908  * Otherwise, fall back to the default selectivity provided by the caller.
909  */
910 double
911 generic_restriction_selectivity(PlannerInfo *root, Oid oproid, Oid collation,
912  List *args, int varRelid,
913  double default_selectivity)
914 {
915  double selec;
916  VariableStatData vardata;
917  Node *other;
918  bool varonleft;
919 
920  /*
921  * If expression is not variable OP something or something OP variable,
922  * then punt and return the default estimate.
923  */
924  if (!get_restriction_variable(root, args, varRelid,
925  &vardata, &other, &varonleft))
926  return default_selectivity;
927 
928  /*
929  * If the something is a NULL constant, assume operator is strict and
930  * return zero, ie, operator will never return TRUE.
931  */
932  if (IsA(other, Const) &&
933  ((Const *) other)->constisnull)
934  {
935  ReleaseVariableStats(vardata);
936  return 0.0;
937  }
938 
939  if (IsA(other, Const))
940  {
941  /* Variable is being compared to a known non-null constant */
942  Datum constval = ((Const *) other)->constvalue;
943  FmgrInfo opproc;
944  double mcvsum;
945  double mcvsel;
946  double nullfrac;
947  int hist_size;
948 
949  fmgr_info(get_opcode(oproid), &opproc);
950 
951  /*
952  * Calculate the selectivity for the column's most common values.
953  */
954  mcvsel = mcv_selectivity(&vardata, &opproc, collation,
955  constval, varonleft,
956  &mcvsum);
957 
958  /*
959  * If the histogram is large enough, see what fraction of it matches
960  * the query, and assume that's representative of the non-MCV
961  * population. Otherwise use the default selectivity for the non-MCV
962  * population.
963  */
964  selec = histogram_selectivity(&vardata, &opproc, collation,
965  constval, varonleft,
966  10, 1, &hist_size);
967  if (selec < 0)
968  {
969  /* Nope, fall back on default */
970  selec = default_selectivity;
971  }
972  else if (hist_size < 100)
973  {
974  /*
975  * For histogram sizes from 10 to 100, we combine the histogram
976  * and default selectivities, putting increasingly more trust in
977  * the histogram for larger sizes.
978  */
979  double hist_weight = hist_size / 100.0;
980 
981  selec = selec * hist_weight +
982  default_selectivity * (1.0 - hist_weight);
983  }
984 
985  /* In any case, don't believe extremely small or large estimates. */
986  if (selec < 0.0001)
987  selec = 0.0001;
988  else if (selec > 0.9999)
989  selec = 0.9999;
990 
991  /* Don't forget to account for nulls. */
992  if (HeapTupleIsValid(vardata.statsTuple))
993  nullfrac = ((Form_pg_statistic) GETSTRUCT(vardata.statsTuple))->stanullfrac;
994  else
995  nullfrac = 0.0;
996 
997  /*
998  * Now merge the results from the MCV and histogram calculations,
999  * realizing that the histogram covers only the non-null values that
1000  * are not listed in MCV.
1001  */
1002  selec *= 1.0 - nullfrac - mcvsum;
1003  selec += mcvsel;
1004  }
1005  else
1006  {
1007  /* Comparison value is not constant, so we can't do anything */
1008  selec = default_selectivity;
1009  }
1010 
1011  ReleaseVariableStats(vardata);
1012 
1013  /* result should be in range, but make sure... */
1014  CLAMP_PROBABILITY(selec);
1015 
1016  return selec;
1017 }
1018 
1019 /*
1020  * ineq_histogram_selectivity - Examine the histogram for scalarineqsel
1021  *
1022  * Determine the fraction of the variable's histogram population that
1023  * satisfies the inequality condition, ie, VAR < (or <=, >, >=) CONST.
1024  * The isgt and iseq flags distinguish which of the four cases apply.
1025  *
1026  * While opproc could be looked up from the operator OID, common callers
1027  * also need to call it separately, so we make the caller pass both.
1028  *
1029  * Returns -1 if there is no histogram (valid results will always be >= 0).
1030  *
1031  * Note that the result disregards both the most-common-values (if any) and
1032  * null entries. The caller is expected to combine this result with
1033  * statistics for those portions of the column population.
1034  *
1035  * This is exported so that some other estimation functions can use it.
1036  */
1037 double
1038 ineq_histogram_selectivity(PlannerInfo *root,
1039  VariableStatData *vardata,
1040  Oid opoid, FmgrInfo *opproc, bool isgt, bool iseq,
1041  Oid collation,
1042  Datum constval, Oid consttype)
1043 {
1044  double hist_selec;
1045  AttStatsSlot sslot;
1046 
1047  hist_selec = -1.0;
1048 
1049  /*
1050  * Someday, ANALYZE might store more than one histogram per rel/att,
1051  * corresponding to more than one possible sort ordering defined for the
1052  * column type. Right now, we know there is only one, so just grab it and
1053  * see if it matches the query.
1054  *
1055  * Note that we can't use opoid as search argument; the staop appearing in
1056  * pg_statistic will be for the relevant '<' operator, but what we have
1057  * might be some other inequality operator such as '>='. (Even if opoid
1058  * is a '<' operator, it could be cross-type.) Hence we must use
1059  * comparison_ops_are_compatible() to see if the operators match.
1060  */
1061  if (HeapTupleIsValid(vardata->statsTuple) &&
1062  statistic_proc_security_check(vardata, opproc->fn_oid) &&
1063  get_attstatsslot(&sslot, vardata->statsTuple,
1064  STATISTIC_KIND_HISTOGRAM, InvalidOid,
1065  ATTSTATSSLOT_VALUES))
1066  {
1067  if (sslot.nvalues > 1 &&
1068  sslot.stacoll == collation &&
1069  comparison_ops_are_compatible(sslot.staop, opoid))
1070  {
1071  /*
1072  * Use binary search to find the desired location, namely the
1073  * right end of the histogram bin containing the comparison value,
1074  * which is the leftmost entry for which the comparison operator
1075  * succeeds (if isgt) or fails (if !isgt).
1076  *
1077  * In this loop, we pay no attention to whether the operator iseq
1078  * or not; that detail will be mopped up below. (We cannot tell,
1079  * anyway, whether the operator thinks the values are equal.)
1080  *
1081  * If the binary search accesses the first or last histogram
1082  * entry, we try to replace that endpoint with the true column min
1083  * or max as found by get_actual_variable_range(). This
1084  * ameliorates misestimates when the min or max is moving as a
1085  * result of changes since the last ANALYZE. Note that this could
1086  * result in effectively including MCVs into the histogram that
1087  * weren't there before, but we don't try to correct for that.
1088  */
1089  double histfrac;
1090  int lobound = 0; /* first possible slot to search */
1091  int hibound = sslot.nvalues; /* last+1 slot to search */
1092  bool have_end = false;
1093 
1094  /*
1095  * If there are only two histogram entries, we'll want up-to-date
1096  * values for both. (If there are more than two, we need at most
1097  * one of them to be updated, so we deal with that within the
1098  * loop.)
1099  */
1100  if (sslot.nvalues == 2)
1101  have_end = get_actual_variable_range(root,
1102  vardata,
1103  sslot.staop,
1104  collation,
1105  &sslot.values[0],
1106  &sslot.values[1]);
1107 
1108  while (lobound < hibound)
1109  {
1110  int probe = (lobound + hibound) / 2;
1111  bool ltcmp;
1112 
1113  /*
1114  * If we find ourselves about to compare to the first or last
1115  * histogram entry, first try to replace it with the actual
1116  * current min or max (unless we already did so above).
1117  */
1118  if (probe == 0 && sslot.nvalues > 2)
1119  have_end = get_actual_variable_range(root,
1120  vardata,
1121  sslot.staop,
1122  collation,
1123  &sslot.values[0],
1124  NULL);
1125  else if (probe == sslot.nvalues - 1 && sslot.nvalues > 2)
1126  have_end = get_actual_variable_range(root,
1127  vardata,
1128  sslot.staop,
1129  collation,
1130  NULL,
1131  &sslot.values[probe]);
1132 
1133  ltcmp = DatumGetBool(FunctionCall2Coll(opproc,
1134  collation,
1135  sslot.values[probe],
1136  constval));
1137  if (isgt)
1138  ltcmp = !ltcmp;
1139  if (ltcmp)
1140  lobound = probe + 1;
1141  else
1142  hibound = probe;
1143  }
1144 
1145  if (lobound <= 0)
1146  {
1147  /*
1148  * Constant is below lower histogram boundary. More
1149  * precisely, we have found that no entry in the histogram
1150  * satisfies the inequality clause (if !isgt) or they all do
1151  * (if isgt). We estimate that that's true of the entire
1152  * table, so set histfrac to 0.0 (which we'll flip to 1.0
1153  * below, if isgt).
1154  */
1155  histfrac = 0.0;
1156  }
1157  else if (lobound >= sslot.nvalues)
1158  {
1159  /*
1160  * Inverse case: constant is above upper histogram boundary.
1161  */
1162  histfrac = 1.0;
1163  }
1164  else
1165  {
1166  /* We have values[i-1] <= constant <= values[i]. */
1167  int i = lobound;
1168  double eq_selec = 0;
1169  double val,
1170  high,
1171  low;
1172  double binfrac;
1173 
1174  /*
1175  * In the cases where we'll need it below, obtain an estimate
1176  * of the selectivity of "x = constval". We use a calculation
1177  * similar to what var_eq_const() does for a non-MCV constant,
1178  * ie, estimate that all distinct non-MCV values occur equally
1179  * often. But multiplication by "1.0 - sumcommon - nullfrac"
1180  * will be done by our caller, so we shouldn't do that here.
1181  * Therefore we can't try to clamp the estimate by reference
1182  * to the least common MCV; the result would be too small.
1183  *
1184  * Note: since this is effectively assuming that constval
1185  * isn't an MCV, it's logically dubious if constval in fact is
1186  * one. But we have to apply *some* correction for equality,
1187  * and anyway we cannot tell if constval is an MCV, since we
1188  * don't have a suitable equality operator at hand.
1189  */
1190  if (i == 1 || isgt == iseq)
1191  {
1192  double otherdistinct;
1193  bool isdefault;
1194  AttStatsSlot mcvslot;
1195 
1196  /* Get estimated number of distinct values */
1197  otherdistinct = get_variable_numdistinct(vardata,
1198  &isdefault);
1199 
1200  /* Subtract off the number of known MCVs */
1201  if (get_attstatsslot(&mcvslot, vardata->statsTuple,
1202  STATISTIC_KIND_MCV, InvalidOid,
1203  ATTSTATSSLOT_NUMBERS))
1204  {
1205  otherdistinct -= mcvslot.nnumbers;
1206  free_attstatsslot(&mcvslot);
1207  }
1208 
1209  /* If result doesn't seem sane, leave eq_selec at 0 */
1210  if (otherdistinct > 1)
1211  eq_selec = 1.0 / otherdistinct;
1212  }
1213 
1214  /*
1215  * Convert the constant and the two nearest bin boundary
1216  * values to a uniform comparison scale, and do a linear
1217  * interpolation within this bin.
1218  */
1219  if (convert_to_scalar(constval, consttype, collation,
1220  &val,
1221  sslot.values[i - 1], sslot.values[i],
1222  vardata->vartype,
1223  &low, &high))
1224  {
1225  if (high <= low)
1226  {
1227  /* cope if bin boundaries appear identical */
1228  binfrac = 0.5;
1229  }
1230  else if (val <= low)
1231  binfrac = 0.0;
1232  else if (val >= high)
1233  binfrac = 1.0;
1234  else
1235  {
1236  binfrac = (val - low) / (high - low);
1237 
1238  /*
1239  * Watch out for the possibility that we got a NaN or
1240  * Infinity from the division. This can happen
1241  * despite the previous checks, if for example "low"
1242  * is -Infinity.
1243  */
1244  if (isnan(binfrac) ||
1245  binfrac < 0.0 || binfrac > 1.0)
1246  binfrac = 0.5;
1247  }
1248  }
1249  else
1250  {
1251  /*
1252  * Ideally we'd produce an error here, on the grounds that
1253  * the given operator shouldn't have scalarXXsel
1254  * registered as its selectivity func unless we can deal
1255  * with its operand types. But currently, all manner of
1256  * stuff is invoking scalarXXsel, so give a default
1257  * estimate until that can be fixed.
1258  */
1259  binfrac = 0.5;
1260  }
1261 
1262  /*
1263  * Now, compute the overall selectivity across the values
1264  * represented by the histogram. We have i-1 full bins and
1265  * binfrac partial bin below the constant.
1266  */
1267  histfrac = (double) (i - 1) + binfrac;
1268  histfrac /= (double) (sslot.nvalues - 1);
1269 
1270  /*
1271  * At this point, histfrac is an estimate of the fraction of
1272  * the population represented by the histogram that satisfies
1273  * "x <= constval". Somewhat remarkably, this statement is
1274  * true regardless of which operator we were doing the probes
1275  * with, so long as convert_to_scalar() delivers reasonable
1276  * results. If the probe constant is equal to some histogram
1277  * entry, we would have considered the bin to the left of that
1278  * entry if probing with "<" or ">=", or the bin to the right
1279  * if probing with "<=" or ">"; but binfrac would have come
1280  * out as 1.0 in the first case and 0.0 in the second, leading
1281  * to the same histfrac in either case. For probe constants
1282  * between histogram entries, we find the same bin and get the
1283  * same estimate with any operator.
1284  *
1285  * The fact that the estimate corresponds to "x <= constval"
1286  * and not "x < constval" is because of the way that ANALYZE
1287  * constructs the histogram: each entry is, effectively, the
1288  * rightmost value in its sample bucket. So selectivity
1289  * values that are exact multiples of 1/(histogram_size-1)
1290  * should be understood as estimates including a histogram
1291  * entry plus everything to its left.
1292  *
1293  * However, that breaks down for the first histogram entry,
1294  * which necessarily is the leftmost value in its sample
1295  * bucket. That means the first histogram bin is slightly
1296  * narrower than the rest, by an amount equal to eq_selec.
1297  * Another way to say that is that we want "x <= leftmost" to
1298  * be estimated as eq_selec not zero. So, if we're dealing
1299  * with the first bin (i==1), rescale to make that true while
1300  * adjusting the rest of that bin linearly.
1301  */
1302  if (i == 1)
1303  histfrac += eq_selec * (1.0 - binfrac);
1304 
1305  /*
1306  * "x <= constval" is good if we want an estimate for "<=" or
1307  * ">", but if we are estimating for "<" or ">=", we now need
1308  * to decrease the estimate by eq_selec.
1309  */
1310  if (isgt == iseq)
1311  histfrac -= eq_selec;
1312  }
1313 
1314  /*
1315  * Now the estimate is finished for "<" and "<=" cases. If we are
1316  * estimating for ">" or ">=", flip it.
1317  */
1318  hist_selec = isgt ? (1.0 - histfrac) : histfrac;
1319 
1320  /*
1321  * The histogram boundaries are only approximate to begin with,
1322  * and may well be out of date anyway. Therefore, don't believe
1323  * extremely small or large selectivity estimates --- unless we
1324  * got actual current endpoint values from the table, in which
1325  * case just do the usual sanity clamp. Somewhat arbitrarily, we
1326  * set the cutoff for other cases at a hundredth of the histogram
1327  * resolution.
1328  */
1329  if (have_end)
1330  CLAMP_PROBABILITY(hist_selec);
1331  else
1332  {
1333  double cutoff = 0.01 / (double) (sslot.nvalues - 1);
1334 
1335  if (hist_selec < cutoff)
1336  hist_selec = cutoff;
1337  else if (hist_selec > 1.0 - cutoff)
1338  hist_selec = 1.0 - cutoff;
1339  }
1340  }
1341  else if (sslot.nvalues > 1)
1342  {
1343  /*
1344  * If we get here, we have a histogram but it's not sorted the way
1345  * we want. Do a brute-force search to see how many of the
1346  * entries satisfy the comparison condition, and take that
1347  * fraction as our estimate. (This is identical to the inner loop
1348  * of histogram_selectivity; maybe share code?)
1349  */
1350  LOCAL_FCINFO(fcinfo, 2);
1351  int nmatch = 0;
1352 
1353  InitFunctionCallInfoData(*fcinfo, opproc, 2, collation,
1354  NULL, NULL);
1355  fcinfo->args[0].isnull = false;
1356  fcinfo->args[1].isnull = false;
1357  fcinfo->args[1].value = constval;
1358  for (int i = 0; i < sslot.nvalues; i++)
1359  {
1360  Datum fresult;
1361 
1362  fcinfo->args[0].value = sslot.values[i];
1363  fcinfo->isnull = false;
1364  fresult = FunctionCallInvoke(fcinfo);
1365  if (!fcinfo->isnull && DatumGetBool(fresult))
1366  nmatch++;
1367  }
1368  hist_selec = ((double) nmatch) / ((double) sslot.nvalues);
1369 
1370  /*
1371  * As above, clamp to a hundredth of the histogram resolution.
1372  * This case is surely even less trustworthy than the normal one,
1373  * so we shouldn't believe exact 0 or 1 selectivity. (Maybe the
1374  * clamp should be more restrictive in this case?)
1375  */
1376  {
1377  double cutoff = 0.01 / (double) (sslot.nvalues - 1);
1378 
1379  if (hist_selec < cutoff)
1380  hist_selec = cutoff;
1381  else if (hist_selec > 1.0 - cutoff)
1382  hist_selec = 1.0 - cutoff;
1383  }
1384  }
1385 
1386  free_attstatsslot(&sslot);
1387  }
1388 
1389  return hist_selec;
1390 }
1391 
1392 /*
1393  * Common wrapper function for the selectivity estimators that simply
1394  * invoke scalarineqsel().
1395  */
1396 static Datum
1397 scalarineqsel_wrapper(PG_FUNCTION_ARGS, bool isgt, bool iseq)
1398 {
1399  PlannerInfo *root = (PlannerInfo *) PG_GETARG_POINTER(0);
1400  Oid operator = PG_GETARG_OID(1);
1401  List *args = (List *) PG_GETARG_POINTER(2);
1402  int varRelid = PG_GETARG_INT32(3);
1403  Oid collation = PG_GET_COLLATION();
1404  VariableStatData vardata;
1405  Node *other;
1406  bool varonleft;
1407  Datum constval;
1408  Oid consttype;
1409  double selec;
1410 
1411  /*
1412  * If expression is not variable op something or something op variable,
1413  * then punt and return a default estimate.
1414  */
1415  if (!get_restriction_variable(root, args, varRelid,
1416  &vardata, &other, &varonleft))
1417  PG_RETURN_FLOAT8(DEFAULT_INEQ_SEL);
1418 
1419  /*
1420  * Can't do anything useful if the something is not a constant, either.
1421  */
1422  if (!IsA(other, Const))
1423  {
1424  ReleaseVariableStats(vardata);
1425  PG_RETURN_FLOAT8(DEFAULT_INEQ_SEL);
1426  }
1427 
1428  /*
1429  * If the constant is NULL, assume operator is strict and return zero, ie,
1430  * operator will never return TRUE.
1431  */
1432  if (((Const *) other)->constisnull)
1433  {
1434  ReleaseVariableStats(vardata);
1435  PG_RETURN_FLOAT8(0.0);
1436  }
1437  constval = ((Const *) other)->constvalue;
1438  consttype = ((Const *) other)->consttype;
1439 
1440  /*
1441  * Force the var to be on the left to simplify logic in scalarineqsel.
1442  */
1443  if (!varonleft)
1444  {
1445  operator = get_commutator(operator);
1446  if (!operator)
1447  {
1448  /* Use default selectivity (should we raise an error instead?) */
1449  ReleaseVariableStats(vardata);
1450  PG_RETURN_FLOAT8(DEFAULT_INEQ_SEL);
1451  }
1452  isgt = !isgt;
1453  }
1454 
1455  /* The rest of the work is done by scalarineqsel(). */
1456  selec = scalarineqsel(root, operator, isgt, iseq, collation,
1457  &vardata, constval, consttype);
1458 
1459  ReleaseVariableStats(vardata);
1460 
1461  PG_RETURN_FLOAT8((float8) selec);
1462 }
1463 
1464 /*
1465  * scalarltsel - Selectivity of "<" for scalars.
1466  */
1467 Datum
1468 scalarltsel(PG_FUNCTION_ARGS)
1469 {
1470  return scalarineqsel_wrapper(fcinfo, false, false);
1471 }
1472 
1473 /*
1474  * scalarlesel - Selectivity of "<=" for scalars.
1475  */
1476 Datum
1477 scalarlesel(PG_FUNCTION_ARGS)
1478 {
1479  return scalarineqsel_wrapper(fcinfo, false, true);
1480 }
1481 
1482 /*
1483  * scalargtsel - Selectivity of ">" for scalars.
1484  */
1485 Datum
1486 scalargtsel(PG_FUNCTION_ARGS)
1487 {
1488  return scalarineqsel_wrapper(fcinfo, true, false);
1489 }
1490 
1491 /*
1492  * scalargesel - Selectivity of ">=" for scalars.
1493  */
1494 Datum
1495 scalargesel(PG_FUNCTION_ARGS)
1496 {
1497  return scalarineqsel_wrapper(fcinfo, true, true);
1498 }
1499 
1500 /*
1501  * boolvarsel - Selectivity of Boolean variable.
1502  *
1503  * This can actually be called on any boolean-valued expression. If it
1504  * involves only Vars of the specified relation, and if there are statistics
1505  * about the Var or expression (the latter is possible if it's indexed) then
1506  * we'll produce a real estimate; otherwise it's just a default.
1507  */
1508 Selectivity
1509 boolvarsel(PlannerInfo *root, Node *arg, int varRelid)
1510 {
1511  VariableStatData vardata;
1512  double selec;
1513 
1514  examine_variable(root, arg, varRelid, &vardata);
1515  if (HeapTupleIsValid(vardata.statsTuple))
1516  {
1517  /*
1518  * A boolean variable V is equivalent to the clause V = 't', so we
1519  * compute the selectivity as if that is what we have.
1520  */
1521  selec = var_eq_const(&vardata, BooleanEqualOperator, InvalidOid,
1522  BoolGetDatum(true), false, true, false);
1523  }
1524  else
1525  {
1526  /* Otherwise, the default estimate is 0.5 */
1527  selec = 0.5;
1528  }
1529  ReleaseVariableStats(vardata);
1530  return selec;
1531 }
1532 
1533 /*
1534  * booltestsel - Selectivity of BooleanTest Node.
1535  */
1536 Selectivity
1537 booltestsel(PlannerInfo *root, BoolTestType booltesttype, Node *arg,
1538  int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo)
1539 {
1540  VariableStatData vardata;
1541  double selec;
1542 
1543  examine_variable(root, arg, varRelid, &vardata);
1544 
1545  if (HeapTupleIsValid(vardata.statsTuple))
1546  {
1547  Form_pg_statistic stats;
1548  double freq_null;
1549  AttStatsSlot sslot;
1550 
1551  stats = (Form_pg_statistic) GETSTRUCT(vardata.statsTuple);
1552  freq_null = stats->stanullfrac;
1553 
1554  if (get_attstatsslot(&sslot, vardata.statsTuple,
1555  STATISTIC_KIND_MCV, InvalidOid,
1556  ATTSTATSSLOT_VALUES | ATTSTATSSLOT_NUMBERS)
1557  && sslot.nnumbers > 0)
1558  {
1559  double freq_true;
1560  double freq_false;
1561 
1562  /*
1563  * Get first MCV frequency and derive frequency for true.
1564  */
1565  if (DatumGetBool(sslot.values[0]))
1566  freq_true = sslot.numbers[0];
1567  else
1568  freq_true = 1.0 - sslot.numbers[0] - freq_null;
1569 
1570  /*
1571  * Next derive frequency for false. Then use these as appropriate
1572  * to derive frequency for each case.
1573  */
1574  freq_false = 1.0 - freq_true - freq_null;
1575 
1576  switch (booltesttype)
1577  {
1578  case IS_UNKNOWN:
1579  /* select only NULL values */
1580  selec = freq_null;
1581  break;
1582  case IS_NOT_UNKNOWN:
1583  /* select non-NULL values */
1584  selec = 1.0 - freq_null;
1585  break;
1586  case IS_TRUE:
1587  /* select only TRUE values */
1588  selec = freq_true;
1589  break;
1590  case IS_NOT_TRUE:
1591  /* select non-TRUE values */
1592  selec = 1.0 - freq_true;
1593  break;
1594  case IS_FALSE:
1595  /* select only FALSE values */
1596  selec = freq_false;
1597  break;
1598  case IS_NOT_FALSE:
1599  /* select non-FALSE values */
1600  selec = 1.0 - freq_false;
1601  break;
1602  default:
1603  elog(ERROR, "unrecognized booltesttype: %d",
1604  (int) booltesttype);
1605  selec = 0.0; /* Keep compiler quiet */
1606  break;
1607  }
1608 
1609  free_attstatsslot(&sslot);
1610  }
1611  else
1612  {
1613  /*
1614  * No most-common-value info available. Still have null fraction
1615  * information, so use it for IS [NOT] UNKNOWN. Otherwise adjust
1616  * for null fraction and assume a 50-50 split of TRUE and FALSE.
1617  */
1618  switch (booltesttype)
1619  {
1620  case IS_UNKNOWN:
1621  /* select only NULL values */
1622  selec = freq_null;
1623  break;
1624  case IS_NOT_UNKNOWN:
1625  /* select non-NULL values */
1626  selec = 1.0 - freq_null;
1627  break;
1628  case IS_TRUE:
1629  case IS_FALSE:
1630  /* Assume we select half of the non-NULL values */
1631  selec = (1.0 - freq_null) / 2.0;
1632  break;
1633  case IS_NOT_TRUE:
1634  case IS_NOT_FALSE:
1635  /* Assume we select NULLs plus half of the non-NULLs */
1636  /* equiv. to freq_null + (1.0 - freq_null) / 2.0 */
1637  selec = (freq_null + 1.0) / 2.0;
1638  break;
1639  default:
1640  elog(ERROR, "unrecognized booltesttype: %d",
1641  (int) booltesttype);
1642  selec = 0.0; /* Keep compiler quiet */
1643  break;
1644  }
1645  }
1646  }
1647  else
1648  {
1649  /*
1650  * If we can't get variable statistics for the argument, perhaps
1651  * clause_selectivity can do something with it. We ignore the
1652  * possibility of a NULL value when using clause_selectivity, and just
1653  * assume the value is either TRUE or FALSE.
1654  */
1655  switch (booltesttype)
1656  {
1657  case IS_UNKNOWN:
1658  selec = DEFAULT_UNK_SEL;
1659  break;
1660  case IS_NOT_UNKNOWN:
1661  selec = DEFAULT_NOT_UNK_SEL;
1662  break;
1663  case IS_TRUE:
1664  case IS_NOT_FALSE:
1665  selec = (double) clause_selectivity(root, arg,
1666  varRelid,
1667  jointype, sjinfo);
1668  break;
1669  case IS_FALSE:
1670  case IS_NOT_TRUE:
1671  selec = 1.0 - (double) clause_selectivity(root, arg,
1672  varRelid,
1673  jointype, sjinfo);
1674  break;
1675  default:
1676  elog(ERROR, "unrecognized booltesttype: %d",
1677  (int) booltesttype);
1678  selec = 0.0; /* Keep compiler quiet */
1679  break;
1680  }
1681  }
1682 
1683  ReleaseVariableStats(vardata);
1684 
1685  /* result should be in range, but make sure... */
1686  CLAMP_PROBABILITY(selec);
1687 
1688  return (Selectivity) selec;
1689 }
1690 
1691 /*
1692  * nulltestsel - Selectivity of NullTest Node.
1693  */
1694 Selectivity
1695 nulltestsel(PlannerInfo *root, NullTestType nulltesttype, Node *arg,
1696  int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo)
1697 {
1698  VariableStatData vardata;
1699  double selec;
1700 
1701  examine_variable(root, arg, varRelid, &vardata);
1702 
1703  if (HeapTupleIsValid(vardata.statsTuple))
1704  {
1705  Form_pg_statistic stats;
1706  double freq_null;
1707 
1708  stats = (Form_pg_statistic) GETSTRUCT(vardata.statsTuple);
1709  freq_null = stats->stanullfrac;
1710 
1711  switch (nulltesttype)
1712  {
1713  case IS_NULL:
1714 
1715  /*
1716  * Use freq_null directly.
1717  */
1718  selec = freq_null;
1719  break;
1720  case IS_NOT_NULL:
1721 
1722  /*
1723  * Select not unknown (not null) values. Calculate from
1724  * freq_null.
1725  */
1726  selec = 1.0 - freq_null;
1727  break;
1728  default:
1729  elog(ERROR, "unrecognized nulltesttype: %d",
1730  (int) nulltesttype);
1731  return (Selectivity) 0; /* keep compiler quiet */
1732  }
1733  }
1734  else if (vardata.var && IsA(vardata.var, Var) &&
1735  ((Var *) vardata.var)->varattno < 0)
1736  {
1737  /*
1738  * There are no stats for system columns, but we know they are never
1739  * NULL.
1740  */
1741  selec = (nulltesttype == IS_NULL) ? 0.0 : 1.0;
1742  }
1743  else
1744  {
1745  /*
1746  * No ANALYZE stats available, so make a guess
1747  */
1748  switch (nulltesttype)
1749  {
1750  case IS_NULL:
1751  selec = DEFAULT_UNK_SEL;
1752  break;
1753  case IS_NOT_NULL:
1754  selec = DEFAULT_NOT_UNK_SEL;
1755  break;
1756  default:
1757  elog(ERROR, "unrecognized nulltesttype: %d",
1758  (int) nulltesttype);
1759  return (Selectivity) 0; /* keep compiler quiet */
1760  }
1761  }
1762 
1763  ReleaseVariableStats(vardata);
1764 
1765  /* result should be in range, but make sure... */
1766  CLAMP_PROBABILITY(selec);
1767 
1768  return (Selectivity) selec;
1769 }
1770 
1771 /*
1772  * strip_array_coercion - strip binary-compatible relabeling from an array expr
1773  *
1774  * For array values, the parser normally generates ArrayCoerceExpr conversions,
1775  * but it seems possible that RelabelType might show up. Also, the planner
1776  * is not currently tense about collapsing stacked ArrayCoerceExpr nodes,
1777  * so we need to be ready to deal with more than one level.
1778  */
1779 static Node *
1780 strip_array_coercion(Node *node)
1781 {
1782  for (;;)
1783  {
1784  if (node && IsA(node, ArrayCoerceExpr))
1785  {
1786  ArrayCoerceExpr *acoerce = (ArrayCoerceExpr *) node;
1787 
1788  /*
1789  * If the per-element expression is just a RelabelType on top of
1790  * CaseTestExpr, then we know it's a binary-compatible relabeling.
1791  */
1792  if (IsA(acoerce->elemexpr, RelabelType) &&
1793  IsA(((RelabelType *) acoerce->elemexpr)->arg, CaseTestExpr))
1794  node = (Node *) acoerce->arg;
1795  else
1796  break;
1797  }
1798  else if (node && IsA(node, RelabelType))
1799  {
1800  /* We don't really expect this case, but may as well cope */
1801  node = (Node *) ((RelabelType *) node)->arg;
1802  }
1803  else
1804  break;
1805  }
1806  return node;
1807 }
1808 
1809 /*
1810  * scalararraysel - Selectivity of ScalarArrayOpExpr Node.
1811  */
1812 Selectivity
1813 scalararraysel(PlannerInfo *root,
1814  ScalarArrayOpExpr *clause,
1815  bool is_join_clause,
1816  int varRelid,
1817  JoinType jointype,
1818  SpecialJoinInfo *sjinfo)
1819 {
1820  Oid operator = clause->opno;
1821  bool useOr = clause->useOr;
1822  bool isEquality = false;
1823  bool isInequality = false;
1824  Node *leftop;
1825  Node *rightop;
1826  Oid nominal_element_type;
1827  Oid nominal_element_collation;
1828  TypeCacheEntry *typentry;
1829  RegProcedure oprsel;
1830  FmgrInfo oprselproc;
1831  Selectivity s1;
1832  Selectivity s1disjoint;
1833 
1834  /* First, deconstruct the expression */
1835  Assert(list_length(clause->args) == 2);
1836  leftop = (Node *) linitial(clause->args);
1837  rightop = (Node *) lsecond(clause->args);
1838 
1839  /* aggressively reduce both sides to constants */
1840  leftop = estimate_expression_value(root, leftop);
1841  rightop = estimate_expression_value(root, rightop);
1842 
1843  /* get nominal (after relabeling) element type of rightop */
1844  nominal_element_type = get_base_element_type(exprType(rightop));
1845  if (!OidIsValid(nominal_element_type))
1846  return (Selectivity) 0.5; /* probably shouldn't happen */
1847  /* get nominal collation, too, for generating constants */
1848  nominal_element_collation = exprCollation(rightop);
1849 
1850  /* look through any binary-compatible relabeling of rightop */
1851  rightop = strip_array_coercion(rightop);
1852 
1853  /*
1854  * Detect whether the operator is the default equality or inequality
1855  * operator of the array element type.
1856  */
1857  typentry = lookup_type_cache(nominal_element_type, TYPECACHE_EQ_OPR);
1858  if (OidIsValid(typentry->eq_opr))
1859  {
1860  if (operator == typentry->eq_opr)
1861  isEquality = true;
1862  else if (get_negator(operator) == typentry->eq_opr)
1863  isInequality = true;
1864  }
1865 
1866  /*
1867  * If it is equality or inequality, we might be able to estimate this as a
1868  * form of array containment; for instance "const = ANY(column)" can be
1869  * treated as "ARRAY[const] <@ column". scalararraysel_containment tries
1870  * that, and returns the selectivity estimate if successful, or -1 if not.
1871  */
1872  if ((isEquality || isInequality) && !is_join_clause)
1873  {
1874  s1 = scalararraysel_containment(root, leftop, rightop,
1875  nominal_element_type,
1876  isEquality, useOr, varRelid);
1877  if (s1 >= 0.0)
1878  return s1;
1879  }
1880 
1881  /*
1882  * Look up the underlying operator's selectivity estimator. Punt if it
1883  * hasn't got one.
1884  */
1885  if (is_join_clause)
1886  oprsel = get_oprjoin(operator);
1887  else
1888  oprsel = get_oprrest(operator);
1889  if (!oprsel)
1890  return (Selectivity) 0.5;
1891  fmgr_info(oprsel, &oprselproc);
1892 
1893  /*
1894  * In the array-containment check above, we must only believe that an
1895  * operator is equality or inequality if it is the default btree equality
1896  * operator (or its negator) for the element type, since those are the
1897  * operators that array containment will use. But in what follows, we can
1898  * be a little laxer, and also believe that any operators using eqsel() or
1899  * neqsel() as selectivity estimator act like equality or inequality.
1900  */
1901  if (oprsel == F_EQSEL || oprsel == F_EQJOINSEL)
1902  isEquality = true;
1903  else if (oprsel == F_NEQSEL || oprsel == F_NEQJOINSEL)
1904  isInequality = true;
1905 
1906  /*
1907  * We consider three cases:
1908  *
1909  * 1. rightop is an Array constant: deconstruct the array, apply the
1910  * operator's selectivity function for each array element, and merge the
1911  * results in the same way that clausesel.c does for AND/OR combinations.
1912  *
1913  * 2. rightop is an ARRAY[] construct: apply the operator's selectivity
1914  * function for each element of the ARRAY[] construct, and merge.
1915  *
1916  * 3. otherwise, make a guess ...
1917  */
1918  if (rightop && IsA(rightop, Const))
1919  {
1920  Datum arraydatum = ((Const *) rightop)->constvalue;
1921  bool arrayisnull = ((Const *) rightop)->constisnull;
1922  ArrayType *arrayval;
1923  int16 elmlen;
1924  bool elmbyval;
1925  char elmalign;
1926  int num_elems;
1927  Datum *elem_values;
1928  bool *elem_nulls;
1929  int i;
1930 
1931  if (arrayisnull) /* qual can't succeed if null array */
1932  return (Selectivity) 0.0;
1933  arrayval = DatumGetArrayTypeP(arraydatum);
1934  get_typlenbyvalalign(ARR_ELEMTYPE(arrayval),
1935  &elmlen, &elmbyval, &elmalign);
1936  deconstruct_array(arrayval,
1937  ARR_ELEMTYPE(arrayval),
1938  elmlen, elmbyval, elmalign,
1939  &elem_values, &elem_nulls, &num_elems);
1940 
1941  /*
1942  * For generic operators, we assume the probability of success is
1943  * independent for each array element. But for "= ANY" or "<> ALL",
1944  * if the array elements are distinct (which'd typically be the case)
1945  * then the probabilities are disjoint, and we should just sum them.
1946  *
1947  * If we were being really tense we would try to confirm that the
1948  * elements are all distinct, but that would be expensive and it
1949  * doesn't seem to be worth the cycles; it would amount to penalizing
1950  * well-written queries in favor of poorly-written ones. However, we
1951  * do protect ourselves a little bit by checking whether the
1952  * disjointness assumption leads to an impossible (out of range)
1953  * probability; if so, we fall back to the normal calculation.
1954  */
1955  s1 = s1disjoint = (useOr ? 0.0 : 1.0);
1956 
1957  for (i = 0; i < num_elems; i++)
1958  {
1959  List *args;
1960  Selectivity s2;
1961 
1962  args = list_make2(leftop,
1963  makeConst(nominal_element_type,
1964  -1,
1965  nominal_element_collation,
1966  elmlen,
1967  elem_values[i],
1968  elem_nulls[i],
1969  elmbyval));
1970  if (is_join_clause)
1971  s2 = DatumGetFloat8(FunctionCall5Coll(&oprselproc,
1972  clause->inputcollid,
1973  PointerGetDatum(root),
1974  ObjectIdGetDatum(operator),
1975  PointerGetDatum(args),
1976  Int16GetDatum(jointype),
1977  PointerGetDatum(sjinfo)));
1978  else
1979  s2 = DatumGetFloat8(FunctionCall4Coll(&oprselproc,
1980  clause->inputcollid,
1981  PointerGetDatum(root),
1982  ObjectIdGetDatum(operator),
1983  PointerGetDatum(args),
1984  Int32GetDatum(varRelid)));
1985 
1986  if (useOr)
1987  {
1988  s1 = s1 + s2 - s1 * s2;
1989  if (isEquality)
1990  s1disjoint += s2;
1991  }
1992  else
1993  {
1994  s1 = s1 * s2;
1995  if (isInequality)
1996  s1disjoint += s2 - 1.0;
1997  }
1998  }
1999 
2000  /* accept disjoint-probability estimate if in range */
2001  if ((useOr ? isEquality : isInequality) &&
2002  s1disjoint >= 0.0 && s1disjoint <= 1.0)
2003  s1 = s1disjoint;
2004  }
2005  else if (rightop && IsA(rightop, ArrayExpr) &&
2006  !((ArrayExpr *) rightop)->multidims)
2007  {
2008  ArrayExpr *arrayexpr = (ArrayExpr *) rightop;
2009  int16 elmlen;
2010  bool elmbyval;
2011  ListCell *l;
2012 
2013  get_typlenbyval(arrayexpr->element_typeid,
2014  &elmlen, &elmbyval);
2015 
2016  /*
2017  * We use the assumption of disjoint probabilities here too, although
2018  * the odds of equal array elements are rather higher if the elements
2019  * are not all constants (which they won't be, else constant folding
2020  * would have reduced the ArrayExpr to a Const). In this path it's
2021  * critical to have the sanity check on the s1disjoint estimate.
2022  */
2023  s1 = s1disjoint = (useOr ? 0.0 : 1.0);
2024 
2025  foreach(l, arrayexpr->elements)
2026  {
2027  Node *elem = (Node *) lfirst(l);
2028  List *args;
2029  Selectivity s2;
2030 
2031  /*
2032  * Theoretically, if elem isn't of nominal_element_type we should
2033  * insert a RelabelType, but it seems unlikely that any operator
2034  * estimation function would really care ...
2035  */
2036  args = list_make2(leftop, elem);
2037  if (is_join_clause)
2038  s2 = DatumGetFloat8(FunctionCall5Coll(&oprselproc,
2039  clause->inputcollid,
2040  PointerGetDatum(root),
2041  ObjectIdGetDatum(operator),
2042  PointerGetDatum(args),
2043  Int16GetDatum(jointype),
2044  PointerGetDatum(sjinfo)));
2045  else
2046  s2 = DatumGetFloat8(FunctionCall4Coll(&oprselproc,
2047  clause->inputcollid,
2048  PointerGetDatum(root),
2049  ObjectIdGetDatum(operator),
2050  PointerGetDatum(args),
2051  Int32GetDatum(varRelid)));
2052 
2053  if (useOr)
2054  {
2055  s1 = s1 + s2 - s1 * s2;
2056  if (isEquality)
2057  s1disjoint += s2;
2058  }
2059  else
2060  {
2061  s1 = s1 * s2;
2062  if (isInequality)
2063  s1disjoint += s2 - 1.0;
2064  }
2065  }
2066 
2067  /* accept disjoint-probability estimate if in range */
2068  if ((useOr ? isEquality : isInequality) &&
2069  s1disjoint >= 0.0 && s1disjoint <= 1.0)
2070  s1 = s1disjoint;
2071  }
2072  else
2073  {
2074  CaseTestExpr *dummyexpr;
2075  List *args;
2076  Selectivity s2;
2077  int i;
2078 
2079  /*
2080  * We need a dummy rightop to pass to the operator selectivity
2081  * routine. It can be pretty much anything that doesn't look like a
2082  * constant; CaseTestExpr is a convenient choice.
2083  */
2084  dummyexpr = makeNode(CaseTestExpr);
2085  dummyexpr->typeId = nominal_element_type;
2086  dummyexpr->typeMod = -1;
2087  dummyexpr->collation = clause->inputcollid;
2088  args = list_make2(leftop, dummyexpr);
2089  if (is_join_clause)
2090  s2 = DatumGetFloat8(FunctionCall5Coll(&oprselproc,
2091  clause->inputcollid,
2092  PointerGetDatum(root),
2093  ObjectIdGetDatum(operator),
2094  PointerGetDatum(args),
2095  Int16GetDatum(jointype),
2096  PointerGetDatum(sjinfo)));
2097  else
2098  s2 = DatumGetFloat8(FunctionCall4Coll(&oprselproc,
2099  clause->inputcollid,
2100  PointerGetDatum(root),
2101  ObjectIdGetDatum(operator),
2102  PointerGetDatum(args),
2103  Int32GetDatum(varRelid)));
2104  s1 = useOr ? 0.0 : 1.0;
2105 
2106  /*
2107  * Arbitrarily assume 10 elements in the eventual array value (see
2108  * also estimate_array_length). We don't risk an assumption of
2109  * disjoint probabilities here.
2110  */
2111  for (i = 0; i < 10; i++)
2112  {
2113  if (useOr)
2114  s1 = s1 + s2 - s1 * s2;
2115  else
2116  s1 = s1 * s2;
2117  }
2118  }
2119 
2120  /* result should be in range, but make sure... */
2121  CLAMP_PROBABILITY(s1);
2122 
2123  return s1;
2124 }
2125 
2126 /*
2127  * Estimate number of elements in the array yielded by an expression.
2128  *
2129  * It's important that this agree with scalararraysel.
2130  */
2131 int
2132 estimate_array_length(Node *arrayexpr)
2133 {
2134  /* look through any binary-compatible relabeling of arrayexpr */
2135  arrayexpr = strip_array_coercion(arrayexpr);
2136 
2137  if (arrayexpr && IsA(arrayexpr, Const))
2138  {
2139  Datum arraydatum = ((Const *) arrayexpr)->constvalue;
2140  bool arrayisnull = ((Const *) arrayexpr)->constisnull;
2141  ArrayType *arrayval;
2142 
2143  if (arrayisnull)
2144  return 0;
2145  arrayval = DatumGetArrayTypeP(arraydatum);
2146  return ArrayGetNItems(ARR_NDIM(arrayval), ARR_DIMS(arrayval));
2147  }
2148  else if (arrayexpr && IsA(arrayexpr, ArrayExpr) &&
2149  !((ArrayExpr *) arrayexpr)->multidims)
2150  {
2151  return list_length(((ArrayExpr *) arrayexpr)->elements);
2152  }
2153  else
2154  {
2155  /* default guess --- see also scalararraysel */
2156  return 10;
2157  }
2158 }
2159 
2160 /*
2161  * rowcomparesel - Selectivity of RowCompareExpr Node.
2162  *
2163  * We estimate RowCompare selectivity by considering just the first (high
2164  * order) columns, which makes it equivalent to an ordinary OpExpr. While
2165  * this estimate could be refined by considering additional columns, it
2166  * seems unlikely that we could do a lot better without multi-column
2167  * statistics.
2168  */
2169 Selectivity
2170 rowcomparesel(PlannerInfo *root,
2171  RowCompareExpr *clause,
2172  int varRelid, JoinType jointype, SpecialJoinInfo *sjinfo)
2173 {
2174  Selectivity s1;
2175  Oid opno = linitial_oid(clause->opnos);
2176  Oid inputcollid = linitial_oid(clause->inputcollids);
2177  List *opargs;
2178  bool is_join_clause;
2179 
2180  /* Build equivalent arg list for single operator */
2181  opargs = list_make2(linitial(clause->largs), linitial(clause->rargs));
2182 
2183  /*
2184  * Decide if it's a join clause. This should match clausesel.c's
2185  * treat_as_join_clause(), except that we intentionally consider only the
2186  * leading columns and not the rest of the clause.
2187  */
2188  if (varRelid != 0)
2189  {
2190  /*
2191  * Caller is forcing restriction mode (eg, because we are examining an
2192  * inner indexscan qual).
2193  */
2194  is_join_clause = false;
2195  }
2196  else if (sjinfo == NULL)
2197  {
2198  /*
2199  * It must be a restriction clause, since it's being evaluated at a
2200  * scan node.
2201  */
2202  is_join_clause = false;
2203  }
2204  else
2205  {
2206  /*
2207  * Otherwise, it's a join if there's more than one relation used.
2208  */
2209  is_join_clause = (NumRelids(root, (Node *) opargs) > 1);
2210  }
2211 
2212  if (is_join_clause)
2213  {
2214  /* Estimate selectivity for a join clause. */
2215  s1 = join_selectivity(root, opno,
2216  opargs,
2217  inputcollid,
2218  jointype,
2219  sjinfo);
2220  }
2221  else
2222  {
2223  /* Estimate selectivity for a restriction clause. */
2224  s1 = restriction_selectivity(root, opno,
2225  opargs,
2226  inputcollid,
2227  varRelid);
2228  }
2229 
2230  return s1;
2231 }
2232 
2233 /*
2234  * eqjoinsel - Join selectivity of "="
2235  */
2236 Datum
2237 eqjoinsel(PG_FUNCTION_ARGS)
2238 {
2239  PlannerInfo *root = (PlannerInfo *) PG_GETARG_POINTER(0);
2240  Oid operator = PG_GETARG_OID(1);
2241  List *args = (List *) PG_GETARG_POINTER(2);
2242 
2243 #ifdef NOT_USED
2244  JoinType jointype = (JoinType) PG_GETARG_INT16(3);
2245 #endif
2246  SpecialJoinInfo *sjinfo = (SpecialJoinInfo *) PG_GETARG_POINTER(4);
2247  Oid collation = PG_GET_COLLATION();
2248  double selec;
2249  double selec_inner;
2250  VariableStatData vardata1;
2251  VariableStatData vardata2;
2252  double nd1;
2253  double nd2;
2254  bool isdefault1;
2255  bool isdefault2;
2256  Oid opfuncoid;
2257  AttStatsSlot sslot1;
2258  AttStatsSlot sslot2;
2259  Form_pg_statistic stats1 = NULL;
2260  Form_pg_statistic stats2 = NULL;
2261  bool have_mcvs1 = false;
2262  bool have_mcvs2 = false;
2263  bool join_is_reversed;
2264  RelOptInfo *inner_rel;
2265 
2266  get_join_variables(root, args, sjinfo,
2267  &vardata1, &vardata2, &join_is_reversed);
2268 
2269  nd1 = get_variable_numdistinct(&vardata1, &isdefault1);
2270  nd2 = get_variable_numdistinct(&vardata2, &isdefault2);
2271 
2272  opfuncoid = get_opcode(operator);
2273 
2274  memset(&sslot1, 0, sizeof(sslot1));
2275  memset(&sslot2, 0, sizeof(sslot2));
2276 
2277  if (HeapTupleIsValid(vardata1.statsTuple))
2278  {
2279  /* note we allow use of nullfrac regardless of security check */
2280  stats1 = (Form_pg_statistic) GETSTRUCT(vardata1.statsTuple);
2281  if (statistic_proc_security_check(&vardata1, opfuncoid))
2282  have_mcvs1 = get_attstatsslot(&sslot1, vardata1.statsTuple,
2283  STATISTIC_KIND_MCV, InvalidOid,
2284  ATTSTATSSLOT_VALUES | ATTSTATSSLOT_NUMBERS);
2285  }
2286 
2287  if (HeapTupleIsValid(vardata2.statsTuple))
2288  {
2289  /* note we allow use of nullfrac regardless of security check */
2290  stats2 = (Form_pg_statistic) GETSTRUCT(vardata2.statsTuple);
2291  if (statistic_proc_security_check(&vardata2, opfuncoid))
2292  have_mcvs2 = get_attstatsslot(&sslot2, vardata2.statsTuple,
2293  STATISTIC_KIND_MCV, InvalidOid,
2294  ATTSTATSSLOT_VALUES | ATTSTATSSLOT_NUMBERS);
2295  }
2296 
2297  /* We need to compute the inner-join selectivity in all cases */
2298  selec_inner = eqjoinsel_inner(opfuncoid, collation,
2299  &vardata1, &vardata2,
2300  nd1, nd2,
2301  isdefault1, isdefault2,
2302  &sslot1, &sslot2,
2303  stats1, stats2,
2304  have_mcvs1, have_mcvs2);
2305 
2306  switch (sjinfo->jointype)
2307  {
2308  case JOIN_INNER:
2309  case JOIN_LEFT:
2310  case JOIN_FULL:
2311  selec = selec_inner;
2312  break;
2313  case JOIN_SEMI:
2314  case JOIN_ANTI:
2315 
2316  /*
2317  * Look up the join's inner relation. min_righthand is sufficient
2318  * information because neither SEMI nor ANTI joins permit any
2319  * reassociation into or out of their RHS, so the righthand will
2320  * always be exactly that set of rels.
2321  */
2322  inner_rel = find_join_input_rel(root, sjinfo->min_righthand);
2323 
2324  if (!join_is_reversed)
2325  selec = eqjoinsel_semi(opfuncoid, collation,
2326  &vardata1, &vardata2,
2327  nd1, nd2,
2328  isdefault1, isdefault2,
2329  &sslot1, &sslot2,
2330  stats1, stats2,
2331  have_mcvs1, have_mcvs2,
2332  inner_rel);
2333  else
2334  {
2335  Oid commop = get_commutator(operator);
2336  Oid commopfuncoid = OidIsValid(commop) ? get_opcode(commop) : InvalidOid;
2337 
2338  selec = eqjoinsel_semi(commopfuncoid, collation,
2339  &vardata2, &vardata1,
2340  nd2, nd1,
2341  isdefault2, isdefault1,
2342  &sslot2, &sslot1,
2343  stats2, stats1,
2344  have_mcvs2, have_mcvs1,
2345  inner_rel);
2346  }
2347 
2348  /*
2349  * We should never estimate the output of a semijoin to be more
2350  * rows than we estimate for an inner join with the same input
2351  * rels and join condition; it's obviously impossible for that to
2352  * happen. The former estimate is N1 * Ssemi while the latter is
2353  * N1 * N2 * Sinner, so we may clamp Ssemi <= N2 * Sinner. Doing
2354  * this is worthwhile because of the shakier estimation rules we
2355  * use in eqjoinsel_semi, particularly in cases where it has to
2356  * punt entirely.
2357  */
2358  selec = Min(selec, inner_rel->rows * selec_inner);
2359  break;
2360  default:
2361  /* other values not expected here */
2362  elog(ERROR, "unrecognized join type: %d",
2363  (int) sjinfo->jointype);
2364  selec = 0; /* keep compiler quiet */
2365  break;
2366  }
2367 
2368  free_attstatsslot(&sslot1);
2369  free_attstatsslot(&sslot2);
2370 
2371  ReleaseVariableStats(vardata1);
2372  ReleaseVariableStats(vardata2);
2373 
2374  CLAMP_PROBABILITY(selec);
2375 
2376  PG_RETURN_FLOAT8((float8) selec);
2377 }
2378 
2379 /*
2380  * eqjoinsel_inner --- eqjoinsel for normal inner join
2381  *
2382  * We also use this for LEFT/FULL outer joins; it's not presently clear
2383  * that it's worth trying to distinguish them here.
2384  */
2385 static double
2386 eqjoinsel_inner(Oid opfuncoid, Oid collation,
2387  VariableStatData *vardata1, VariableStatData *vardata2,
2388  double nd1, double nd2,
2389  bool isdefault1, bool isdefault2,
2390  AttStatsSlot *sslot1, AttStatsSlot *sslot2,
2391  Form_pg_statistic stats1, Form_pg_statistic stats2,
2392  bool have_mcvs1, bool have_mcvs2)
2393 {
2394  double selec;
2395 
2396  if (have_mcvs1 && have_mcvs2)
2397  {
2398  /*
2399  * We have most-common-value lists for both relations. Run through
2400  * the lists to see which MCVs actually join to each other with the
2401  * given operator. This allows us to determine the exact join
2402  * selectivity for the portion of the relations represented by the MCV
2403  * lists. We still have to estimate for the remaining population, but
2404  * in a skewed distribution this gives us a big leg up in accuracy.
2405  * For motivation see the analysis in Y. Ioannidis and S.
2406  * Christodoulakis, "On the propagation of errors in the size of join
2407  * results", Technical Report 1018, Computer Science Dept., University
2408  * of Wisconsin, Madison, March 1991 (available from ftp.cs.wisc.edu).
2409  */
2410  LOCAL_FCINFO(fcinfo, 2);
2411  FmgrInfo eqproc;
2412  bool *hasmatch1;
2413  bool *hasmatch2;
2414  double nullfrac1 = stats1->stanullfrac;
2415  double nullfrac2 = stats2->stanullfrac;
2416  double matchprodfreq,
2417  matchfreq1,
2418  matchfreq2,
2419  unmatchfreq1,
2420  unmatchfreq2,
2421  otherfreq1,
2422  otherfreq2,
2423  totalsel1,
2424  totalsel2;
2425  int i,
2426  nmatches;
2427 
2428  fmgr_info(opfuncoid, &eqproc);
2429 
2430  /*
2431  * Save a few cycles by setting up the fcinfo struct just once. Using
2432  * FunctionCallInvoke directly also avoids failure if the eqproc
2433  * returns NULL, though really equality functions should never do
2434  * that.
2435  */
2436  InitFunctionCallInfoData(*fcinfo, &eqproc, 2, collation,
2437  NULL, NULL);
2438  fcinfo->args[0].isnull = false;
2439  fcinfo->args[1].isnull = false;
2440 
2441  hasmatch1 = (bool *) palloc0(sslot1->nvalues * sizeof(bool));
2442  hasmatch2 = (bool *) palloc0(sslot2->nvalues * sizeof(bool));
2443 
2444  /*
2445  * Note we assume that each MCV will match at most one member of the
2446  * other MCV list. If the operator isn't really equality, there could
2447  * be multiple matches --- but we don't look for them, both for speed
2448  * and because the math wouldn't add up...
2449  */
2450  matchprodfreq = 0.0;
2451  nmatches = 0;
2452  for (i = 0; i < sslot1->nvalues; i++)
2453  {
2454  int j;
2455 
2456  fcinfo->args[0].value = sslot1->values[i];
2457 
2458  for (j = 0; j < sslot2->nvalues; j++)
2459  {
2460  Datum fresult;
2461 
2462  if (hasmatch2[j])
2463  continue;
2464  fcinfo->args[1].value = sslot2->values[j];
2465  fcinfo->isnull = false;
2466  fresult = FunctionCallInvoke(fcinfo);
2467  if (!fcinfo->isnull && DatumGetBool(fresult))
2468  {
2469  hasmatch1[i] = hasmatch2[j] = true;
2470  matchprodfreq += sslot1->numbers[i] * sslot2->numbers[j];
2471  nmatches++;
2472  break;
2473  }
2474  }
2475  }
2476  CLAMP_PROBABILITY(matchprodfreq);
2477  /* Sum up frequencies of matched and unmatched MCVs */
2478  matchfreq1 = unmatchfreq1 = 0.0;
2479  for (i = 0; i < sslot1->nvalues; i++)
2480  {
2481  if (hasmatch1[i])
2482  matchfreq1 += sslot1->numbers[i];
2483  else
2484  unmatchfreq1 += sslot1->numbers[i];
2485  }
2486  CLAMP_PROBABILITY(matchfreq1);
2487  CLAMP_PROBABILITY(unmatchfreq1);
2488  matchfreq2 = unmatchfreq2 = 0.0;
2489  for (i = 0; i < sslot2->nvalues; i++)
2490  {
2491  if (hasmatch2[i])
2492  matchfreq2 += sslot2->numbers[i];
2493  else
2494  unmatchfreq2 += sslot2->numbers[i];
2495  }
2496  CLAMP_PROBABILITY(matchfreq2);
2497  CLAMP_PROBABILITY(unmatchfreq2);
2498  pfree(hasmatch1);
2499  pfree(hasmatch2);
2500 
2501  /*
2502  * Compute total frequency of non-null values that are not in the MCV
2503  * lists.
2504  */
2505  otherfreq1 = 1.0 - nullfrac1 - matchfreq1 - unmatchfreq1;
2506  otherfreq2 = 1.0 - nullfrac2 - matchfreq2 - unmatchfreq2;
2507  CLAMP_PROBABILITY(otherfreq1);
2508  CLAMP_PROBABILITY(otherfreq2);
2509 
2510  /*
2511  * We can estimate the total selectivity from the point of view of
2512  * relation 1 as: the known selectivity for matched MCVs, plus
2513  * unmatched MCVs that are assumed to match against random members of
2514  * relation 2's non-MCV population, plus non-MCV values that are
2515  * assumed to match against random members of relation 2's unmatched
2516  * MCVs plus non-MCV values.
2517  */
2518  totalsel1 = matchprodfreq;
2519  if (nd2 > sslot2->nvalues)
2520  totalsel1 += unmatchfreq1 * otherfreq2 / (nd2 - sslot2->nvalues);
2521  if (nd2 > nmatches)
2522  totalsel1 += otherfreq1 * (otherfreq2 + unmatchfreq2) /
2523  (nd2 - nmatches);
2524  /* Same estimate from the point of view of relation 2. */
2525  totalsel2 = matchprodfreq;
2526  if (nd1 > sslot1->nvalues)
2527  totalsel2 += unmatchfreq2 * otherfreq1 / (nd1 - sslot1->nvalues);
2528  if (nd1 > nmatches)
2529  totalsel2 += otherfreq2 * (otherfreq1 + unmatchfreq1) /
2530  (nd1 - nmatches);
2531 
2532  /*
2533  * Use the smaller of the two estimates. This can be justified in
2534  * essentially the same terms as given below for the no-stats case: to
2535  * a first approximation, we are estimating from the point of view of
2536  * the relation with smaller nd.
2537  */
2538  selec = (totalsel1 < totalsel2) ? totalsel1 : totalsel2;
2539  }
2540  else
2541  {
2542  /*
2543  * We do not have MCV lists for both sides. Estimate the join
2544  * selectivity as MIN(1/nd1,1/nd2)*(1-nullfrac1)*(1-nullfrac2). This
2545  * is plausible if we assume that the join operator is strict and the
2546  * non-null values are about equally distributed: a given non-null
2547  * tuple of rel1 will join to either zero or N2*(1-nullfrac2)/nd2 rows
2548  * of rel2, so total join rows are at most
2549  * N1*(1-nullfrac1)*N2*(1-nullfrac2)/nd2 giving a join selectivity of
2550  * not more than (1-nullfrac1)*(1-nullfrac2)/nd2. By the same logic it
2551  * is not more than (1-nullfrac1)*(1-nullfrac2)/nd1, so the expression
2552  * with MIN() is an upper bound. Using the MIN() means we estimate
2553  * from the point of view of the relation with smaller nd (since the
2554  * larger nd is determining the MIN). It is reasonable to assume that
2555  * most tuples in this rel will have join partners, so the bound is
2556  * probably reasonably tight and should be taken as-is.
2557  *
2558  * XXX Can we be smarter if we have an MCV list for just one side? It
2559  * seems that if we assume equal distribution for the other side, we
2560  * end up with the same answer anyway.
2561  */
2562  double nullfrac1 = stats1 ? stats1->stanullfrac : 0.0;
2563  double nullfrac2 = stats2 ? stats2->stanullfrac : 0.0;
2564 
2565  selec = (1.0 - nullfrac1) * (1.0 - nullfrac2);
2566  if (nd1 > nd2)
2567  selec /= nd1;
2568  else
2569  selec /= nd2;
2570  }
2571 
2572  return selec;
2573 }
2574 
2575 /*
2576  * eqjoinsel_semi --- eqjoinsel for semi join
2577  *
2578  * (Also used for anti join, which we are supposed to estimate the same way.)
2579  * Caller has ensured that vardata1 is the LHS variable.
2580  * Unlike eqjoinsel_inner, we have to cope with opfuncoid being InvalidOid.
2581  */
2582 static double
2583 eqjoinsel_semi(Oid opfuncoid, Oid collation,
2584  VariableStatData *vardata1, VariableStatData *vardata2,
2585  double nd1, double nd2,
2586  bool isdefault1, bool isdefault2,
2587  AttStatsSlot *sslot1, AttStatsSlot *sslot2,
2588  Form_pg_statistic stats1, Form_pg_statistic stats2,
2589  bool have_mcvs1, bool have_mcvs2,
2590  RelOptInfo *inner_rel)
2591 {
2592  double selec;
2593 
2594  /*
2595  * We clamp nd2 to be not more than what we estimate the inner relation's
2596  * size to be. This is intuitively somewhat reasonable since obviously
2597  * there can't be more than that many distinct values coming from the
2598  * inner rel. The reason for the asymmetry (ie, that we don't clamp nd1
2599  * likewise) is that this is the only pathway by which restriction clauses
2600  * applied to the inner rel will affect the join result size estimate,
2601  * since set_joinrel_size_estimates will multiply SEMI/ANTI selectivity by
2602  * only the outer rel's size. If we clamped nd1 we'd be double-counting
2603  * the selectivity of outer-rel restrictions.
2604  *
2605  * We can apply this clamping both with respect to the base relation from
2606  * which the join variable comes (if there is just one), and to the
2607  * immediate inner input relation of the current join.
2608  *
2609  * If we clamp, we can treat nd2 as being a non-default estimate; it's not
2610  * great, maybe, but it didn't come out of nowhere either. This is most
2611  * helpful when the inner relation is empty and consequently has no stats.
2612  */
2613  if (vardata2->rel)
2614  {
2615  if (nd2 >= vardata2->rel->rows)
2616  {
2617  nd2 = vardata2->rel->rows;
2618  isdefault2 = false;
2619  }
2620  }
2621  if (nd2 >= inner_rel->rows)
2622  {
2623  nd2 = inner_rel->rows;
2624  isdefault2 = false;
2625  }
2626 
2627  if (have_mcvs1 && have_mcvs2 && OidIsValid(opfuncoid))
2628  {
2629  /*
2630  * We have most-common-value lists for both relations. Run through
2631  * the lists to see which MCVs actually join to each other with the
2632  * given operator. This allows us to determine the exact join
2633  * selectivity for the portion of the relations represented by the MCV
2634  * lists. We still have to estimate for the remaining population, but
2635  * in a skewed distribution this gives us a big leg up in accuracy.
2636  */
2637  LOCAL_FCINFO(fcinfo, 2);
2638  FmgrInfo eqproc;
2639  bool *hasmatch1;
2640  bool *hasmatch2;
2641  double nullfrac1 = stats1->stanullfrac;
2642  double matchfreq1,
2643  uncertainfrac,
2644  uncertain;
2645  int i,
2646  nmatches,
2647  clamped_nvalues2;
2648 
2649  /*
2650  * The clamping above could have resulted in nd2 being less than
2651  * sslot2->nvalues; in which case, we assume that precisely the nd2
2652  * most common values in the relation will appear in the join input,
2653  * and so compare to only the first nd2 members of the MCV list. Of
2654  * course this is frequently wrong, but it's the best bet we can make.
2655  */
2656  clamped_nvalues2 = Min(sslot2->nvalues, nd2);
2657 
2658  fmgr_info(opfuncoid, &eqproc);
2659 
2660  /*
2661  * Save a few cycles by setting up the fcinfo struct just once. Using
2662  * FunctionCallInvoke directly also avoids failure if the eqproc
2663  * returns NULL, though really equality functions should never do
2664  * that.
2665  */
2666  InitFunctionCallInfoData(*fcinfo, &eqproc, 2, collation,
2667  NULL, NULL);
2668  fcinfo->args[0].isnull = false;
2669  fcinfo->args[1].isnull = false;
2670 
2671  hasmatch1 = (bool *) palloc0(sslot1->nvalues * sizeof(bool));
2672  hasmatch2 = (bool *) palloc0(clamped_nvalues2 * sizeof(bool));
2673 
2674  /*
2675  * Note we assume that each MCV will match at most one member of the
2676  * other MCV list. If the operator isn't really equality, there could
2677  * be multiple matches --- but we don't look for them, both for speed
2678  * and because the math wouldn't add up...
2679  */
2680  nmatches = 0;
2681  for (i = 0; i < sslot1->nvalues; i++)
2682  {
2683  int j;
2684 
2685  fcinfo->args[0].value = sslot1->values[i];
2686 
2687  for (j = 0; j < clamped_nvalues2; j++)
2688  {
2689  Datum fresult;
2690 
2691  if (hasmatch2[j])
2692  continue;
2693  fcinfo->args[1].value = sslot2->values[j];
2694  fcinfo->isnull = false;
2695  fresult = FunctionCallInvoke(fcinfo);
2696  if (!fcinfo->isnull && DatumGetBool(fresult))
2697  {
2698  hasmatch1[i] = hasmatch2[j] = true;
2699  nmatches++;
2700  break;
2701  }
2702  }
2703  }
2704  /* Sum up frequencies of matched MCVs */
2705  matchfreq1 = 0.0;
2706  for (i = 0; i < sslot1->nvalues; i++)
2707  {
2708  if (hasmatch1[i])
2709  matchfreq1 += sslot1->numbers[i];
2710  }
2711  CLAMP_PROBABILITY(matchfreq1);
2712  pfree(hasmatch1);
2713  pfree(hasmatch2);
2714 
2715  /*
2716  * Now we need to estimate the fraction of relation 1 that has at
2717  * least one join partner. We know for certain that the matched MCVs
2718  * do, so that gives us a lower bound, but we're really in the dark
2719  * about everything else. Our crude approach is: if nd1 <= nd2 then
2720  * assume all non-null rel1 rows have join partners, else assume for
2721  * the uncertain rows that a fraction nd2/nd1 have join partners. We
2722  * can discount the known-matched MCVs from the distinct-values counts
2723  * before doing the division.
2724  *
2725  * Crude as the above is, it's completely useless if we don't have
2726  * reliable ndistinct values for both sides. Hence, if either nd1 or
2727  * nd2 is default, punt and assume half of the uncertain rows have
2728  * join partners.
2729  */
2730  if (!isdefault1 && !isdefault2)
2731  {
2732  nd1 -= nmatches;
2733  nd2 -= nmatches;
2734  if (nd1 <= nd2 || nd2 < 0)
2735  uncertainfrac = 1.0;
2736  else
2737  uncertainfrac = nd2 / nd1;
2738  }
2739  else
2740  uncertainfrac = 0.5;
2741  uncertain = 1.0 - matchfreq1 - nullfrac1;
2742  CLAMP_PROBABILITY(uncertain);
2743  selec = matchfreq1 + uncertainfrac * uncertain;
2744  }
2745  else
2746  {
2747  /*
2748  * Without MCV lists for both sides, we can only use the heuristic
2749  * about nd1 vs nd2.
2750  */
2751  double nullfrac1 = stats1 ? stats1->stanullfrac : 0.0;
2752 
2753  if (!isdefault1 && !isdefault2)
2754  {
2755  if (nd1 <= nd2 || nd2 < 0)
2756  selec = 1.0 - nullfrac1;
2757  else
2758  selec = (nd2 / nd1) * (1.0 - nullfrac1);
2759  }
2760  else
2761  selec = 0.5 * (1.0 - nullfrac1);
2762  }
2763 
2764  return selec;
2765 }
2766 
2767 /*
2768  * neqjoinsel - Join selectivity of "!="
2769  */
2770 Datum
2771 neqjoinsel(PG_FUNCTION_ARGS)
2772 {
2773  PlannerInfo *root = (PlannerInfo *) PG_GETARG_POINTER(0);
2774  Oid operator = PG_GETARG_OID(1);
2775  List *args = (List *) PG_GETARG_POINTER(2);
2776  JoinType jointype = (JoinType) PG_GETARG_INT16(3);
2777  SpecialJoinInfo *sjinfo = (SpecialJoinInfo *) PG_GETARG_POINTER(4);
2778  Oid collation = PG_GET_COLLATION();
2779  float8 result;
2780 
2781  if (jointype == JOIN_SEMI || jointype == JOIN_ANTI)
2782  {
2783  /*
2784  * For semi-joins, if there is more than one distinct value in the RHS
2785  * relation then every non-null LHS row must find a row to join since
2786  * it can only be equal to one of them. We'll assume that there is
2787  * always more than one distinct RHS value for the sake of stability,
2788  * though in theory we could have special cases for empty RHS
2789  * (selectivity = 0) and single-distinct-value RHS (selectivity =
2790  * fraction of LHS that has the same value as the single RHS value).
2791  *
2792  * For anti-joins, if we use the same assumption that there is more
2793  * than one distinct key in the RHS relation, then every non-null LHS
2794  * row must be suppressed by the anti-join.
2795  *
2796  * So either way, the selectivity estimate should be 1 - nullfrac.
2797  */
2798  VariableStatData leftvar;
2799  VariableStatData rightvar;
2800  bool reversed;
2801  HeapTuple statsTuple;
2802  double nullfrac;
2803 
2804  get_join_variables(root, args, sjinfo, &leftvar, &rightvar, &reversed);
2805  statsTuple = reversed ? rightvar.statsTuple : leftvar.statsTuple;
2806  if (HeapTupleIsValid(statsTuple))
2807  nullfrac = ((Form_pg_statistic) GETSTRUCT(statsTuple))->stanullfrac;
2808  else
2809  nullfrac = 0.0;
2810  ReleaseVariableStats(leftvar);
2811  ReleaseVariableStats(rightvar);
2812 
2813  result = 1.0 - nullfrac;
2814  }
2815  else
2816  {
2817  /*
2818  * We want 1 - eqjoinsel() where the equality operator is the one
2819  * associated with this != operator, that is, its negator.
2820  */
2821  Oid eqop = get_negator(operator);
2822 
2823  if (eqop)
2824  {
2825  result =
2826  DatumGetFloat8(DirectFunctionCall5Coll(eqjoinsel,
2827  collation,
2828  PointerGetDatum(root),
2829  ObjectIdGetDatum(eqop),
2830  PointerGetDatum(args),
2831  Int16GetDatum(jointype),
2832  PointerGetDatum(sjinfo)));
2833  }
2834  else
2835  {
2836  /* Use default selectivity (should we raise an error instead?) */
2837  result = DEFAULT_EQ_SEL;
2838  }
2839  result = 1.0 - result;
2840  }
2841 
2842  PG_RETURN_FLOAT8(result);
2843 }
2844 
2845 /*
2846  * scalarltjoinsel - Join selectivity of "<" for scalars
2847  */
2848 Datum
2849 scalarltjoinsel(PG_FUNCTION_ARGS)
2850 {
2851  PG_RETURN_FLOAT8(DEFAULT_INEQ_SEL);
2852 }
2853 
2854 /*
2855  * scalarlejoinsel - Join selectivity of "<=" for scalars
2856  */
2857 Datum
2858 scalarlejoinsel(PG_FUNCTION_ARGS)
2859 {
2860  PG_RETURN_FLOAT8(DEFAULT_INEQ_SEL);
2861 }
2862 
2863 /*
2864  * scalargtjoinsel - Join selectivity of ">" for scalars
2865  */
2866 Datum
2867 scalargtjoinsel(PG_FUNCTION_ARGS)
2868 {
2869  PG_RETURN_FLOAT8(DEFAULT_INEQ_SEL);
2870 }
2871 
2872 /*
2873  * scalargejoinsel - Join selectivity of ">=" for scalars
2874  */
2875 Datum
2876 scalargejoinsel(PG_FUNCTION_ARGS)
2877 {
2878  PG_RETURN_FLOAT8(DEFAULT_INEQ_SEL);
2879 }
2880 
2881 
2882 /*
2883  * mergejoinscansel - Scan selectivity of merge join.
2884  *
2885  * A merge join will stop as soon as it exhausts either input stream.
2886  * Therefore, if we can estimate the ranges of both input variables,
2887  * we can estimate how much of the input will actually be read. This
2888  * can have a considerable impact on the cost when using indexscans.
2889  *
2890  * Also, we can estimate how much of each input has to be read before the
2891  * first join pair is found, which will affect the join's startup time.
2892  *
2893  * clause should be a clause already known to be mergejoinable. opfamily,
2894  * strategy, and nulls_first specify the sort ordering being used.
2895  *
2896  * The outputs are:
2897  * *leftstart is set to the fraction of the left-hand variable expected
2898  * to be scanned before the first join pair is found (0 to 1).
2899  * *leftend is set to the fraction of the left-hand variable expected
2900  * to be scanned before the join terminates (0 to 1).
2901  * *rightstart, *rightend similarly for the right-hand variable.
2902  */
2903 void
2904 mergejoinscansel(PlannerInfo *root, Node *clause,
2905  Oid opfamily, int strategy, bool nulls_first,
2906  Selectivity *leftstart, Selectivity *leftend,
2907  Selectivity *rightstart, Selectivity *rightend)
2908 {
2909  Node *left,
2910  *right;
2911  VariableStatData leftvar,
2912  rightvar;
2913  int op_strategy;
2914  Oid op_lefttype;
2915  Oid op_righttype;
2916  Oid opno,
2917  collation,
2918  lsortop,
2919  rsortop,
2920  lstatop,
2921  rstatop,
2922  ltop,
2923  leop,
2924  revltop,
2925  revleop;
2926  bool isgt;
2927  Datum leftmin,
2928  leftmax,
2929  rightmin,
2930  rightmax;
2931  double selec;
2932 
2933  /* Set default results if we can't figure anything out. */
2934  /* XXX should default "start" fraction be a bit more than 0? */
2935  *leftstart = *rightstart = 0.0;
2936  *leftend = *rightend = 1.0;
2937 
2938  /* Deconstruct the merge clause */
2939  if (!is_opclause(clause))
2940  return; /* shouldn't happen */
2941  opno = ((OpExpr *) clause)->opno;
2942  collation = ((OpExpr *) clause)->inputcollid;
2943  left = get_leftop((Expr *) clause);
2944  right = get_rightop((Expr *) clause);
2945  if (!right)
2946  return; /* shouldn't happen */
2947 
2948  /* Look for stats for the inputs */
2949  examine_variable(root, left, 0, &leftvar);
2950  examine_variable(root, right, 0, &rightvar);
2951 
2952  /* Extract the operator's declared left/right datatypes */
2953  get_op_opfamily_properties(opno, opfamily, false,
2954  &op_strategy,
2955  &op_lefttype,
2956  &op_righttype);
2957  Assert(op_strategy == BTEqualStrategyNumber);
2958 
2959  /*
2960  * Look up the various operators we need. If we don't find them all, it
2961  * probably means the opfamily is broken, but we just fail silently.
2962  *
2963  * Note: we expect that pg_statistic histograms will be sorted by the '<'
2964  * operator, regardless of which sort direction we are considering.
2965  */
2966  switch (strategy)
2967  {
2968  case BTLessStrategyNumber:
2969  isgt = false;
2970  if (op_lefttype == op_righttype)
2971  {
2972  /* easy case */
2973  ltop = get_opfamily_member(opfamily,
2974  op_lefttype, op_righttype,
2975  BTLessStrategyNumber);
2976  leop = get_opfamily_member(opfamily,
2977  op_lefttype, op_righttype,
2978  BTLessEqualStrategyNumber);
2979  lsortop = ltop;
2980  rsortop = ltop;
2981  lstatop = lsortop;
2982  rstatop = rsortop;
2983  revltop = ltop;
2984  revleop = leop;
2985  }
2986  else
2987  {
2988  ltop = get_opfamily_member(opfamily,
2989  op_lefttype, op_righttype,
2990  BTLessStrategyNumber);
2991  leop = get_opfamily_member(opfamily,
2992  op_lefttype, op_righttype,
2993  BTLessEqualStrategyNumber);
2994  lsortop = get_opfamily_member(opfamily,
2995  op_lefttype, op_lefttype,
2996  BTLessStrategyNumber);
2997  rsortop = get_opfamily_member(opfamily,
2998  op_righttype, op_righttype,
2999  BTLessStrategyNumber);
3000  lstatop = lsortop;
3001  rstatop = rsortop;
3002  revltop = get_opfamily_member(opfamily,
3003  op_righttype, op_lefttype,
3004  BTLessStrategyNumber);
3005  revleop = get_opfamily_member(opfamily,
3006  op_righttype, op_lefttype,
3007  BTLessEqualStrategyNumber);
3008  }
3009  break;
3010  case BTGreaterStrategyNumber:
3011  /* descending-order case */
3012  isgt = true;
3013  if (op_lefttype == op_righttype)
3014  {
3015  /* easy case */
3016  ltop = get_opfamily_member(opfamily,
3017  op_lefttype, op_righttype,
3018  BTGreaterStrategyNumber);
3019  leop = get_opfamily_member(opfamily,
3020  op_lefttype, op_righttype,
3021  BTGreaterEqualStrategyNumber);
3022  lsortop = ltop;
3023  rsortop = ltop;
3024  lstatop = get_opfamily_member(opfamily,
3025  op_lefttype, op_lefttype,
3026  BTLessStrategyNumber);
3027  rstatop = lstatop;
3028  revltop = ltop;
3029  revleop = leop;
3030  }
3031  else
3032  {
3033  ltop = get_opfamily_member(opfamily,
3034  op_lefttype, op_righttype,
3035  BTGreaterStrategyNumber);
3036  leop = get_opfamily_member(opfamily,
3037  op_lefttype, op_righttype,
3038  BTGreaterEqualStrategyNumber);
3039  lsortop = get_opfamily_member(opfamily,
3040  op_lefttype, op_lefttype,
3041  BTGreaterStrategyNumber);
3042  rsortop = get_opfamily_member(opfamily,
3043  op_righttype, op_righttype,
3044  BTGreaterStrategyNumber);
3045  lstatop = get_opfamily_member(opfamily,
3046  op_lefttype, op_lefttype,
3047  BTLessStrategyNumber);
3048  rstatop = get_opfamily_member(opfamily,
3049  op_righttype, op_righttype,
3050  BTLessStrategyNumber);
3051  revltop = get_opfamily_member(opfamily,
3052  op_righttype, op_lefttype,
3053  BTGreaterStrategyNumber);
3054  revleop = get_opfamily_member(opfamily,
3055  op_righttype, op_lefttype,
3056  BTGreaterEqualStrategyNumber);
3057  }
3058  break;
3059  default:
3060  goto fail; /* shouldn't get here */
3061  }
3062 
3063  if (!OidIsValid(lsortop) ||
3064  !OidIsValid(rsortop) ||
3065  !OidIsValid(lstatop) ||
3066  !OidIsValid(rstatop) ||
3067  !OidIsValid(ltop) ||
3068  !OidIsValid(leop) ||
3069  !OidIsValid(revltop) ||
3070  !OidIsValid(revleop))
3071  goto fail; /* insufficient info in catalogs */
3072 
3073  /* Try to get ranges of both inputs */
3074  if (!isgt)
3075  {
3076  if (!get_variable_range(root, &leftvar, lstatop, collation,
3077  &leftmin, &leftmax))
3078  goto fail; /* no range available from stats */
3079  if (!get_variable_range(root, &rightvar, rstatop, collation,
3080  &rightmin, &rightmax))
3081  goto fail; /* no range available from stats */
3082  }
3083  else
3084  {
3085  /* need to swap the max and min */
3086  if (!get_variable_range(root, &leftvar, lstatop, collation,
3087  &leftmax, &leftmin))
3088  goto fail; /* no range available from stats */
3089  if (!get_variable_range(root, &rightvar, rstatop, collation,
3090  &rightmax, &rightmin))
3091  goto fail; /* no range available from stats */
3092  }
3093 
3094  /*
3095  * Now, the fraction of the left variable that will be scanned is the
3096  * fraction that's <= the right-side maximum value. But only believe
3097  * non-default estimates, else stick with our 1.0.
3098  */
3099  selec = scalarineqsel(root, leop, isgt, true, collation, &leftvar,
3100  rightmax, op_righttype);
3101  if (selec != DEFAULT_INEQ_SEL)
3102  *leftend = selec;
3103 
3104  /* And similarly for the right variable. */
3105  selec = scalarineqsel(root, revleop, isgt, true, collation, &rightvar,
3106  leftmax, op_lefttype);
3107  if (selec != DEFAULT_INEQ_SEL)
3108  *rightend = selec;
3109 
3110  /*
3111  * Only one of the two "end" fractions can really be less than 1.0;
3112  * believe the smaller estimate and reset the other one to exactly 1.0. If
3113  * we get exactly equal estimates (as can easily happen with self-joins),
3114  * believe neither.
3115  */
3116  if (*leftend > *rightend)
3117  *leftend = 1.0;
3118  else if (*leftend < *rightend)
3119  *rightend = 1.0;
3120  else
3121  *leftend = *rightend = 1.0;
3122 
3123  /*
3124  * Also, the fraction of the left variable that will be scanned before the
3125  * first join pair is found is the fraction that's < the right-side
3126  * minimum value. But only believe non-default estimates, else stick with
3127  * our own default.
3128  */
3129  selec = scalarineqsel(root, ltop, isgt, false, collation, &leftvar,
3130  rightmin, op_righttype);
3131  if (selec != DEFAULT_INEQ_SEL)
3132  *leftstart = selec;
3133 
3134  /* And similarly for the right variable. */
3135  selec = scalarineqsel(root, revltop, isgt, false, collation, &rightvar,
3136  leftmin, op_lefttype);
3137  if (selec != DEFAULT_INEQ_SEL)
3138  *rightstart = selec;
3139 
3140  /*
3141  * Only one of the two "start" fractions can really be more than zero;
3142  * believe the larger estimate and reset the other one to exactly 0.0. If
3143  * we get exactly equal estimates (as can easily happen with self-joins),
3144  * believe neither.
3145  */
3146  if (*leftstart < *rightstart)
3147  *leftstart = 0.0;
3148  else if (*leftstart > *rightstart)
3149  *rightstart = 0.0;
3150  else
3151  *leftstart = *rightstart = 0.0;
3152 
3153  /*
3154  * If the sort order is nulls-first, we're going to have to skip over any
3155  * nulls too. These would not have been counted by scalarineqsel, and we
3156  * can safely add in this fraction regardless of whether we believe
3157  * scalarineqsel's results or not. But be sure to clamp the sum to 1.0!
3158  */
3159  if (nulls_first)
3160  {
3161  Form_pg_statistic stats;
3162 
3163  if (HeapTupleIsValid(leftvar.statsTuple))
3164  {
3165  stats = (Form_pg_statistic) GETSTRUCT(leftvar.statsTuple);
3166  *leftstart += stats->stanullfrac;
3167  CLAMP_PROBABILITY(*leftstart);
3168  *leftend += stats->stanullfrac;
3169  CLAMP_PROBABILITY(*leftend);
3170  }
3171  if (HeapTupleIsValid(rightvar.statsTuple))
3172  {
3173  stats = (Form_pg_statistic) GETSTRUCT(rightvar.statsTuple);
3174  *rightstart += stats->stanullfrac;
3175  CLAMP_PROBABILITY(*rightstart);
3176  *rightend += stats->stanullfrac;
3177  CLAMP_PROBABILITY(*rightend);
3178  }
3179  }
3180 
3181  /* Disbelieve start >= end, just in case that can happen */
3182  if (*leftstart >= *leftend)
3183  {
3184  *leftstart = 0.0;
3185  *leftend = 1.0;
3186  }
3187  if (*rightstart >= *rightend)
3188  {
3189  *rightstart = 0.0;
3190  *rightend = 1.0;
3191  }
3192 
3193 fail:
3194  ReleaseVariableStats(leftvar);
3195  ReleaseVariableStats(rightvar);
3196 }
3197 
3198 
3199 /*
3200  * matchingsel -- generic matching-operator selectivity support
3201  *
3202  * Use these for any operators that (a) are on data types for which we collect
3203  * standard statistics, and (b) have behavior for which the default estimate
3204  * (twice DEFAULT_EQ_SEL) is sane. Typically that is good for match-like
3205  * operators.
3206  */
3207 
3208 Datum
3209 matchingsel(PG_FUNCTION_ARGS)
3210 {
3211  PlannerInfo *root = (PlannerInfo *) PG_GETARG_POINTER(0);
3212  Oid operator = PG_GETARG_OID(1);
3213  List *args = (List *) PG_GETARG_POINTER(2);
3214  int varRelid = PG_GETARG_INT32(3);
3215  Oid collation = PG_GET_COLLATION();
3216  double selec;
3217 
3218  /* Use generic restriction selectivity logic. */
3219  selec = generic_restriction_selectivity(root, operator, collation,
3220  args, varRelid,
3221  DEFAULT_MATCHING_SEL);
3222 
3223  PG_RETURN_FLOAT8((float8) selec);
3224 }
3225 
3226 Datum
3227 matchingjoinsel(PG_FUNCTION_ARGS)
3228 {
3229  /* Just punt, for the moment. */
3230  PG_RETURN_FLOAT8(DEFAULT_MATCHING_SEL);
3231 }
3232 
3233 
3234 /*
3235  * Helper routine for estimate_num_groups: add an item to a list of
3236  * GroupVarInfos, but only if it's not known equal to any of the existing
3237  * entries.
3238  */
3239 typedef struct
3240 {
3241  Node *var; /* might be an expression, not just a Var */
3242  RelOptInfo *rel; /* relation it belongs to */
3243  double ndistinct; /* # distinct values */
3244  bool isdefault; /* true if DEFAULT_NUM_DISTINCT was used */
3245 } GroupVarInfo;
3246 
3247 static List *
3248 add_unique_group_var(PlannerInfo *root, List *varinfos,
3249  Node *var, VariableStatData *vardata)
3250 {
3251  GroupVarInfo *varinfo;
3252  double ndistinct;
3253  bool isdefault;
3254  ListCell *lc;
3255 
3256  ndistinct = get_variable_numdistinct(vardata, &isdefault);
3257 
3258  foreach(lc, varinfos)
3259  {
3260  varinfo = (GroupVarInfo *) lfirst(lc);
3261 
3262  /* Drop exact duplicates */
3263  if (equal(var, varinfo->var))
3264  return varinfos;
3265 
3266  /*
3267  * Drop known-equal vars, but only if they belong to different
3268  * relations (see comments for estimate_num_groups)
3269  */
3270  if (vardata->rel != varinfo->rel &&
3271  exprs_known_equal(root, var, varinfo->var))
3272  {
3273  if (varinfo->ndistinct <= ndistinct)
3274  {
3275  /* Keep older item, forget new one */
3276  return varinfos;
3277  }
3278  else
3279  {
3280  /* Delete the older item */
3281  varinfos = foreach_delete_current(varinfos, lc);
3282  }
3283  }
3284  }
3285 
3286  varinfo = (GroupVarInfo *) palloc(sizeof(GroupVarInfo));
3287 
3288  varinfo->var = var;
3289  varinfo->rel = vardata->rel;
3290  varinfo->ndistinct = ndistinct;
3291  varinfo->isdefault = isdefault;
3292  varinfos = lappend(varinfos, varinfo);
3293  return varinfos;
3294 }
3295 
3296 /*
3297  * estimate_num_groups - Estimate number of groups in a grouped query
3298  *
3299  * Given a query having a GROUP BY clause, estimate how many groups there
3300  * will be --- ie, the number of distinct combinations of the GROUP BY
3301  * expressions.
3302  *
3303  * This routine is also used to estimate the number of rows emitted by
3304  * a DISTINCT filtering step; that is an isomorphic problem. (Note:
3305  * actually, we only use it for DISTINCT when there's no grouping or
3306  * aggregation ahead of the DISTINCT.)
3307  *
3308  * Inputs:
3309  * root - the query
3310  * groupExprs - list of expressions being grouped by
3311  * input_rows - number of rows estimated to arrive at the group/unique
3312  * filter step
3313  * pgset - NULL, or a List** pointing to a grouping set to filter the
3314  * groupExprs against
3315  *
3316  * Outputs:
3317  * estinfo - When passed as non-NULL, the function will set bits in the
3318  * "flags" field in order to provide callers with additional information
3319  * about the estimation. Currently, we only set the SELFLAG_USED_DEFAULT
3320  * bit if we used any default values in the estimation.
3321  *
3322  * Given the lack of any cross-correlation statistics in the system, it's
3323  * impossible to do anything really trustworthy with GROUP BY conditions
3324  * involving multiple Vars. We should however avoid assuming the worst
3325  * case (all possible cross-product terms actually appear as groups) since
3326  * very often the grouped-by Vars are highly correlated. Our current approach
3327  * is as follows:
3328  * 1. Expressions yielding boolean are assumed to contribute two groups,
3329  * independently of their content, and are ignored in the subsequent
3330  * steps. This is mainly because tests like "col IS NULL" break the
3331  * heuristic used in step 2 especially badly.
3332  * 2. Reduce the given expressions to a list of unique Vars used. For
3333  * example, GROUP BY a, a + b is treated the same as GROUP BY a, b.
3334  * It is clearly correct not to count the same Var more than once.
3335  * It is also reasonable to treat f(x) the same as x: f() cannot
3336  * increase the number of distinct values (unless it is volatile,
3337  * which we consider unlikely for grouping), but it probably won't
3338  * reduce the number of distinct values much either.
3339  * As a special case, if a GROUP BY expression can be matched to an
3340  * expressional index for which we have statistics, then we treat the
3341  * whole expression as though it were just a Var.
3342  * 3. If the list contains Vars of different relations that are known equal
3343  * due to equivalence classes, then drop all but one of the Vars from each
3344  * known-equal set, keeping the one with smallest estimated # of values
3345  * (since the extra values of the others can't appear in joined rows).
3346  * Note the reason we only consider Vars of different relations is that
3347  * if we considered ones of the same rel, we'd be double-counting the
3348  * restriction selectivity of the equality in the next step.
3349  * 4. For Vars within a single source rel, we multiply together the numbers
3350  * of values, clamp to the number of rows in the rel (divided by 10 if
3351  * more than one Var), and then multiply by a factor based on the
3352  * selectivity of the restriction clauses for that rel. When there's
3353  * more than one Var, the initial product is probably too high (it's the
3354  * worst case) but clamping to a fraction of the rel's rows seems to be a
3355  * helpful heuristic for not letting the estimate get out of hand. (The
3356  * factor of 10 is derived from pre-Postgres-7.4 practice.) The factor
3357  * we multiply by to adjust for the restriction selectivity assumes that
3358  * the restriction clauses are independent of the grouping, which may not
3359  * be a valid assumption, but it's hard to do better.
3360  * 5. If there are Vars from multiple rels, we repeat step 4 for each such
3361  * rel, and multiply the results together.
3362  * Note that rels not containing grouped Vars are ignored completely, as are
3363  * join clauses. Such rels cannot increase the number of groups, and we
3364  * assume such clauses do not reduce the number either (somewhat bogus,
3365  * but we don't have the info to do better).
3366  */
3367 double
3368 estimate_num_groups(PlannerInfo *root, List *groupExprs, double input_rows,
3369  List **pgset, EstimationInfo *estinfo)
3370 {
3371  List *varinfos = NIL;
3372  double srf_multiplier = 1.0;
3373  double numdistinct;
3374  ListCell *l;
3375  int i;
3376 
3377  /* Zero the estinfo output parameter, if non-NULL */
3378  if (estinfo != NULL)
3379  memset(estinfo, 0, sizeof(EstimationInfo));
3380 
3381  /*
3382  * We don't ever want to return an estimate of zero groups, as that tends
3383  * to lead to division-by-zero and other unpleasantness. The input_rows
3384  * estimate is usually already at least 1, but clamp it just in case it
3385  * isn't.
3386  */
3387  input_rows = clamp_row_est(input_rows);
3388 
3389  /*
3390  * If no grouping columns, there's exactly one group. (This can't happen
3391  * for normal cases with GROUP BY or DISTINCT, but it is possible for
3392  * corner cases with set operations.)
3393  */
3394  if (groupExprs == NIL || (pgset && list_length(*pgset) < 1))
3395  return 1.0;
3396 
3397  /*
3398  * Count groups derived from boolean grouping expressions. For other
3399  * expressions, find the unique Vars used, treating an expression as a Var
3400  * if we can find stats for it. For each one, record the statistical
3401  * estimate of number of distinct values (total in its table, without
3402  * regard for filtering).
3403  */
3404  numdistinct = 1.0;
3405 
3406  i = 0;
3407  foreach(l, groupExprs)
3408  {
3409  Node *groupexpr = (Node *) lfirst(l);
3410  double this_srf_multiplier;
3411  VariableStatData vardata;
3412  List *varshere;
3413  ListCell *l2;
3414 
3415  /* is expression in this grouping set? */
3416  if (pgset && !list_member_int(*pgset, i++))
3417  continue;
3418 
3419  /*
3420  * Set-returning functions in grouping columns are a bit problematic.
3421  * The code below will effectively ignore their SRF nature and come up
3422  * with a numdistinct estimate as though they were scalar functions.
3423  * We compensate by scaling up the end result by the largest SRF
3424  * rowcount estimate. (This will be an overestimate if the SRF
3425  * produces multiple copies of any output value, but it seems best to
3426  * assume the SRF's outputs are distinct. In any case, it's probably
3427  * pointless to worry too much about this without much better
3428  * estimates for SRF output rowcounts than we have today.)
3429  */
3430  this_srf_multiplier = expression_returns_set_rows(root, groupexpr);
3431  if (srf_multiplier < this_srf_multiplier)
3432  srf_multiplier = this_srf_multiplier;
3433 
3434  /* Short-circuit for expressions returning boolean */
3435  if (exprType(groupexpr) == BOOLOID)
3436  {
3437  numdistinct *= 2.0;
3438  continue;
3439  }
3440 
3441  /*
3442  * If examine_variable is able to deduce anything about the GROUP BY
3443  * expression, treat it as a single variable even if it's really more
3444  * complicated.
3445  *
3446  * XXX This has the consequence that if there's a statistics on the
3447  * expression, we don't split it into individual Vars. This affects
3448  * our selection of statistics in estimate_multivariate_ndistinct,
3449  * because it's probably better to use more accurate estimate for each
3450  * expression and treat them as independent, than to combine estimates
3451  * for the extracted variables when we don't know how that relates to
3452  * the expressions.
3453  */
3454  examine_variable(root, groupexpr, 0, &vardata);
3455  if (HeapTupleIsValid(vardata.statsTuple) || vardata.isunique)
3456  {
3457  varinfos = add_unique_group_var(root, varinfos,
3458  groupexpr, &vardata);
3459  ReleaseVariableStats(vardata);
3460  continue;
3461  }
3462  ReleaseVariableStats(vardata);
3463 
3464  /*
3465  * Else pull out the component Vars. Handle PlaceHolderVars by
3466  * recursing into their arguments (effectively assuming that the
3467  * PlaceHolderVar doesn't change the number of groups, which boils
3468  * down to ignoring the possible addition of nulls to the result set).
3469  */
3470  varshere = pull_var_clause(groupexpr,
3471  PVC_RECURSE_AGGREGATES |
3472  PVC_RECURSE_WINDOWFUNCS |
3473  PVC_RECURSE_PLACEHOLDERS);
3474 
3475  /*
3476  * If we find any variable-free GROUP BY item, then either it is a
3477  * constant (and we can ignore it) or it contains a volatile function;
3478  * in the latter case we punt and assume that each input row will
3479  * yield a distinct group.
3480  */
3481  if (varshere == NIL)
3482  {
3483  if (contain_volatile_functions(groupexpr))
3484  return input_rows;
3485  continue;
3486  }
3487 
3488  /*
3489  * Else add variables to varinfos list
3490  */
3491  foreach(l2, varshere)
3492  {
3493  Node *var = (Node *) lfirst(l2);
3494 
3495  examine_variable(root, var, 0, &vardata);
3496  varinfos = add_unique_group_var(root, varinfos, var, &vardata);
3497  ReleaseVariableStats(vardata);
3498  }
3499  }
3500 
3501  /*
3502  * If now no Vars, we must have an all-constant or all-boolean GROUP BY
3503  * list.
3504  */
3505  if (varinfos == NIL)
3506  {
3507  /* Apply SRF multiplier as we would do in the long path */
3508  numdistinct *= srf_multiplier;
3509  /* Round off */
3510  numdistinct = ceil(numdistinct);
3511  /* Guard against out-of-range answers */
3512  if (numdistinct > input_rows)
3513  numdistinct = input_rows;
3514  if (numdistinct < 1.0)
3515  numdistinct = 1.0;
3516  return numdistinct;
3517  }
3518 
3519  /*
3520  * Group Vars by relation and estimate total numdistinct.
3521  *
3522  * For each iteration of the outer loop, we process the frontmost Var in
3523  * varinfos, plus all other Vars in the same relation. We remove these
3524  * Vars from the newvarinfos list for the next iteration. This is the
3525  * easiest way to group Vars of same rel together.
3526  */
3527  do
3528  {
3529  GroupVarInfo *varinfo1 = (GroupVarInfo *) linitial(varinfos);
3530  RelOptInfo *rel = varinfo1->rel;
3531  double reldistinct = 1;
3532  double relmaxndistinct = reldistinct;
3533  int relvarcount = 0;
3534  List *newvarinfos = NIL;
3535  List *relvarinfos = NIL;
3536 
3537  /*
3538  * Split the list of varinfos in two - one for the current rel, one
3539  * for remaining Vars on other rels.
3540  */
3541  relvarinfos = lappend(relvarinfos, varinfo1);
3542  for_each_from(l, varinfos, 1)
3543  {
3544  GroupVarInfo *varinfo2 = (GroupVarInfo *) lfirst(l);
3545 
3546  if (varinfo2->rel == varinfo1->rel)
3547  {
3548  /* varinfos on current rel */
3549  relvarinfos = lappend(relvarinfos, varinfo2);
3550  }
3551  else
3552  {
3553  /* not time to process varinfo2 yet */
3554  newvarinfos = lappend(newvarinfos, varinfo2);
3555  }
3556  }
3557 
3558  /*
3559  * Get the numdistinct estimate for the Vars of this rel. We
3560  * iteratively search for multivariate n-distinct with maximum number
3561  * of vars; assuming that each var group is independent of the others,
3562  * we multiply them together. Any remaining relvarinfos after no more
3563  * multivariate matches are found are assumed independent too, so
3564  * their individual ndistinct estimates are multiplied also.
3565  *
3566  * While iterating, count how many separate numdistinct values we
3567  * apply. We apply a fudge factor below, but only if we multiplied
3568  * more than one such values.
3569  */
3570  while (relvarinfos)
3571  {
3572  double mvndistinct;
3573 
3574  if (estimate_multivariate_ndistinct(root, rel, &relvarinfos,
3575  &mvndistinct))
3576  {
3577  reldistinct *= mvndistinct;
3578  if (relmaxndistinct < mvndistinct)
3579  relmaxndistinct = mvndistinct;
3580  relvarcount++;
3581  }
3582  else
3583  {
3584  foreach(l, relvarinfos)
3585  {
3586  GroupVarInfo *varinfo2 = (GroupVarInfo *) lfirst(l);
3587 
3588  reldistinct *= varinfo2->ndistinct;
3589  if (relmaxndistinct < varinfo2->ndistinct)
3590  relmaxndistinct = varinfo2->ndistinct;
3591  relvarcount++;
3592 
3593  /*
3594  * When varinfo2's isdefault is set then we'd better set
3595  * the SELFLAG_USED_DEFAULT bit in the EstimationInfo.
3596  */
3597  if (estinfo != NULL && varinfo2->isdefault)
3598  estinfo->flags |= SELFLAG_USED_DEFAULT;
3599 
3600  }
3601 
3602  /* we're done with this relation */
3603  relvarinfos = NIL;
3604  }
3605  }
3606 
3607  /*
3608  * Sanity check --- don't divide by zero if empty relation.
3609  */
3610  Assert(IS_SIMPLE_REL(rel));
3611  if (rel->tuples > 0)
3612  {
3613  /*
3614  * Clamp to size of rel, or size of rel / 10 if multiple Vars. The
3615  * fudge factor is because the Vars are probably correlated but we
3616  * don't know by how much. We should never clamp to less than the
3617  * largest ndistinct value for any of the Vars, though, since
3618  * there will surely be at least that many groups.
3619  */
3620  double clamp = rel->tuples;
3621 
3622  if (relvarcount > 1)
3623  {
3624  clamp *= 0.1;
3625  if (clamp < relmaxndistinct)
3626  {
3627  clamp = relmaxndistinct;
3628  /* for sanity in case some ndistinct is too large: */
3629  if (clamp > rel->tuples)
3630  clamp = rel->tuples;
3631  }
3632  }
3633  if (reldistinct > clamp)
3634  reldistinct = clamp;
3635 
3636  /*
3637  * Update the estimate based on the restriction selectivity,
3638  * guarding against division by zero when reldistinct is zero.
3639  * Also skip this if we know that we are returning all rows.
3640  */
3641  if (reldistinct > 0 && rel->rows < rel->tuples)
3642  {
3643  /*
3644  * Given a table containing N rows with n distinct values in a
3645  * uniform distribution, if we select p rows at random then
3646  * the expected number of distinct values selected is
3647  *
3648  * n * (1 - product((N-N/n-i)/(N-i), i=0..p-1))
3649  *
3650  * = n * (1 - (N-N/n)! / (N-N/n-p)! * (N-p)! / N!)
3651  *
3652  * See "Approximating block accesses in database
3653  * organizations", S. B. Yao, Communications of the ACM,
3654  * Volume 20 Issue 4, April 1977 Pages 260-261.
3655  *
3656  * Alternatively, re-arranging the terms from the factorials,
3657  * this may be written as
3658  *
3659  * n * (1 - product((N-p-i)/(N-i), i=0..N/n-1))
3660  *
3661  * This form of the formula is more efficient to compute in
3662  * the common case where p is larger than N/n. Additionally,
3663  * as pointed out by Dell'Era, if i << N for all terms in the
3664  * product, it can be approximated by
3665  *
3666  * n * (1 - ((N-p)/N)^(N/n))
3667  *
3668  * See "Expected distinct values when selecting from a bag
3669  * without replacement", Alberto Dell'Era,
3670  * http://www.adellera.it/investigations/distinct_balls/.
3671  *
3672  * The condition i << N is equivalent to n >> 1, so this is a
3673  * good approximation when the number of distinct values in
3674  * the table is large. It turns out that this formula also
3675  * works well even when n is small.
3676  */
3677  reldistinct *=
3678  (1 - pow((rel->tuples - rel->rows) / rel->tuples,
3679  rel->tuples / reldistinct));
3680  }
3681  reldistinct = clamp_row_est(reldistinct);
3682 
3683  /*
3684  * Update estimate of total distinct groups.
3685  */
3686  numdistinct *= reldistinct;
3687  }
3688 
3689  varinfos = newvarinfos;
3690  } while (varinfos != NIL);
3691 
3692  /* Now we can account for the effects of any SRFs */
3693  numdistinct *= srf_multiplier;
3694 
3695  /* Round off */
3696  numdistinct = ceil(numdistinct);
3697 
3698  /* Guard against out-of-range answers */
3699  if (numdistinct > input_rows)
3700  numdistinct = input_rows;
3701  if (numdistinct < 1.0)
3702  numdistinct = 1.0;
3703 
3704  return numdistinct;
3705 }
3706 
3707 /*
3708  * Estimate hash bucket statistics when the specified expression is used
3709  * as a hash key for the given number of buckets.
3710  *
3711  * This attempts to determine two values:
3712  *
3713  * 1. The frequency of the most common value of the expression (returns
3714  * zero into *mcv_freq if we can't get that).
3715  *
3716  * 2. The "bucketsize fraction", ie, average number of entries in a bucket
3717  * divided by total tuples in relation.
3718  *
3719  * XXX This is really pretty bogus since we're effectively assuming that the
3720  * distribution of hash keys will be the same after applying restriction
3721  * clauses as it was in the underlying relation. However, we are not nearly
3722  * smart enough to figure out how the restrict clauses might change the
3723  * distribution, so this will have to do for now.
3724  *
3725  * We are passed the number of buckets the executor will use for the given
3726  * input relation. If the data were perfectly distributed, with the same
3727  * number of tuples going into each available bucket, then the bucketsize
3728  * fraction would be 1/nbuckets. But this happy state of affairs will occur
3729  * only if (a) there are at least nbuckets distinct data values, and (b)
3730  * we have a not-too-skewed data distribution. Otherwise the buckets will
3731  * be nonuniformly occupied. If the other relation in the join has a key
3732  * distribution similar to this one's, then the most-loaded buckets are
3733  * exactly those that will be probed most often. Therefore, the "average"
3734  * bucket size for costing purposes should really be taken as something close
3735  * to the "worst case" bucket size. We try to estimate this by adjusting the
3736  * fraction if there are too few distinct data values, and then scaling up
3737  * by the ratio of the most common value's frequency to the average frequency.
3738  *
3739  * If no statistics are available, use a default estimate of 0.1. This will
3740  * discourage use of a hash rather strongly if the inner relation is large,
3741  * which is what we want. We do not want to hash unless we know that the
3742  * inner rel is well-dispersed (or the alternatives seem much worse).
3743  *
3744  * The caller should also check that the mcv_freq is not so large that the
3745  * most common value would by itself require an impractically large bucket.
3746  * In a hash join, the executor can split buckets if they get too big, but
3747  * obviously that doesn't help for a bucket that contains many duplicates of
3748  * the same value.
3749  */
3750 void
3751 estimate_hash_bucket_stats(PlannerInfo *root, Node *hashkey, double nbuckets,
3752  Selectivity *mcv_freq,
3753  Selectivity *bucketsize_frac)
3754 {
3755  VariableStatData vardata;
3756  double estfract,
3757  ndistinct,
3758  stanullfrac,
3759  avgfreq;
3760  bool isdefault;
3761  AttStatsSlot sslot;
3762 
3763  examine_variable(root, hashkey, 0, &vardata);
3764 
3765  /* Look up the frequency of the most common value, if available */
3766  *mcv_freq = 0.0;
3767 
3768  if (HeapTupleIsValid(vardata.statsTuple))
3769  {
3770  if (get_attstatsslot(&sslot, vardata.statsTuple,
3771  STATISTIC_KIND_MCV, InvalidOid,
3772  ATTSTATSSLOT_NUMBERS))
3773  {
3774  /*
3775  * The first MCV stat is for the most common value.
3776  */
3777  if (sslot.nnumbers > 0)
3778  *mcv_freq = sslot.numbers[0];
3779  free_attstatsslot(&sslot);
3780  }
3781  }
3782 
3783  /* Get number of distinct values */
3784  ndistinct = get_variable_numdistinct(&vardata, &isdefault);
3785 
3786  /*
3787  * If ndistinct isn't real, punt. We normally return 0.1, but if the
3788  * mcv_freq is known to be even higher than that, use it instead.
3789  */
3790  if (isdefault)
3791  {
3792  *bucketsize_frac = (Selectivity) Max(0.1, *mcv_freq);
3793  ReleaseVariableStats(vardata);
3794  return;
3795  }
3796 
3797  /* Get fraction that are null */
3798  if (HeapTupleIsValid(vardata.statsTuple))
3799  {
3800  Form_pg_statistic stats;
3801 
3802  stats = (Form_pg_statistic) GETSTRUCT(vardata.statsTuple);
3803  stanullfrac = stats->stanullfrac;
3804  }
3805  else
3806  stanullfrac = 0.0;
3807 
3808  /* Compute avg freq of all distinct data values in raw relation */
3809  avgfreq = (1.0 - stanullfrac) / ndistinct;
3810 
3811  /*
3812  * Adjust ndistinct to account for restriction clauses. Observe we are
3813  * assuming that the data distribution is affected uniformly by the
3814  * restriction clauses!
3815  *
3816  * XXX Possibly better way, but much more expensive: multiply by
3817  * selectivity of rel's restriction clauses that mention the target Var.
3818  */
3819  if (vardata.rel && vardata.rel->tuples > 0)
3820  {
3821  ndistinct *= vardata.rel->rows / vardata.rel->tuples;
3822  ndistinct = clamp_row_est(ndistinct);
3823  }
3824 
3825  /*
3826  * Initial estimate of bucketsize fraction is 1/nbuckets as long as the
3827  * number of buckets is less than the expected number of distinct values;
3828  * otherwise it is 1/ndistinct.
3829  */
3830  if (ndistinct > nbuckets)
3831  estfract = 1.0 / nbuckets;
3832  else
3833  estfract = 1.0 / ndistinct;
3834 
3835  /*
3836  * Adjust estimated bucketsize upward to account for skewed distribution.
3837  */
3838  if (avgfreq > 0.0 && *mcv_freq > avgfreq)
3839  estfract *= *mcv_freq / avgfreq;
3840 
3841  /*
3842  * Clamp bucketsize to sane range (the above adjustment could easily
3843  * produce an out-of-range result). We set the lower bound a little above
3844  * zero, since zero isn't a very sane result.
3845  */
3846  if (estfract < 1.0e-6)
3847  estfract = 1.0e-6;
3848  else if (estfract > 1.0)
3849  estfract = 1.0;
3850 
3851  *bucketsize_frac = (Selectivity) estfract;
3852 
3853  ReleaseVariableStats(vardata);
3854 }
3855 
3856 /*
3857  * estimate_hashagg_tablesize
3858  * estimate the number of bytes that a hash aggregate hashtable will
3859  * require based on the agg_costs, path width and number of groups.
3860  *
3861  * We return the result as "double" to forestall any possible overflow
3862  * problem in the multiplication by dNumGroups.
3863  *
3864  * XXX this may be over-estimating the size now that hashagg knows to omit
3865  * unneeded columns from the hashtable. Also for mixed-mode grouping sets,
3866  * grouping columns not in the hashed set are counted here even though hashagg
3867  * won't store them. Is this a problem?
3868  */
3869 double
3870 estimate_hashagg_tablesize(PlannerInfo *root, Path *path,
3871  const AggClauseCosts *agg_costs, double dNumGroups)
3872 {
3873  Size hashentrysize;
3874 
3875  hashentrysize = hash_agg_entry_size(list_length(root->aggtransinfos),
3876  path->pathtarget->width,
3877  agg_costs->transitionSpace);
3878 
3879  /*
3880  * Note that this disregards the effect of fill-factor and growth policy
3881  * of the hash table. That's probably ok, given that the default
3882  * fill-factor is relatively high. It'd be hard to meaningfully factor in
3883  * "double-in-size" growth policies here.
3884  */
3885  return hashentrysize * dNumGroups;
3886 }
3887 
3888 
3889 /*-------------------------------------------------------------------------
3890  *
3891  * Support routines
3892  *
3893  *-------------------------------------------------------------------------
3894  */
3895 
3896 /*
3897  * Find applicable ndistinct statistics for the given list of VarInfos (which
3898  * must all belong to the given rel), and update *ndistinct to the estimate of
3899  * the MVNDistinctItem that best matches. If a match it found, *varinfos is
3900  * updated to remove the list of matched varinfos.
3901  *
3902  * Varinfos that aren't for simple Vars are ignored.
3903  *
3904  * Return true if we're able to find a match, false otherwise.
3905  */
3906 static bool
3907 estimate_multivariate_ndistinct(PlannerInfo *root, RelOptInfo *rel,
3908  List **varinfos, double *ndistinct)
3909 {
3910  ListCell *lc;
3911  int nmatches_vars;
3912  int nmatches_exprs;
3913  Oid statOid = InvalidOid;
3914  MVNDistinct *stats;
3915  StatisticExtInfo *matched_info = NULL;
3916 
3917  /* bail out immediately if the table has no extended statistics */
3918  if (!rel->statlist)
3919  return false;
3920 
3921  /* look for the ndistinct statistics matching the most vars */
3922  nmatches_vars = 0; /* we require at least two matches */
3923  nmatches_exprs = 0;
3924  foreach(lc, rel->statlist)
3925  {
3926  ListCell *lc2;
3927  StatisticExtInfo *info = (StatisticExtInfo *) lfirst(lc);
3928  int nshared_vars = 0;
3929  int nshared_exprs = 0;
3930 
3931  /* skip statistics of other kinds */
3932  if (info->kind != STATS_EXT_NDISTINCT)
3933  continue;
3934 
3935  /*
3936  * Determine how many expressions (and variables in non-matched
3937  * expressions) match. We'll then use these numbers to pick the
3938  * statistics object that best matches the clauses.
3939  */
3940  foreach(lc2, *varinfos)
3941  {
3942  ListCell *lc3;
3943  GroupVarInfo *varinfo = (GroupVarInfo *) lfirst(lc2);
3944  AttrNumber attnum;
3945 
3946  Assert(varinfo->rel == rel);
3947 
3948  /* simple Var, search in statistics keys directly */
3949  if (IsA(varinfo->var, Var))
3950  {
3951  attnum = ((Var *) varinfo->var)->varattno;
3952 
3953  /*
3954  * Ignore system attributes - we don't support statistics on
3955  * them, so can't match them (and it'd fail as the values are
3956  * negative).
3957  */
3958  if (!AttrNumberIsForUserDefinedAttr(attnum))
3959  continue;
3960 
3961  if (bms_is_member(attnum, info->keys))
3962  nshared_vars++;
3963 
3964  continue;
3965  }
3966 
3967  /* expression - see if it's in the statistics */
3968  foreach(lc3, info->exprs)
3969  {
3970  Node *expr = (Node *) lfirst(lc3);
3971 
3972  if (equal(varinfo->var, expr))
3973  {
3974  nshared_exprs++;
3975  break;
3976  }
3977  }
3978  }
3979 
3980  if (nshared_vars + nshared_exprs < 2)
3981  continue;
3982 
3983  /*
3984  * Does this statistics object match more columns than the currently
3985  * best object? If so, use this one instead.
3986  *
3987  * XXX This should break ties using name of the object, or something
3988  * like that, to make the outcome stable.
3989  */
3990  if ((nshared_exprs > nmatches_exprs) ||
3991  (((nshared_exprs == nmatches_exprs)) && (nshared_vars > nmatches_vars)))
3992  {
3993  statOid = info->statOid;
3994  nmatches_vars = nshared_vars;
3995  nmatches_exprs = nshared_exprs;
3996  matched_info = info;
3997  }
3998  }
3999 
4000  /* No match? */
4001  if (statOid == InvalidOid)
4002  return false;
4003 
4004  Assert(nmatches_vars + nmatches_exprs > 1);
4005 
4006  stats = statext_ndistinct_load(statOid);
4007 
4008  /*
4009  * If we have a match, search it for the specific item that matches (there
4010  * must be one), and construct the output values.
4011  */
4012  if (stats)
4013  {
4014  int i;
4015  List *newlist = NIL;
4016  MVNDistinctItem *item = NULL;
4017  ListCell *lc2;
4018  Bitmapset *matched = NULL;
4019  AttrNumber attnum_offset;
4020 
4021  /*
4022  * How much we need to offset the attnums? If there are no
4023  * expressions, no offset is needed. Otherwise offset enough to move
4024  * the lowest one (which is equal to number of expressions) to 1.
4025  */
4026  if (matched_info->exprs)
4027  attnum_offset = (list_length(matched_info->exprs) + 1);
4028  else
4029  attnum_offset = 0;
4030 
4031  /* see what actually matched */
4032  foreach(lc2, *varinfos)
4033  {
4034  ListCell *lc3;
4035  int idx;
4036  bool found = false;
4037 
4038  GroupVarInfo *varinfo = (GroupVarInfo *) lfirst(lc2);
4039 
4040  /*
4041  * Process a simple Var expression, by matching it to keys
4042  * directly. If there's a matching expression, we'll try matching
4043  * it later.
4044  */
4045  if (IsA(varinfo->var, Var))
4046  {
4047  AttrNumber attnum = ((Var *) varinfo->var)->varattno;
4048 
4049  /*
4050  * Ignore expressions on system attributes. Can't rely on the
4051  * bms check for negative values.
4052  */
4053  if (!AttrNumberIsForUserDefinedAttr(attnum))
4054  continue;
4055 
4056  /* Is the variable covered by the statistics? */
4057  if (!bms_is_member(attnum, matched_info->keys))
4058  continue;
4059 
4060  attnum = attnum + attnum_offset;
4061 
4062  /* ensure sufficient offset */
4063  Assert(AttrNumberIsForUserDefinedAttr(attnum));
4064 
4065  matched = bms_add_member(matched, attnum);
4066 
4067  found = true;
4068  }
4069 
4070  /*
4071  * XXX Maybe we should allow searching the expressions even if we
4072  * found an attribute matching the expression? That would handle
4073  * trivial expressions like "(a)" but it seems fairly useless.
4074  */
4075  if (found)
4076  continue;
4077 
4078  /* expression - see if it's in the statistics */
4079  idx = 0;
4080  foreach(lc3, matched_info->exprs)
4081  {
4082  Node *expr = (Node *) lfirst(lc3);
4083 
4084  if (equal(varinfo->var, expr))
4085  {
4086  AttrNumber attnum = -(idx + 1);
4087 
4088  attnum = attnum + attnum_offset;
4089 
4090  /* ensure sufficient offset */
4091  Assert(AttrNumberIsForUserDefinedAttr(attnum));
4092 
4093  matched = bms_add_member(matched, attnum);
4094 
4095  /* there should be just one matching expression */
4096  break;
4097  }
4098 
4099  idx++;
4100  }
4101  }
4102 
4103  /* Find the specific item that exactly matches the combination */
4104  for (i = 0; i < stats->nitems; i++)
4105  {
4106  int j;
4107  MVNDistinctItem *tmpitem = &stats->items[i];
4108 
4109  if (tmpitem->nattributes != bms_num_members(matched))
4110  continue;
4111 
4112  /* assume it's the right item */
4113  item = tmpitem;
4114 
4115  /* check that all item attributes/expressions fit the match */
4116  for (j = 0; j < tmpitem->nattributes; j++)
4117  {
4118  AttrNumber attnum = tmpitem->attributes[j];
4119 
4120  /*
4121  * Thanks to how we constructed the matched bitmap above, we
4122  * can just offset all attnums the same way.
4123  */
4124  attnum = attnum + attnum_offset;
4125 
4126  if (!bms_is_member(attnum, matched))
4127  {
4128  /* nah, it's not this item */
4129  item = NULL;
4130  break;
4131  }
4132  }
4133 
4134  /*
4135  * If the item has all the matched attributes, we know it's the
4136  * right one - there can't be a better one. matching more.
4137  */
4138  if (item)
4139  break;
4140  }
4141 
4142  /*
4143  * Make sure we found an item. There has to be one, because ndistinct
4144  * statistics includes all combinations of attributes.
4145  */
4146  if (!item)
4147  elog(ERROR, "corrupt MVNDistinct entry");
4148 
4149  /* Form the output varinfo list, keeping only unmatched ones */
4150  foreach(lc, *varinfos)
4151  {
4152  GroupVarInfo *varinfo = (GroupVarInfo *) lfirst(lc);
4153  ListCell *lc3;
4154  bool found = false;
4155 
4156  /*
4157  * Let's look at plain variables first, because it's the most
4158  * common case and the check is quite cheap. We can simply get the
4159  * attnum and check (with an offset) matched bitmap.
4160  */
4161  if (IsA(varinfo->var, Var))
4162  {
4163  AttrNumber attnum = ((Var *) varinfo->var)->varattno;
4164 
4165  /*
4166  * If it's a system attribute, we're done. We don't support
4167  * extended statistics on system attributes, so it's clearly
4168  * not matched. Just keep the expression and continue.
4169  */
4170  if (!AttrNumberIsForUserDefinedAttr(attnum))
4171  {
4172  newlist = lappend(newlist, varinfo);
4173  continue;
4174  }
4175 
4176  /* apply the same offset as above */
4177  attnum += attnum_offset;
4178 
4179  /* if it's not matched, keep the varinfo */
4180  if (!bms_is_member(attnum, matched))
4181  newlist = lappend(newlist, varinfo);
4182 
4183  /* The rest of the loop deals with complex expressions. */
4184  continue;
4185  }
4186 
4187  /*
4188  * Process complex expressions, not just simple Vars.
4189  *
4190  * First, we search for an exact match of an expression. If we
4191  * find one, we can just discard the whole GroupExprInfo, with all
4192  * the variables we extracted from it.
4193  *
4194  * Otherwise we inspect the individual vars, and try matching it
4195  * to variables in the item.
4196  */
4197  foreach(lc3, matched_info->exprs)
4198  {
4199  Node *expr = (Node *) lfirst(lc3);
4200 
4201  if (equal(varinfo->var, expr))
4202  {
4203  found = true;
4204  break;
4205  }
4206  }
4207 
4208  /* found exact match, skip */
4209  if (found)
4210  continue;
4211 
4212  newlist = lappend(newlist, varinfo);
4213  }
4214 
4215  *varinfos = newlist;
4216  *ndistinct = item->ndistinct;
4217  return true;
4218  }
4219 
4220  return false;
4221 }
4222 
4223 /*
4224  * convert_to_scalar
4225  * Convert non-NULL values of the indicated types to the comparison
4226  * scale needed by scalarineqsel().
4227  * Returns "true" if successful.
4228  *
4229  * XXX this routine is a hack: ideally we should look up the conversion
4230  * subroutines in pg_type.
4231  *
4232  * All numeric datatypes are simply converted to their equivalent
4233  * "double" values. (NUMERIC values that are outside the range of "double"
4234  * are clamped to +/- HUGE_VAL.)
4235  *
4236  * String datatypes are converted by convert_string_to_scalar(),
4237  * which is explained below. The reason why this routine deals with
4238  * three values at a time, not just one, is that we need it for strings.
4239  *
4240  * The bytea datatype is just enough different from strings that it has
4241  * to be treated separately.
4242  *
4243  * The several datatypes representing absolute times are all converted
4244  * to Timestamp, which is actually an int64, and then we promote that to
4245  * a double. Note this will give correct results even for the "special"
4246  * values of Timestamp, since those are chosen to compare correctly;
4247  * see timestamp_cmp.
4248  *
4249  * The several datatypes representing relative times (intervals) are all
4250  * converted to measurements expressed in seconds.
4251  */
4252 static bool
4253 convert_to_scalar(Datum value, Oid valuetypid, Oid collid, double *scaledvalue,
4254  Datum lobound, Datum hibound, Oid boundstypid,
4255  double *scaledlobound, double *scaledhibound)
4256 {
4257  bool failure = false;
4258 
4259  /*
4260  * Both the valuetypid and the boundstypid should exactly match the
4261  * declared input type(s) of the operator we are invoked for. However,
4262  * extensions might try to use scalarineqsel as estimator for operators
4263  * with input type(s) we don't handle here; in such cases, we want to
4264  * return false, not fail. In any case, we mustn't assume that valuetypid
4265  * and boundstypid are identical.
4266  *
4267  * XXX The histogram we are interpolating between points of could belong
4268  * to a column that's only binary-compatible with the declared type. In
4269  * essence we are assuming that the semantics of binary-compatible types
4270  * are enough alike that we can use a histogram generated with one type's
4271  * operators to estimate selectivity for the other's. This is outright
4272  * wrong in some cases --- in particular signed versus unsigned
4273  * interpretation could trip us up. But it's useful enough in the
4274  * majority of cases that we do it anyway. Should think about more
4275  * rigorous ways to do it.
4276  */
4277  switch (valuetypid)
4278  {
4279  /*
4280  * Built-in numeric types
4281  */
4282  case BOOLOID:
4283  case INT2OID:
4284  case INT4OID:
4285  case INT8OID:
4286  case FLOAT4OID:
4287  case FLOAT8OID:
4288  case NUMERICOID:
4289  case OIDOID:
4290  case REGPROCOID:
4291  case REGPROCEDUREOID:
4292  case REGOPEROID:
4293  case REGOPERATOROID:
4294  case REGCLASSOID:
4295  case REGTYPEOID:
4296  case REGCONFIGOID:
4297  case REGDICTIONARYOID:
4298  case REGROLEOID:
4299  case REGNAMESPACEOID:
4300  *scaledvalue = convert_numeric_to_scalar(value, valuetypid,
4301  &failure);
4302  *scaledlobound = convert_numeric_to_scalar(lobound, boundstypid,
4303  &failure);
4304  *scaledhibound = convert_numeric_to_scalar(hibound, boundstypid,
4305  &failure);
4306  return !failure;
4307 
4308  /*
4309  * Built-in string types
4310  */
4311  case CHAROID:
4312  case BPCHAROID:
4313  case VARCHAROID:
4314  case TEXTOID:
4315  case NAMEOID:
4316  {
4317  char *valstr = convert_string_datum(value, valuetypid,
4318  collid, &failure);
4319  char *lostr = convert_string_datum(lobound, boundstypid,
4320  collid, &failure);
4321  char *histr = convert_string_datum(hibound, boundstypid,
4322  collid, &failure);
4323 
4324  /*
4325  * Bail out if any of the values is not of string type. We
4326  * might leak converted strings for the other value(s), but
4327  * that's not worth troubling over.
4328  */
4329  if (failure)
4330  return false;
4331 
4332  convert_string_to_scalar(valstr, scaledvalue,
4333  lostr, scaledlobound,
4334  histr, scaledhibound);
4335  pfree(valstr);
4336  pfree(lostr);
4337  pfree(histr);
4338  return true;
4339  }
4340 
4341  /*
4342  * Built-in bytea type
4343  */
4344  case BYTEAOID:
4345  {
4346  /* We only support bytea vs bytea comparison */
4347  if (boundstypid != BYTEAOID)
4348  return false;
4349  convert_bytea_to_scalar(value, scaledvalue,
4350  lobound, scaledlobound,
4351  hibound, scaledhibound);
4352  return true;
4353  }
4354 
4355  /*
4356  * Built-in time types
4357  */
4358  case TIMESTAMPOID:
4359  case TIMESTAMPTZOID:
4360  case DATEOID:
4361  case INTERVALOID:
4362  case TIMEOID:
4363  case TIMETZOID:
4364  *scaledvalue = convert_timevalue_to_scalar(value, valuetypid,
4365  &failure);
4366  *scaledlobound = convert_timevalue_to_scalar(lobound, boundstypid,
4367  &failure);
4368  *scaledhibound = convert_timevalue_to_scalar(hibound, boundstypid,
4369  &failure);
4370  return !failure;
4371 
4372  /*
4373  * Built-in network types
4374  */
4375  case INETOID:
4376  case CIDROID:
4377  case MACADDROID:
4378  case MACADDR8OID:
4379  *scaledvalue = convert_network_to_scalar(value, valuetypid,
4380  &failure);
4381  *scaledlobound = convert_network_to_scalar(lobound, boundstypid,
4382  &failure);
4383  *scaledhibound = convert_network_to_scalar(hibound, boundstypid,
4384  &failure);
4385  return !failure;
4386  }
4387  /* Don't know how to convert */
4388  *scaledvalue = *scaledlobound = *scaledhibound = 0;
4389  return false;
4390 }
4391 
4392 /*
4393  * Do convert_to_scalar()'s work for any numeric data type.
4394  *
4395  * On failure (e.g., unsupported typid), set *failure to true;
4396  * otherwise, that variable is not changed.
4397  */
4398 static double
4399 convert_numeric_to_scalar(Datum value, Oid typid, bool *failure)
4400 {
4401  switch (typid)
4402  {
4403  case BOOLOID:
4404  return (double) DatumGetBool(value);
4405  case INT2OID:
4406  return (double) DatumGetInt16(value);
4407  case INT4OID:
4408  return (double) DatumGetInt32(value);
4409  case INT8OID:
4410  return (double) DatumGetInt64(value);
4411  case FLOAT4OID:
4412  return (double) DatumGetFloat4(value);
4413  case FLOAT8OID:
4414  return (double) DatumGetFloat8(value);
4415  case NUMERICOID:
4416  /* Note: out-of-range values will be clamped to +-HUGE_VAL */
4417  return (double)
4418  DatumGetFloat8(DirectFunctionCall1(numeric_float8_no_overflow,
4419  value));
4420  case OIDOID:
4421  case REGPROCOID:
4422  case REGPROCEDUREOID:
4423  case REGOPEROID:
4424  case REGOPERATOROID:
4425  case REGCLASSOID:
4426  case REGTYPEOID:
4427  case REGCONFIGOID:
4428  case REGDICTIONARYOID:
4429  case REGROLEOID:
4430  case REGNAMESPACEOID:
4431  /* we can treat OIDs as integers... */
4432  return (double) DatumGetObjectId(value);
4433  }
4434 
4435  *failure = true;
4436  return 0;
4437 }
4438 
4439 /*
4440  * Do convert_to_scalar()'s work for any character-string data type.
4441  *
4442  * String datatypes are converted to a scale that ranges from 0 to 1,
4443  * where we visualize the bytes of the string as fractional digits.
4444  *
4445  * We do not want the base to be 256, however, since that tends to
4446  * generate inflated selectivity estimates; few databases will have
4447  * occurrences of all 256 possible byte values at each position.
4448  * Instead, use the smallest and largest byte values seen in the bounds
4449  * as the estimated range for each byte, after some fudging to deal with
4450  * the fact that we probably aren't going to see the full range that way.
4451  *
4452  * An additional refinement is that we discard any common prefix of the
4453  * three strings before computing the scaled values. This allows us to
4454  * "zoom in" when we encounter a narrow data range. An example is a phone
4455  * number database where all the values begin with the same area code.
4456  * (Actually, the bounds will be adjacent histogram-bin-boundary values,
4457  * so this is more likely to happen than you might think.)
4458  */
4459 static void
4460 convert_string_to_scalar(char *value,
4461  double *scaledvalue,
4462  char *lobound,
4463  double *scaledlobound,
4464  char *hibound,
4465  double *scaledhibound)
4466 {
4467  int rangelo,
4468  rangehi;
4469  char *sptr;
4470 
4471  rangelo = rangehi = (unsigned char) hibound[0];
4472  for (sptr = lobound; *sptr; sptr++)
4473  {
4474  if (rangelo > (unsigned char) *sptr)
4475  rangelo = (unsigned char) *sptr;
4476  if (rangehi < (unsigned char) *sptr)
4477  rangehi = (unsigned char) *sptr;
4478  }
4479  for (sptr = hibound; *sptr; sptr++)
4480  {
4481  if (rangelo > (unsigned char) *sptr)
4482  rangelo = (unsigned char) *sptr;
4483  if (rangehi < (unsigned char) *sptr)
4484  rangehi = (unsigned char) *sptr;
4485  }
4486  /* If range includes any upper-case ASCII chars, make it include all */
4487  if (rangelo <= 'Z' && rangehi >= 'A')
4488  {
4489  if (rangelo > 'A')
4490  rangelo = 'A';
4491  if (rangehi < 'Z')
4492  rangehi = 'Z';
4493  }
4494  /* Ditto lower-case */
4495  if (rangelo <= 'z' && rangehi >= 'a')
4496  {
4497  if (rangelo > 'a')
4498  rangelo = 'a';
4499  if (rangehi < 'z')
4500  rangehi = 'z';
4501  }
4502  /* Ditto digits */
4503  if (rangelo <= '9' && rangehi >= '0')
4504  {
4505  if (rangelo > '0')
4506  rangelo = '0';
4507  if (rangehi < '9')
4508  rangehi = '9';
4509  }
4510 
4511  /*
4512  * If range includes less than 10 chars, assume we have not got enough
4513  * data, and make it include regular ASCII set.
4514  */
4515  if (rangehi - rangelo < 9)
4516  {
4517  rangelo = ' ';
4518  rangehi = 127;
4519  }
4520 
4521  /*
4522  * Now strip any common prefix of the three strings.
4523  */
4524  while (*lobound)
4525  {
4526  if (*lobound != *hibound || *lobound != *value)
4527  break;
4528  lobound++, hibound++, value++;
4529  }
4530 
4531  /*
4532  * Now we can do the conversions.
4533  */
4534  *scaledvalue = convert_one_string_to_scalar(value, rangelo, rangehi);
4535  *scaledlobound = convert_one_string_to_scalar(lobound, rangelo, rangehi);
4536  *scaledhibound = convert_one_string_to_scalar(hibound, rangelo, rangehi);
4537 }
4538 
4539 static double
4540 convert_one_string_to_scalar(char *value, int rangelo, int rangehi)
4541 {
4542  int slen = strlen(value);
4543  double num,
4544  denom,
4545  base;
4546 
4547  if (slen <= 0)
4548  return 0.0; /* empty string has scalar value 0 */
4549 
4550  /*
4551  * There seems little point in considering more than a dozen bytes from
4552  * the string. Since base is at least 10, that will give us nominal
4553  * resolution of at least 12 decimal digits, which is surely far more
4554  * precision than this estimation technique has got anyway (especially in
4555  * non-C locales). Also, even with the maximum possible base of 256, this
4556  * ensures denom cannot grow larger than 256^13 = 2.03e31, which will not
4557  * overflow on any known machine.
4558  */
4559  if (slen > 12)
4560  slen = 12;
4561 
4562  /* Convert initial characters to fraction */
4563  base = rangehi - rangelo + 1;
4564  num = 0.0;
4565  denom = base;
4566  while (slen-- > 0)
4567  {
4568  int ch = (unsigned char) *value++;
4569 
4570  if (ch < rangelo)
4571  ch = rangelo - 1;
4572  else if (ch > rangehi)
4573  ch = rangehi + 1;
4574  num += ((double) (ch - rangelo)) / denom;
4575  denom *= base;
4576  }
4577 
4578  return num;
4579 }
4580 
4581 /*
4582  * Convert a string-type Datum into a palloc'd, null-terminated string.
4583  *
4584  * On failure (e.g., unsupported typid), set *failure to true;
4585  * otherwise, that variable is not changed. (We'll return NULL on failure.)
4586  *
4587  * When using a non-C locale, we must pass the string through strxfrm()
4588  * before continuing, so as to generate correct locale-specific results.
4589  */
4590 static char *
4591 convert_string_datum(Datum value, Oid typid, Oid collid, bool *failure)
4592 {
4593  char *val;
4594 
4595  switch (typid)
4596  {
4597  case CHAROID:
4598  val = (char *) palloc(2);
4599  val[0] = DatumGetChar(value);
4600  val[1] = '\0';
4601  break;
4602  case BPCHAROID:
4603  case VARCHAROID:
4604  case TEXTOID:
4605  val = TextDatumGetCString(value);
4606  break;
4607  case NAMEOID:
4608  {
4609  NameData *nm = (NameData *) DatumGetPointer(value);
4610 
4611  val = pstrdup(NameStr(*nm));
4612  break;
4613  }
4614  default:
4615  *failure = true;
4616  return NULL;
4617  }
4618 
4619  if (!lc_collate_is_c(collid))
4620  {
4621  char *xfrmstr;
4622  size_t xfrmlen;
4623  size_t xfrmlen2 PG_USED_FOR_ASSERTS_ONLY;
4624 
4625  /*
4626  * XXX: We could guess at a suitable output buffer size and only call
4627  * strxfrm twice if our guess is too small.
4628  *
4629  * XXX: strxfrm doesn't support UTF-8 encoding on Win32, it can return
4630  * bogus data or set an error. This is not really a problem unless it
4631  * crashes since it will only give an estimation error and nothing
4632  * fatal.
4633  */
4634  xfrmlen = strxfrm(NULL, val, 0);
4635 #ifdef WIN32
4636 
4637  /*
4638  * On Windows, strxfrm returns INT_MAX when an error occurs. Instead
4639  * of trying to allocate this much memory (and fail), just return the
4640  * original string unmodified as if we were in the C locale.
4641  */
4642  if (xfrmlen == INT_MAX)
4643  return val;
4644 #endif
4645  xfrmstr = (char *) palloc(xfrmlen + 1);
4646  xfrmlen2 = strxfrm(xfrmstr, val, xfrmlen + 1);
4647 
4648  /*
4649  * Some systems (e.g., glibc) can return a smaller value from the
4650  * second call than the first; thus the Assert must be <= not ==.
4651  */
4652  Assert(xfrmlen2 <= xfrmlen);
4653  pfree(val);
4654  val = xfrmstr;
4655  }
4656 
4657  return val;
4658 }
4659 
4660 /*
4661  * Do convert_to_scalar()'s work for any bytea data type.
4662  *
4663  * Very similar to convert_string_to_scalar except we can't assume
4664  * null-termination and therefore pass explicit lengths around.
4665  *
4666  * Also, assumptions about likely "normal" ranges of characters have been
4667  * removed - a data range of 0..255 is always used, for now. (Perhaps
4668  * someday we will add information about actual byte data range to
4669  * pg_statistic.)
4670  */
4671 static void
4672 convert_bytea_to_scalar(Datum value,
4673  double *scaledvalue,
4674  Datum lobound,
4675  double *scaledlobound,
4676  Datum hibound,
4677  double *scaledhibound)
4678 {
4679  bytea *valuep = DatumGetByteaPP(value);
4680  bytea *loboundp = DatumGetByteaPP(lobound);
4681  bytea *hiboundp = DatumGetByteaPP(hibound);
4682  int rangelo,
4683  rangehi,
4684  valuelen = VARSIZE_ANY_EXHDR(valuep),
4685  loboundlen = VARSIZE_ANY_EXHDR(loboundp),
4686  hiboundlen = VARSIZE_ANY_EXHDR(hiboundp),
4687  i,
4688  minlen;
4689  unsigned char *valstr = (unsigned char *) VARDATA_ANY(valuep);
4690  unsigned char *lostr = (unsigned char *) VARDATA_ANY(loboundp);
4691  unsigned char *histr = (unsigned char *) VARDATA_ANY(hiboundp);
4692 
4693  /*
4694  * Assume bytea data is uniformly distributed across all byte values.
4695  */
4696  rangelo = 0;
4697  rangehi = 255;
4698 
4699  /*
4700  * Now strip any common prefix of the three strings.
4701  */
4702  minlen = Min(Min(valuelen, loboundlen), hiboundlen);
4703  for (i = 0; i < minlen; i++)
4704  {
4705  if (*lostr != *histr || *lostr != *valstr)
4706  break;
4707  lostr++, histr++, valstr++;
4708  loboundlen--, hiboundlen--, valuelen--;
4709  }
4710 
4711  /*
4712  * Now we can do the conversions.
4713  */
4714  *scaledvalue = convert_one_bytea_to_scalar(valstr, valuelen, rangelo, rangehi);
4715  *scaledlobound = convert_one_bytea_to_scalar(lostr, loboundlen, rangelo, rangehi);
4716  *scaledhibound = convert_one_bytea_to_scalar(histr, hiboundlen, rangelo, rangehi);
4717 }
4718 
4719 static double
4720 convert_one_bytea_to_scalar(unsigned char *value, int valuelen,
4721  int rangelo, int rangehi)
4722 {
4723  double num,
4724  denom,
4725  base;
4726 
4727  if (valuelen <= 0)
4728  return 0.0; /* empty string has scalar value 0 */
4729 
4730  /*
4731  * Since base is 256, need not consider more than about 10 chars (even
4732  * this many seems like overkill)
4733  */
4734  if (valuelen > 10)
4735  valuelen = 10;
4736 
4737  /* Convert initial characters to fraction */
4738  base = rangehi - rangelo + 1;
4739  num = 0.0;
4740  denom = base;
4741  while (valuelen-- > 0)
4742  {
4743  int ch = *value++;
4744 
4745  if (ch < rangelo)
4746  ch = rangelo - 1;
4747  else if (ch > rangehi)
4748  ch = rangehi + 1;
4749  num += ((double) (ch - rangelo)) / denom;
4750  denom *= base;
4751  }
4752 
4753  return num;
4754 }
4755 
4756 /*
4757  * Do convert_to_scalar()'s work for any timevalue data type.
4758  *
4759  * On failure (e.g., unsupported typid), set *failure to true;
4760  * otherwise, that variable is not changed.
4761  */
4762 static double
4763 convert_timevalue_to_scalar(Datum value, Oid typid, bool *failure)
4764 {
4765  switch (typid)
4766  {
4767  case TIMESTAMPOID:
4768  return DatumGetTimestamp(value);
4769  case TIMESTAMPTZOID:
4770  return DatumGetTimestampTz(value);
4771  case DATEOID:
4772  return date2timestamp_no_overflow(DatumGetDateADT(value));
4773  case INTERVALOID:
4774  {
4775  Interval *interval = DatumGetIntervalP(value);
4776 
4777  /*
4778  * Convert the month part of Interval to days using assumed
4779  * average month length of 365.25/12.0 days. Not too
4780  * accurate, but plenty good enough for our purposes.
4781  */
4782  return interval->time + interval->day * (double) USECS_PER_DAY +
4783  interval->month * ((DAYS_PER_YEAR / (double) MONTHS_PER_YEAR) * USECS_PER_DAY);
4784  }
4785  case TIMEOID:
4786  return DatumGetTimeADT(value);
4787  case TIMETZOID:
4788  {
4789  TimeTzADT *timetz = DatumGetTimeTzADTP(value);
4790 
4791  /* use GMT-equivalent time */
4792  return (double) (timetz->time + (timetz->zone * 1000000.0));
4793  }
4794  }
4795 
4796  *failure = true;
4797  return 0;
4798 }
4799 
4800 
4801 /*
4802  * get_restriction_variable
4803  * Examine the args of a restriction clause to see if it's of the
4804  * form (variable op pseudoconstant) or (pseudoconstant op variable),
4805  * where "variable" could be either a Var or an expression in vars of a
4806  * single relation. If so, extract information about the variable,
4807  * and also indicate which side it was on and the other argument.
4808  *
4809  * Inputs:
4810  * root: the planner info
4811  * args: clause argument list
4812  * varRelid: see specs for restriction selectivity functions
4813  *
4814  * Outputs: (these are valid only if true is returned)
4815  * *vardata: gets information about variable (see examine_variable)
4816  * *other: gets other clause argument, aggressively reduced to a constant
4817  * *varonleft: set true if variable is on the left, false if on the right
4818  *
4819  * Returns true if a variable is identified, otherwise false.
4820  *
4821  * Note: if there are Vars on both sides of the clause, we must fail, because
4822  * callers are expecting that the other side will act like a pseudoconstant.
4823  */
4824 bool
4825 get_restriction_variable(PlannerInfo *root, List *args, int varRelid,
4826  VariableStatData *vardata, Node **other,
4827  bool *varonleft)
4828 {
4829  Node *left,
4830  *right;
4831  VariableStatData rdata;
4832 
4833  /* Fail if not a binary opclause (probably shouldn't happen) */
4834  if (list_length(args) != 2)
4835  return false;
4836 
4837  left = (Node *) linitial(args);
4838  right = (Node *) lsecond(args);
4839 
4840  /*
4841  * Examine both sides. Note that when varRelid is nonzero, Vars of other
4842  * relations will be treated as pseudoconstants.
4843  */
4844  examine_variable(root, left, varRelid, vardata);
4845  examine_variable(root, right, varRelid, &rdata);
4846 
4847  /*
4848  * If one side is a variable and the other not, we win.
4849  */
4850  if (vardata->rel && rdata.rel == NULL)
4851  {
4852  *varonleft = true;
4853  *other = estimate_expression_value(root, rdata.var);
4854  /* Assume we need no ReleaseVariableStats(rdata) here */
4855  return true;
4856  }
4857 
4858  if (vardata->rel == NULL && rdata.rel)
4859  {
4860  *varonleft = false;
4861  *other = estimate_expression_value(root, vardata->var);
4862  /* Assume we need no ReleaseVariableStats(*vardata) here */
4863  *vardata = rdata;
4864  return true;
4865  }
4866 
4867  /* Oops, clause has wrong structure (probably var op var) */
4868  ReleaseVariableStats(*vardata);
4869  ReleaseVariableStats(rdata);
4870 
4871  return false;
4872 }
4873 
4874 /*
4875  * get_join_variables
4876  * Apply examine_variable() to each side of a join clause.
4877  * Also, attempt to identify whether the join clause has the same
4878  * or reversed sense compared to the SpecialJoinInfo.
4879  *
4880  * We consider the join clause "normal" if it is "lhs_var OP rhs_var",
4881  * or "reversed" if it is "rhs_var OP lhs_var". In complicated cases
4882  * where we can't tell for sure, we default to assuming it's normal.
4883  */
4884 void
4885 get_join_variables(PlannerInfo *root, List *args, SpecialJoinInfo *sjinfo,
4886  VariableStatData *vardata1, VariableStatData *vardata2,
4887  bool *join_is_reversed)
4888 {
4889  Node *left,
4890  *right;
4891 
4892  if (list_length(args) != 2)
4893  elog(ERROR, "join operator should take two arguments");
4894 
4895  left = (Node *) linitial(args);
4896  right = (Node *) lsecond(args);
4897 
4898  examine_variable(root, left, 0, vardata1);
4899  examine_variable(root, right, 0, vardata2);
4900 
4901  if (vardata1->rel &&
4902  bms_is_subset(vardata1->rel->relids, sjinfo->syn_righthand))
4903  *join_is_reversed = true; /* var1 is on RHS */
4904  else if (vardata2->rel &&
4905  bms_is_subset(vardata2->rel->relids, sjinfo->syn_lefthand))
4906  *join_is_reversed = true; /* var2 is on LHS */
4907  else
4908  *join_is_reversed = false;
4909 }
4910 
4911 /* statext_expressions_load copies the tuple, so just pfree it. */
4912 static void
4913 ReleaseDummy(HeapTuple tuple)
4914 {
4915  pfree(tuple);
4916 }
4917 
4918 /*
4919  * examine_variable
4920  * Try to look up statistical data about an expression.
4921  * Fill in a VariableStatData struct to describe the expression.
4922  *
4923  * Inputs:
4924  * root: the planner info
4925  * node: the expression tree to examine
4926  * varRelid: see specs for restriction selectivity functions
4927  *
4928  * Outputs: *vardata is filled as follows:
4929  * var: the input expression (with any binary relabeling stripped, if
4930  * it is or contains a variable; but otherwise the type is preserved)
4931  * rel: RelOptInfo for relation containing variable; NULL if expression
4932  * contains no Vars (NOTE this could point to a RelOptInfo of a
4933  * subquery, not one in the current query).
4934  * statsTuple: the pg_statistic entry for the variable, if one exists;
4935  * otherwise NULL.
4936  * freefunc: pointer to a function to release statsTuple with.
4937  * vartype: exposed type of the expression; this should always match
4938  * the declared input type of the operator we are estimating for.
4939  * atttype, atttypmod: actual type/typmod of the "var" expression. This is
4940  * commonly the same as the exposed type of the variable argument,
4941  * but can be different in binary-compatible-type cases.
4942  * isunique: true if we were able to match the var to a unique index or a
4943  * single-column DISTINCT clause, implying its values are unique for
4944  * this query. (Caution: this should be trusted for statistical
4945  * purposes only, since we do not check indimmediate nor verify that
4946  * the exact same definition of equality applies.)
4947  * acl_ok: true if current user has permission to read the column(s)
4948  * underlying the pg_statistic entry. This is consulted by
4949  * statistic_proc_security_check().
4950  *
4951  * Caller is responsible for doing ReleaseVariableStats() before exiting.
4952  */
4953 void
4954 examine_variable(PlannerInfo *root, Node *node, int varRelid,
4955  VariableStatData *vardata)
4956 {
4957  Node *basenode;
4958  Relids varnos;
4959  RelOptInfo *onerel;
4960 
4961  /* Make sure we don't return dangling pointers in vardata */
4962  MemSet(vardata, 0, sizeof(VariableStatData));
4963 
4964  /* Save the exposed type of the expression */
4965  vardata->vartype = exprType(node);
4966 
4967  /* Look inside any binary-compatible relabeling */
4968 
4969  if (IsA(node, RelabelType))
4970  basenode = (Node *) ((RelabelType *) node)->arg;
4971  else
4972  basenode = node;
4973 
4974  /* Fast path for a simple Var */
4975 
4976  if (IsA(basenode, Var) &&
4977  (varRelid == 0 || varRelid == ((Var *) basenode)->varno))
4978  {
4979  Var *var = (Var *) basenode;
4980 
4981  /* Set up result fields other than the stats tuple */
4982  vardata->var = basenode; /* return Var without relabeling */
4983  vardata->rel = find_base_rel(root, var->varno);
4984  vardata->atttype = var->vartype;
4985  vardata->atttypmod = var->vartypmod;
4986  vardata->isunique = has_unique_index(vardata->rel, var->varattno);
4987 
4988  /* Try to locate some stats */
4989  examine_simple_variable(root, var, vardata);
4990 
4991  return;
4992  }
4993 
4994  /*
4995  * Okay, it's a more complicated expression. Determine variable
4996  * membership. Note that when varRelid isn't zero, only vars of that
4997  * relation are considered "real" vars.
4998  */
4999  varnos = pull_varnos(root, basenode);
5000 
5001  onerel = NULL;
5002 
5003  switch (bms_membership(varnos))
5004  {
5005  case BMS_EMPTY_SET:
5006  /* No Vars at all ... must be pseudo-constant clause */
5007  break;
5008  case BMS_SINGLETON:
5009  if (varRelid == 0 || bms_is_member(varRelid, varnos))
5010  {
5011  onerel = find_base_rel(root,
5012  (varRelid ? varRelid : bms_singleton_member(varnos)));
5013  vardata->rel = onerel;
5014  node = basenode; /* strip any relabeling */
5015  }
5016  /* else treat it as a constant */
5017  break;
5018  case BMS_MULTIPLE:
5019  if (varRelid == 0)
5020  {
5021  /* treat it as a variable of a join relation */
5022  vardata->rel = find_join_rel(root, varnos);
5023  node = basenode; /* strip any relabeling */
5024  }
5025  else if (bms_is_member(varRelid, varnos))
5026  {
5027  /* ignore the vars belonging to other relations */
5028  vardata->rel = find_base_rel(root, varRelid);
5029  node = basenode; /* strip any relabeling */
5030  /* note: no point in expressional-index search here */
5031  }
5032  /* else treat it as a constant */
5033  break;
5034  }
5035 
5036  bms_free(varnos);
5037 
5038  vardata->var = node;
5039  vardata->atttype = exprType(node);
5040  vardata->atttypmod = exprTypmod(node);
5041 
5042  if (onerel)
5043  {
5044  /*
5045  * We have an expression in vars of a single relation. Try to match
5046  * it to expressional index columns, in hopes of finding some
5047  * statistics.
5048  *
5049  * Note that we consider all index columns including INCLUDE columns,
5050  * since there could be stats for such columns. But the test for
5051  * uniqueness needs to be warier.
5052  *
5053  * XXX it's conceivable that there are multiple matches with different
5054  * index opfamilies; if so, we need to pick one that matches the
5055  * operator we are estimating for. FIXME later.
5056  */
5057  ListCell *ilist;
5058  ListCell *slist;
5059 
5060  foreach(ilist, onerel->indexlist)
5061  {
5062  IndexOptInfo *index = (IndexOptInfo *) lfirst(ilist);
5063  ListCell *indexpr_item;
5064  int pos;
5065 
5066  indexpr_item = list_head(index->indexprs);
5067  if (indexpr_item == NULL)
5068  continue; /* no expressions here... */
5069 
5070  for (pos = 0; pos < index->ncolumns; pos++)
5071  {
5072  if (index->indexkeys[pos] == 0)
5073  {
5074  Node *indexkey;
5075 
5076  if (indexpr_item == NULL)
5077  elog(ERROR, "too few entries in indexprs list");
5078  indexkey = (Node *) lfirst(indexpr_item);
5079  if (indexkey && IsA(indexkey, RelabelType))
5080  indexkey = (Node *) ((RelabelType *) indexkey)->arg;
5081  if (equal(node, indexkey))
5082  {
5083  /*
5084  * Found a match ... is it a unique index? Tests here
5085  * should match has_unique_index().
5086  */
5087  if (index->unique &&
5088  index->nkeycolumns == 1 &&
5089  pos == 0 &&
5090  (index->indpred == NIL || index->predOK))
5091  vardata->isunique = true;
5092 
5093  /*
5094  * Has it got stats? We only consider stats for
5095  * non-partial indexes, since partial indexes probably
5096  * don't reflect whole-relation statistics; the above
5097  * check for uniqueness is the only info we take from
5098  * a partial index.
5099  *
5100  * An index stats hook, however, must make its own
5101  * decisions about what to do with partial indexes.
5102  */
5103  if (get_index_stats_hook &&
5104  (*get_index_stats_hook) (root, index->indexoid,
5105  pos + 1, vardata))
5106  {
5107  /*
5108  * The hook took control of acquiring a stats
5109  * tuple. If it did supply a tuple, it'd better
5110  * have supplied a freefunc.
5111  */
5112  if (HeapTupleIsValid(vardata->statsTuple) &&
5113  !vardata->freefunc)
5114  elog(ERROR, "no function provided to release variable stats with");
5115  }
5116  else if (index->indpred == NIL)
5117  {
5118  vardata->statsTuple =
5119  SearchSysCache3(STATRELATTINH,
5120  ObjectIdGetDatum(index->indexoid),
5121  Int16GetDatum(pos + 1),
5122  BoolGetDatum(false));
5123  vardata->freefunc = ReleaseSysCache;
5124 
5125  if (HeapTupleIsValid(vardata->statsTuple))
5126  {
5127  /* Get index's table for permission check */
5128  RangeTblEntry *rte;
5129  Oid userid;
5130 
5131  rte = planner_rt_fetch(index->rel->relid, root);
5132  Assert(rte->rtekind == RTE_RELATION);
5133 
5134  /*
5135  * Use checkAsUser if it's set, in case we're
5136  * accessing the table via a view.
5137  */
5138  userid = rte->checkAsUser ? rte->checkAsUser : GetUserId();
5139 
5140  /*
5141  * For simplicity, we insist on the whole
5142  * table being selectable, rather than trying
5143  * to identify which column(s) the index
5144  * depends on. Also require all rows to be
5145  * selectable --- there must be no
5146  * securityQuals from security barrier views
5147  * or RLS policies.
5148  */
5149  vardata->acl_ok =
5150  rte->securityQuals == NIL &&
5151  (pg_class_aclcheck(rte->relid, userid,
5152  ACL_SELECT) == ACLCHECK_OK);
5153 
5154  /*
5155  * If the user doesn't have permissions to
5156  * access an inheritance child relation, check
5157  * the permissions of the table actually
5158  * mentioned in the query, since most likely
5159  * the user does have that permission. Note
5160  * that whole-table select privilege on the
5161  * parent doesn't quite guarantee that the
5162  * user could read all columns of the child.
5163  * But in practice it's unlikely that any
5164  * interesting security violation could result
5165  * from allowing access to the expression
5166  * index's stats, so we allow it anyway. See
5167  * similar code in examine_simple_variable()
5168  * for additional comments.
5169  */
5170  if (!vardata->acl_ok &&
5171  root->append_rel_array != NULL)
5172  {
5173  AppendRelInfo *appinfo;
5174  Index varno = index->rel->relid;
5175 
5176  appinfo = root->append_rel_array[varno];
5177  while (appinfo &&
5178  planner_rt_fetch(appinfo->parent_relid,
5179  root)->rtekind == RTE_RELATION)
5180  {
5181  varno = appinfo->parent_relid;
5182  appinfo = root->append_rel_array[varno];
5183  }
5184  if (varno != index->rel->relid)
5185  {
5186  /* Repeat access check on this rel */
5187  rte = planner_rt_fetch(varno, root);
5188  Assert(rte->rtekind == RTE_RELATION);
5189 
5190  userid = rte->checkAsUser ? rte->checkAsUser : GetUserId();
5191 
5192  vardata->acl_ok =
5193  rte->securityQuals == NIL &&
5194  (pg_class_aclcheck(rte->relid,
5195  userid,
5196  ACL_SELECT) == ACLCHECK_OK);
5197  }
5198  }
5199  }
5200  else
5201  {
5202  /* suppress leakproofness checks later */
5203  vardata->acl_ok = true;
5204  }
5205  }
5206  if (vardata->statsTuple)
5207  break;
5208  }
5209  indexpr_item = lnext(index->indexprs, indexpr_item);
5210  }
5211  }
5212  if (vardata->statsTuple)
5213  break;
5214  }
5215 
5216  /*
5217  * Search extended statistics for one with a matching expression.
5218  * There might be multiple ones, so just grab the first one. In the
5219  * future, we might consider the statistics target (and pick the most
5220  * accurate statistics) and maybe some other parameters.
5221  */
5222  foreach(slist, onerel->statlist)
5223  {
5224  StatisticExtInfo *info = (StatisticExtInfo *) lfirst(slist);
5225  ListCell *expr_item;
5226  int pos;
5227 
5228  /*
5229  * Stop once we've found statistics for the expression (either
5230  * from extended stats, or for an index in the preceding loop).
5231  */
5232  if (vardata->statsTuple)
5233  break;
5234 
5235  /* skip stats without per-expression stats */
5236  if (info->kind != STATS_EXT_EXPRESSIONS)
5237  continue;
5238 
5239  pos = 0;
5240  foreach(expr_item, info->exprs)
5241  {
5242  Node *expr = (Node *) lfirst(expr_item);
5243 
5244  Assert(expr);
5245 
5246  /* strip RelabelType before comparing it */
5247  if (expr && IsA(expr, RelabelType))
5248  expr = (Node *) ((RelabelType *) expr)->arg;
5249 
5250  /* found a match, see if we can extract pg_statistic row */
5251  if (equal(node, expr))
5252  {
5253  HeapTuple t = statext_expressions_load(info->statOid, pos);
5254 
5255  /* Get index's table for permission check */
5256  RangeTblEntry *rte;
5257  Oid userid;
5258 
5259  vardata->statsTuple = t;
5260 
5261  /*
5262  * XXX Not sure if we should cache the tuple somewhere.
5263  * Now we just create a new copy every time.
5264  */
5265  vardata->freefunc = ReleaseDummy;
5266 
5267  rte = planner_rt_fetch(onerel->relid, root);
5268  Assert(rte->rtekind == RTE_RELATION);
5269 
5270  /*
5271  * Use checkAsUser if it's set, in case we're accessing
5272  * the table via a view.
5273  */
5274  userid = rte->checkAsUser ? rte->checkAsUser : GetUserId();
5275 
5276  /*
5277  * For simplicity, we insist on the whole table being
5278  * selectable, rather than trying to identify which
5279  * column(s) the statistics depends on. Also require all
5280  * rows to be selectable --- there must be no
5281  * securityQuals from security barrier views or RLS
5282  * policies.
5283  */
5284  vardata->acl_ok =
5285  rte->securityQuals == NIL &&
5286  (pg_class_aclcheck(rte->relid, userid,
5287  ACL_SELECT) == ACLCHECK_OK);
5288 
5289  /*
5290  * If the user doesn't have permissions to access an
5291  * inheritance child relation, check the permissions of
5292  * the table actually mentioned in the query, since most
5293  * likely the user does have that permission. Note that
5294  * whole-table select privilege on the parent doesn't
5295  * quite guarantee that the user could read all columns of
5296  * the child. But in practice it's unlikely that any
5297  * interesting security violation could result from
5298  * allowing access to the expression stats, so we allow it
5299  * anyway. See similar code in examine_simple_variable()
5300  * for additional comments.
5301  */
5302  if (!vardata->acl_ok &&
5303  root->append_rel_array != NULL)
5304  {
5305  AppendRelInfo *appinfo;
5306  Index varno = onerel->relid;
5307 
5308  appinfo = root->append_rel_array[varno];
5309  while (appinfo &&
5310  planner_rt_fetch(appinfo->parent_relid,
5311  root)->rtekind == RTE_RELATION)
5312  {
5313  varno = appinfo->parent_relid;
5314  appinfo = root->append_rel_array[varno];
5315  }
5316  if (varno != onerel->relid)
5317  {
5318  /* Repeat access check on this rel */
5319  rte = planner_rt_fetch(varno, root);
5320  Assert(rte->rtekind == RTE_RELATION);
5321 
5322  userid = rte->checkAsUser ? rte->checkAsUser : GetUserId();
5323 
5324  vardata->acl_ok =
5325  rte->securityQuals == NIL &&
5326  (pg_class_aclcheck(rte->relid,
5327  userid,
5328  ACL_SELECT) == ACLCHECK_OK);
5329  }
5330  }
5331 
5332  break;
5333  }
5334 
5335  pos++;
5336  }
5337  }
5338  }
5339 }
5340 
5341 /*
5342  * examine_simple_variable
5343  * Handle a simple Var for examine_variable
5344  *
5345  * This is split out as a subroutine so that we can recurse to deal with
5346  * Vars referencing subqueries.
5347  *
5348  * We already filled in all the fields of *vardata except for the stats tuple.
5349  */
5350 static void
5351 examine_simple_variable(PlannerInfo *root, Var *var,
5352  VariableStatData *vardata)
5353 {
5354  RangeTblEntry *rte = root->simple_rte_array[var->varno];
5355 
5356  Assert(IsA(rte, RangeTblEntry));
5357 
5358  if (get_relation_stats_hook &&
5359  (*get_relation_stats_hook) (root, rte, var->varattno, vardata))
5360  {
5361  /*
5362  * The hook took control of acquiring a stats tuple. If it did supply
5363  * a tuple, it'd better have supplied a freefunc.
5364  */
5365  if (HeapTupleIsValid(vardata->statsTuple) &&
5366  !vardata->freefunc)
5367  elog(ERROR, "no function provided to release variable stats with");
5368  }
5369  else if (rte->rtekind == RTE_RELATION)
5370  {
5371  /*
5372  * Plain table or parent of an inheritance appendrel, so look up the
5373  * column in pg_statistic
5374  */
5375  vardata->statsTuple = SearchSysCache3(STATRELATTINH,
5376  ObjectIdGetDatum(rte->relid),
5377  Int16GetDatum(var->varattno),
5378  BoolGetDatum(rte->inh));
5379  vardata->freefunc = ReleaseSysCache;
5380 
5381  if (HeapTupleIsValid(vardata->statsTuple))
5382  {
5383  Oid userid;
5384 
5385  /*
5386  * Check if user has permission to read this column. We require
5387  * all rows to be accessible, so there must be no securityQuals
5388  * from security barrier views or RLS policies. Use checkAsUser
5389  * if it's set, in case we're accessing the table via a view.
5390  */
5391  userid = rte->checkAsUser ? rte->checkAsUser : GetUserId();
5392 
5393  vardata->acl_ok =
5394  rte->securityQuals == NIL &&
5395  ((pg_class_aclcheck(rte->relid, userid,
5396  ACL_SELECT) == ACLCHECK_OK) ||
5397  (pg_attribute_aclcheck(rte->relid, var->varattno, userid,
5398  ACL_SELECT) == ACLCHECK_OK));
5399 
5400  /*
5401  * If the user doesn't have permissions to access an inheritance
5402  * child relation or specifically this attribute, check the
5403  * permissions of the table/column actually mentioned in the
5404  * query, since most likely the user does have that permission
5405  * (else the query will fail at runtime), and if the user can read
5406  * the column there then he can get the values of the child table
5407  * too. To do that, we must find out which of the root parent's
5408  * attributes the child relation's attribute corresponds to.
5409  */
5410  if (!vardata->acl_ok && var->varattno > 0 &&
5411  root->append_rel_array != NULL)
5412  {
5413  AppendRelInfo *appinfo;
5414  Index varno = var->varno;
5415  int varattno = var->varattno;
5416  bool found = false;
5417 
5418  appinfo = root->append_rel_array[varno];
5419 
5420  /*
5421  * Partitions are mapped to their immediate parent, not the
5422  * root parent, so must be ready to walk up multiple
5423  * AppendRelInfos. But stop if we hit a parent that is not
5424  * RTE_RELATION --- that's a flattened UNION ALL subquery, not
5425  * an inheritance parent.
5426  */
5427  while (appinfo &&
5428  planner_rt_fetch(appinfo->parent_relid,
5429  root)->rtekind == RTE_RELATION)
5430  {
5431  int parent_varattno;
5432 
5433  found = false;
5434  if (varattno <= 0 || varattno > appinfo->num_child_cols)
5435  break; /* safety check */
5436  parent_varattno = appinfo->parent_colnos[varattno - 1];
5437  if (parent_varattno == 0)
5438  break; /* Var is local to child */
5439 
5440  varno = appinfo->parent_relid;
5441  varattno = parent_varattno;
5442  found = true;
5443 
5444  /* If the parent is itself a child, continue up. */
5445  appinfo = root->append_rel_array[varno];
5446  }
5447 
5448  /*
5449  * In rare cases, the Var may be local to the child table, in
5450  * which case, we've got to live with having no access to this
5451  * column's stats.
5452  */
5453  if (!found)
5454  return;
5455 
5456  /* Repeat the access check on this parent rel & column */
5457  rte = planner_rt_fetch(varno, root);
5458  Assert(rte->rtekind == RTE_RELATION);
5459 
5460  userid = rte->checkAsUser ? rte->checkAsUser : GetUserId();
5461 
5462  vardata->acl_ok =
5463  rte->securityQuals == NIL &&
5464  ((pg_class_aclcheck(rte->relid, userid,
5465  ACL_SELECT) == ACLCHECK_OK) ||
5466  (pg_attribute_aclcheck(rte->relid, varattno, userid,
5467  ACL_SELECT) == ACLCHECK_OK));
5468  }
5469  }
5470  else
5471  {
5472  /* suppress any possible leakproofness checks later */
5473  vardata->acl_ok = true;
5474  }
5475  }
5476  else if (rte->rtekind == RTE_SUBQUERY && !rte->inh)
5477  {
5478  /*
5479  * Plain subquery (not one that was converted to an appendrel).
5480  */
5481  Query *subquery = rte->subquery;
5482  RelOptInfo *rel;
5483  TargetEntry *ste;
5484 
5485  /*
5486  * Punt if it's a whole-row var rather than a plain column reference.
5487  */
5488  if (var->varattno == InvalidAttrNumber)
5489  return;
5490 
5491  /*
5492  * Punt if subquery uses set operations or GROUP BY, as these will
5493  * mash underlying columns' stats beyond recognition. (Set ops are
5494  * particularly nasty; if we forged ahead, we would return stats
5495  * relevant to only the leftmost subselect...) DISTINCT is also
5496  * problematic, but we check that later because there is a possibility
5497  * of learning something even with it.
5498  */
5499  if (subquery->setOperations ||
5500  subquery->groupClause ||
5501  subquery->groupingSets)
5502  return;
5503 
5504  /*
5505  * OK, fetch RelOptInfo for subquery. Note that we don't change the
5506  * rel returned in vardata, since caller expects it to be a rel of the
5507  * caller's query level. Because we might already be recursing, we
5508  * can't use that rel pointer either, but have to look up the Var's
5509  * rel afresh.
5510  */
5511  rel = find_base_rel(root, var->varno);
5512 
5513  /* If the subquery hasn't been planned yet, we have to punt */
5514  if (rel->subroot == NULL)
5515  return;
5516  Assert(IsA(rel->subroot, PlannerInfo));
5517 
5518  /*
5519  * Switch our attention to the subquery as mangled by the planner. It
5520  * was okay to look at the pre-planning version for the tests above,
5521  * but now we need a Var that will refer to the subroot's live
5522  * RelOptInfos. For instance, if any subquery pullup happened during
5523  * planning, Vars in the targetlist might have gotten replaced, and we
5524  * need to see the replacement expressions.
5525  */
5526  subquery = rel->subroot->parse;
5527  Assert(IsA(subquery, Query));
5528 
5529  /* Get the subquery output expression referenced by the upper Var */
5530  ste = get_tle_by_resno(subquery->targetList, var->varattno);
5531  if (ste == NULL || ste->resjunk)
5532  elog(ERROR, "subquery %s does not have attribute %d",
5533  rte->eref->aliasname, var->varattno);
5534  var = (Var *) ste->expr;
5535 
5536  /*
5537  * If subquery uses DISTINCT, we can't make use of any stats for the
5538  * variable ... but, if it's the only DISTINCT column, we are entitled
5539  * to consider it unique. We do the test this way so that it works
5540  * for cases involving DISTINCT ON.
5541  */
5542  if (subquery->distinctClause)
5543  {
5544  if (list_length(subquery->distinctClause) == 1 &&
5545  targetIsInSortList(ste, InvalidOid, subquery->distinctClause))
5546  vardata->isunique = true;
5547  /* cannot go further */
5548  return;
5549  }
5550 
5551  /*
5552  * If the sub-query originated from a view with the security_barrier
5553  * attribute, we must not look at the variable's statistics, though it
5554  * seems all right to notice the existence of a DISTINCT clause. So
5555  * stop here.
5556  *
5557  * This is probably a harsher restriction than necessary; it's
5558  * certainly OK for the selectivity estimator (which is a C function,
5559  * and therefore omnipotent anyway) to look at the statistics. But
5560  * many selectivity estimators will happily *invoke the operator
5561  * function* to try to work out a good estimate - and that's not OK.
5562  * So for now, don't dig down for stats.
5563  */
5564  if (rte->security_barrier)
5565  return;
5566 
5567  /* Can only handle a simple Var of subquery's query level */
5568  if (var && IsA(var, Var) &&
5569  var->varlevelsup == 0)
5570  {
5571  /*
5572  * OK, recurse into the subquery. Note that the original setting
5573  * of vardata->isunique (which will surely be false) is left
5574  * unchanged in this situation. That's what we want, since even
5575  * if the underlying column is unique, the subquery may have
5576  * joined to other tables in a way that creates duplicates.
5577  */
5578  examine_simple_variable(rel->subroot, var, vardata);
5579  }
5580  }
5581  else
5582  {
5583  /*
5584  * Otherwise, the Var comes from a FUNCTION, VALUES, or CTE RTE. (We
5585  * won't see RTE_JOIN here because join alias Vars have already been
5586  * flattened.) There's not much we can do with function outputs, but
5587  * maybe someday try to be smarter about VALUES and/or CTEs.
5588  */
5589  }
5590 }
5591 
5592 /*
5593  * Check whether it is permitted to call func_oid passing some of the
5594  * pg_statistic data in vardata. We allow this either if the user has SELECT
5595  * privileges on the table or column underlying the pg_statistic data or if
5596  * the function is marked leak-proof.
5597  */
5598 bool
5599 statistic_proc_security_check(VariableStatData *vardata, Oid func_oid)
5600 {
5601  if (vardata->acl_ok)
5602  return true;
5603 
5604  if (!OidIsValid(func_oid))
5605  return false;
5606 
5607  if (get_func_leakproof(func_oid))
5608  return true;
5609 
5610  ereport(DEBUG2,
5611  (errmsg_internal("not using statistics because function \"%s\" is not leak-proof",
5612  get_func_name(func_oid))));
5613  return false;
5614 }
5615 
5616 /*
5617  * get_variable_numdistinct
5618  * Estimate the number of distinct values of a variable.
5619  *
5620  * vardata: results of examine_variable
5621  * *isdefault: set to true if the result is a default rather than based on
5622  * anything meaningful.
5623  *
5624  * NB: be careful to produce a positive integral result, since callers may
5625  * compare the result to exact integer counts, or might divide by it.
5626  */
5627 double
5628 get_variable_numdistinct(VariableStatData *vardata, bool *isdefault)
5629 {
5630  double stadistinct;
5631  double stanullfrac = 0.0;
5632  double ntuples;
5633 
5634  *isdefault = false;
5635 
5636  /*
5637  * Determine the stadistinct value to use. There are cases where we can
5638  * get an estimate even without a pg_statistic entry, or can get a better
5639  * value than is in pg_statistic. Grab stanullfrac too if we can find it
5640  * (otherwise, assume no nulls, for lack of any better idea).
5641  */
5642  if (HeapTupleIsValid(vardata->statsTuple))
5643  {
5644  /* Use the pg_statistic entry */
5645  Form_pg_statistic stats;
5646 
5647  stats = (Form_pg_statistic) GETSTRUCT(vardata->statsTuple);
5648  stadistinct = stats->stadistinct;
5649  stanullfrac = stats->stanullfrac;
5650  }
5651  else if (vardata->vartype == BOOLOID)
5652  {
5653  /*
5654  * Special-case boolean columns: presumably, two distinct values.
5655  *
5656  * Are there any other datatypes we should wire in special estimates
5657  * for?
5658  */
5659  stadistinct = 2.0;
5660  }
5661  else if (vardata->rel && vardata->rel->rtekind == RTE_VALUES)
5662  {
5663  /*
5664  * If the Var represents a column of a VALUES RTE, assume it's unique.
5665  * This could of course be very wrong, but it should tend to be true
5666  * in well-written queries. We could consider examining the VALUES'
5667  * contents to get some real statistics; but that only works if the
5668  * entries are all constants, and it would be pretty expensive anyway.
5669  */
5670  stadistinct = -1.0; /* unique (and all non null) */
5671  }
5672  else
5673  {
5674  /*
5675  * We don't keep statistics for system columns, but in some cases we
5676  * can infer distinctness anyway.
5677  */
5678  if (vardata->var && IsA(vardata->var, Var))
5679  {
5680  switch (((Var *) vardata->var)->varattno)
5681  {
5682  case SelfItemPointerAttributeNumber:
5683  stadistinct = -1.0; /* unique (and all non null) */
5684  break;
5685  case TableOidAttributeNumber:
5686  stadistinct = 1.0; /* only 1 value */
5687  break;
5688  default:
5689  stadistinct = 0.0; /* means "unknown" */
5690  break;
5691  }
5692  }
5693  else
5694  stadistinct = 0.0; /* means "unknown" */
5695 
5696  /*
5697  * XXX consider using estimate_num_groups on expressions?
5698  */
5699  }
5700 
5701  /*
5702  * If there is a unique index or DISTINCT clause for the variable, assume
5703  * it is unique no matter what pg_statistic says; the statistics could be
5704  * out of date, or we might have found a partial unique index that proves
5705  * the var is unique for this query. However, we'd better still believe
5706  * the null-fraction statistic.
5707  */
5708  if (vardata->isunique)
5709  stadistinct = -1.0 * (1.0 - stanullfrac);
5710 
5711  /*
5712  * If we had an absolute estimate, use that.
5713  */
5714  if (stadistinct > 0.0)
5715  return clamp_row_est(stadistinct);
5716 
5717  /*
5718  * Otherwise we need to get the relation size; punt if not available.
5719  */
5720  if (vardata->rel == NULL)
5721  {
5722  *isdefault = true;
5723  return DEFAULT_NUM_DISTINCT;
5724  }
5725  ntuples = vardata->rel->tuples;
5726  if (ntuples <= 0.0)
5727  {
5728  *isdefault = true;
5729  return DEFAULT_NUM_DISTINCT;
5730  }
5731 
5732  /*
5733  * If we had a relative estimate, use that.
5734  */
5735  if (stadistinct < 0.0)
5736  return clamp_row_est(-stadistinct * ntuples);
5737 
5738  /*
5739  * With no data, estimate ndistinct = ntuples if the table is small, else
5740  * use default. We use DEFAULT_NUM_DISTINCT as the cutoff for "small" so
5741  * that the behavior isn't discontinuous.
5742  */
5743  if (ntuples < DEFAULT_NUM_DISTINCT)
5744  return clamp_row_est(ntuples);
5745 
5746  *isdefault = true;
5747  return DEFAULT_NUM_DISTINCT;
5748 }
5749 
5750 /*
5751  * get_variable_range
5752  * Estimate the minimum and maximum value of the specified variable.
5753  * If successful, store values in *min and *max, and return true.
5754  * If no data available, return false.
5755  *
5756  * sortop is the "<" comparison operator to use. This should generally
5757  * be "<" not ">", as only the former is likely to be found in pg_statistic.
5758  * The collation must be specified too.
5759  */
5760 static bool
5761 get_variable_range(PlannerInfo *root, VariableStatData *vardata,
5762  Oid sortop, Oid collation,
5763  Datum *min, Datum *max)
5764 {
5765  Datum tmin = 0;
5766  Datum tmax = 0;
5767  bool have_data = false;
5768  int16 typLen;
5769  bool typByVal;
5770  Oid opfuncoid;
5771  FmgrInfo opproc;
5772  AttStatsSlot sslot;
5773 
5774  /*
5775  * XXX It's very tempting to try to use the actual column min and max, if
5776  * we can get them relatively-cheaply with an index probe. However, since
5777  * this function is called many times during join planning, that could
5778  * have unpleasant effects on planning speed. Need more investigation
5779  * before enabling this.
5780  */
5781 #ifdef NOT_USED
5782  if (get_actual_variable_range(root, vardata, sortop, collation, min, max))
5783  return true;
5784 #endif
5785 
5786  if (!HeapTupleIsValid(vardata->statsTuple))
5787  {
5788  /* no stats available, so default result */
5789  return false;
5790  }
5791 
5792  /*
5793  * If we can't apply the sortop to the stats data, just fail. In
5794  * principle, if there's a histogram and no MCVs, we could return the
5795  * histogram endpoints without ever applying the sortop ... but it's
5796  * probably not worth trying, because whatever the caller wants to do with
5797  * the endpoints would likely fail the security check too.
5798  */
5799  if (!statistic_proc_security_check(vardata,
5800  (opfuncoid = get_opcode(sortop))))
5801  return false;
5802 
5803  opproc.fn_oid = InvalidOid; /* mark this as not looked up yet */
5804 
5805  get_typlenbyval(vardata->atttype, &typLen, &typByVal);
5806 
5807  /*
5808  * If there is a histogram with the ordering we want, grab the first and
5809  * last values.
5810  */
5811  if (get_attstatsslot(&sslot, vardata->statsTuple,
5812  STATISTIC_KIND_HISTOGRAM, sortop,
5813  ATTSTATSSLOT_VALUES))
5814  {
5815  if (sslot.stacoll == collation && sslot.nvalues > 0)
5816  {
5817  tmin = datumCopy(sslot.values[0], typByVal, typLen);
5818  tmax = datumCopy(sslot.values[sslot.nvalues - 1], typByVal, typLen);
5819  have_data = true;
5820  }
5821  free_attstatsslot(&sslot);
5822  }
5823 
5824  /*
5825  * Otherwise, if there is a histogram with some other ordering, scan it
5826  * and get the min and max values according to the ordering we want. This
5827  * of course may not find values that are really extremal according to our
5828  * ordering, but it beats ignoring available data.
5829  */
5830  if (!have_data &&
5831  get_attstatsslot(&sslot, vardata->statsTuple,
5832  STATISTIC_KIND_HISTOGRAM, InvalidOid,
5833  ATTSTATSSLOT_VALUES))
5834  {
5835  get_stats_slot_range(&sslot, opfuncoid, &opproc,
5836  collation, typLen, typByVal,
5837  &tmin, &tmax, &have_data);
5838  free_attstatsslot(&sslot);
5839  }
5840 
5841  /*
5842  * If we have most-common-values info, look for extreme MCVs. This is
5843  * needed even if we also have a histogram, since the histogram excludes
5844  * the MCVs.
5845  */
5846  if (get_attstatsslot(&sslot, vardata->statsTuple,
5847  STATISTIC_KIND_MCV, InvalidOid,
5848  ATTSTATSSLOT_VALUES))
5849  {
5850  get_stats_slot_range(&sslot, opfuncoid, &opproc,
5851  collation, typLen, typByVal,
5852  &tmin, &tmax, &have_data);
5853  free_attstatsslot(&sslot);
5854  }
5855 
5856  *min = tmin;
5857  *max = tmax;
5858  return have_data;
5859 }
5860 
5861 /*
5862  * get_stats_slot_range: scan sslot for min/max values
5863  *
5864  * Subroutine for get_variable_range: update min/max/have_data according
5865  * to what we find in the statistics array.
5866  */
5867 static void
5868 get_stats_slot_range(AttStatsSlot *sslot, Oid opfuncoid, FmgrInfo *opproc,
5869  Oid collation, int16 typLen, bool typByVal,
5870  Datum *min, Datum *max, bool *p_have_data)
5871 {
5872  Datum tmin = *min;
5873  Datum tmax = *max;
5874  bool have_data = *p_have_data;
5875  bool found_tmin = false;
5876  bool found_tmax = false;
5877 
5878  /* Look up the comparison function, if we didn't already do so */
5879  if (opproc->fn_oid != opfuncoid)
5880  fmgr_info(opfuncoid, opproc);
5881 
5882  /* Scan all the slot's values */
5883  for (int i = 0; i < sslot->nvalues; i++)
5884  {
5885  if (!have_data)
5886  {
5887  tmin = tmax = sslot->values[i];
5888  found_tmin = found_tmax = true;
5889  *p_have_data = have_data = true;
5890  continue;
5891  }
5892  if (DatumGetBool(FunctionCall2Coll(opproc,
5893  collation,
5894  sslot->values[i], tmin)))
5895  {
5896  tmin = sslot->values[i];
5897  found_tmin = true;
5898  }
5899  if (DatumGetBool(FunctionCall2Coll(opproc,
5900  collation,
5901  tmax, sslot->values[i])))
5902  {
5903  tmax = sslot->values[i];
5904  found_tmax = true;
5905  }
5906  }
5907 
5908  /*
5909  * Copy the slot's values, if we found new extreme values.
5910  */
5911  if (found_tmin)
5912  *min = datumCopy(tmin, typByVal, typLen);
5913  if (found_tmax)
5914  *max = datumCopy(tmax, typByVal, typLen);
5915 }
5916 
5917 
5918 /*
5919  * get_actual_variable_range
5920  * Attempt to identify the current *actual* minimum and/or maximum
5921  * of the specified variable, by looking for a suitable btree index
5922  * and fetching its low and/or high values.
5923  * If successful, store values in *min and *max, and return true.
5924  * (Either pointer can be NULL if that endpoint isn't needed.)
5925  * If no data available, return false.
5926  *
5927  * sortop is the "<" comparison operator to use.
5928  * collation is the required collation.
5929  */
5930 static bool
5931 get_actual_variable_range(PlannerInfo *root, VariableStatData *vardata,
5932  Oid sortop, Oid collation,
5933  Datum *min, Datum *max)
5934 {
5935  bool have_data = false;
5936  RelOptInfo *rel = vardata->rel;
5937  RangeTblEntry *rte;
5938  ListCell *lc;
5939 
5940  /* No hope if no relation or it doesn't have indexes */
5941  if (rel == NULL || rel->indexlist == NIL)
5942  return false;
5943  /* If it has indexes it must be a plain relation */
5944  rte = root->simple_rte_array[rel->relid];
5945  Assert(rte->rtekind == RTE_RELATION);
5946 
5947  /* Search through the indexes to see if any match our problem */
5948  foreach(lc, rel->indexlist)
5949  {
5950  IndexOptInfo *index = (IndexOptInfo *) lfirst(lc);
5951  ScanDirection indexscandir;
5952 
5953  /* Ignore non-btree indexes */
5954  if (index->relam != BTREE_AM_OID)
5955  continue;
5956 
5957  /*
5958  * Ignore partial indexes --- we only want stats that cover the entire
5959  * relation.
5960  */
5961  if (index->indpred != NIL)
5962  continue;
5963 
5964  /*
5965  * The index list might include hypothetical indexes inserted by a
5966  * get_relation_info hook --- don't try to access them.
5967  */
5968  if (index->hypothetical)
5969  continue;
5970 
5971  /*
5972  * The first index column must match the desired variable, sortop, and
5973  * collation --- but we can use a descending-order index.
5974  */
5975  if (collation != index->indexcollations[0])
5976  continue; /* test first 'cause it's cheapest */
5977  if (!match_index_to_operand(vardata->var, 0, index))
5978  continue;
5979  switch (get_op_opfamily_strategy(sortop, index->sortopfamily[0]))
5980  {
5981  case BTLessStrategyNumber:
5982  if (index->reverse_sort[0])
5983  indexscandir = BackwardScanDirection;
5984  else
5985  indexscandir = ForwardScanDirection;
5986  break;
5987  case BTGreaterStrategyNumber:
5988  if (index->reverse_sort[0])
5989  indexscandir = ForwardScanDirection;
5990  else
5991  indexscandir = BackwardScanDirection;
5992  break;
5993  default:
5994  /* index doesn't match the sortop */
5995  continue;
5996  }
5997 
5998  /*
5999  * Found a suitable index to extract data from. Set up some data that
6000  * can be used by both invocations of get_actual_variable_endpoint.
6001  */
6002  {
6003  MemoryContext tmpcontext;
6004  MemoryContext oldcontext;
6005  Relation heapRel;
6006  Relation indexRel;
6007  TupleTableSlot *slot;
6008  int16 typLen;
6009  bool typByVal;
6010  ScanKeyData scankeys[1];
6011 
6012  /* Make sure any cruft gets recycled when we're done */
6013  tmpcontext = AllocSetContextCreate(CurrentMemoryContext,
6014  "get_actual_variable_range workspace",
6015  ALLOCSET_DEFAULT_SIZES);
6016  oldcontext = MemoryContextSwitchTo(tmpcontext);
6017 
6018  /*
6019  * Open the table and index so we can read from them. We should
6020  * already have some type of lock on each.
6021  */
6022  heapRel = table_open(rte->relid, NoLock);
6023  indexRel = index_open(index->indexoid, NoLock);
6024 
6025  /* build some stuff needed for indexscan execution */
6026  slot = table_slot_create(heapRel, NULL);
6027  get_typlenbyval(vardata->atttype, &typLen, &typByVal);
6028 
6029  /* set up an IS NOT NULL scan key so that we ignore nulls */
6030  ScanKeyEntryInitialize(&scankeys[0],
6031  SK_ISNULL | SK_SEARCHNOTNULL,
6032  1, /* index col to scan */
6033  InvalidStrategy, /* no strategy */
6034  InvalidOid, /* no strategy subtype */
6035  InvalidOid, /* no collation */
6036  InvalidOid, /* no reg proc for this */
6037  (Datum) 0); /* constant */
6038 
6039  /* If min is requested ... */
6040  if (min)
6041  {
6042  have_data = get_actual_variable_endpoint(heapRel,
6043  indexRel,
6044  indexscandir,
6045  scankeys,
6046  typLen,
6047  typByVal,
6048  slot,
6049  oldcontext,
6050  min);
6051  }
6052  else
6053  {
6054  /* If min not requested, assume index is nonempty */
6055  have_data = true;
6056  }
6057 
6058  /* If max is requested, and we didn't find the index is empty */
6059  if (max && have_data)
6060  {
6061  /* scan in the opposite direction; all else is the same */
6062  have_data = get_actual_variable_endpoint(heapRel,
6063  indexRel,
6064  -indexscandir,
6065  scankeys,
6066  typLen,
6067  typByVal,
6068  slot,
6069  oldcontext,
6070  max);
6071  }
6072 
6073  /* Clean everything up */
6074  ExecDropSingleTupleTableSlot(slot);
6075 
6076  index_close(indexRel, NoLock);
6077  table_close(heapRel, NoLock);
6078 
6079  MemoryContextSwitchTo(oldcontext);
6080  MemoryContextDelete(tmpcontext);
6081 
6082  /* And we're done */
6083  break;
6084  }
6085  }
6086 
6087  return have_data;
6088 }
6089 
6090 /*
6091  * Get one endpoint datum (min or max depending on indexscandir) from the
6092  * specified index. Return true if successful, false if index is empty.
6093  * On success, endpoint value is stored to *endpointDatum (and copied into
6094  * outercontext).
6095  *
6096  * scankeys is a 1-element scankey array set up to reject nulls.
6097  * typLen/typByVal describe the datatype of the index's first column.
6098  * tableslot is a slot suitable to hold table tuples, in case we need
6099  * to probe the heap.
6100  * (We could compute these values locally, but that would mean computing them
6101  * twice when get_actual_variable_range needs both the min and the max.)
6102  */
6103 static bool
6104 get_actual_variable_endpoint(Relation heapRel,
6105  Relation indexRel,
6106  ScanDirection indexscandir,
6107  ScanKey scankeys,
6108  int16 typLen,
6109  bool typByVal,
6110  TupleTableSlot *tableslot,
6111  MemoryContext outercontext,
6112  Datum *endpointDatum)
6113 {
6114  bool have_data = false;
6115  SnapshotData SnapshotNonVacuumable;
6116  IndexScanDesc index_scan;
6117  Buffer vmbuffer = InvalidBuffer;
6118  ItemPointer tid;
6119  Datum values[INDEX_MAX_KEYS];
6120  bool isnull[INDEX_MAX_KEYS];
6121  MemoryContext oldcontext;
6122 
6123  /*
6124  * We use the index-only-scan machinery for this. With mostly-static
6125  * tables that's a win because it avoids a heap visit. It's also a win
6126  * for dynamic data, but the reason is less obvious; read on for details.
6127  *
6128  * In principle, we should scan the index with our current active
6129  * snapshot, which is the best approximation we've got to what the query
6130  * will see when executed. But that won't be exact if a new snap is taken
6131  * before running the query, and it can be very expensive if a lot of
6132  * recently-dead or uncommitted rows exist at the beginning or end of the
6133  * index (because we'll laboriously fetch each one and reject it).
6134  * Instead, we use SnapshotNonVacuumable. That will accept recently-dead
6135  * and uncommitted rows as well as normal visible rows. On the other
6136  * hand, it will reject known-dead rows, and thus not give a bogus answer
6137  * when the extreme value has been deleted (unless the deletion was quite
6138  * recent); that case motivates not using SnapshotAny here.
6139  *
6140  * A crucial point here is that SnapshotNonVacuumable, with
6141  * GlobalVisTestFor(heapRel) as horizon, yields the inverse of the
6142  * condition that the indexscan will use to decide that index entries are
6143  * killable (see heap_hot_search_buffer()). Therefore, if the snapshot
6144  * rejects a tuple (or more precisely, all tuples of a HOT chain) and we
6145  * have to continue scanning past it, we know that the indexscan will mark
6146  * that index entry killed. That means that the next
6147  * get_actual_variable_endpoint() call will not have to re-consider that
6148  * index entry. In this way we avoid repetitive work when this function
6149  * is used a lot during planning.
6150  *
6151  * But using SnapshotNonVacuumable creates a hazard of its own. In a
6152  * recently-created index, some index entries may point at "broken" HOT
6153  * chains in which not all the tuple versions contain data matching the
6154  * index entry. The live tuple version(s) certainly do match the index,
6155  * but SnapshotNonVacuumable can accept recently-dead tuple versions that
6156  * don't match. Hence, if we took data from the selected heap tuple, we
6157  * might get a bogus answer that's not close to the index extremal value,
6158  * or could even be NULL. We avoid this hazard because we take the data
6159  * from the index entry not the heap.
6160  */
6161  InitNonVacuumableSnapshot(SnapshotNonVacuumable,
6162  GlobalVisTestFor(heapRel));
6163 
6164  index_scan = index_beginscan(heapRel, indexRel,
6165  &SnapshotNonVacuumable,
6166  1, 0);
6167  /* Set it up for index-only scan */
6168  index_scan->xs_want_itup = true;
6169  index_rescan(index_scan, scankeys, 1, NULL, 0);
6170 
6171  /* Fetch first/next tuple in specified direction */
6172  while ((tid = index_getnext_tid(index_scan, indexscandir)) != NULL)
6173  {
6174  if (!VM_ALL_VISIBLE(heapRel,
6175  ItemPointerGetBlockNumber(tid),
6176  &vmbuffer))
6177  {
6178  /* Rats, we have to visit the heap to check visibility */
6179  if (!index_fetch_heap(index_scan, tableslot))
6180  continue; /* no visible tuple, try next index entry */
6181 
6182  /* We don't actually need the heap tuple for anything */