First step in fixing selectivity-estimation code. eqsel and

neqsel now behave as per my suggestions in pghackers a few days ago.
selectivity for < > <= >= should work OK for integral types as well, but
still need work for nonintegral types.  Since these routines have never
actually executed before :-(, this may result in some significant changes
in the optimizer's choices of execution plans.  Let me know if you see
any serious misbehavior.
CAUTION: THESE CHANGES REQUIRE INITDB.  pg_statistic table has changed.

3 files changed, 44 insertions, 19 deletions

diff --git a/src/include/catalog/pg_statistic.h b/src/include/catalog/pg_statistic.h
index 19b87b68b1b..1c719443282 100644
--- a/src/include/catalog/pg_statistic.h
+++ b/src/include/catalog/pg_statistic.h
@@ -7,7 +7,7 @@
  *
  * Copyright (c) 1994, Regents of the University of California
  *
- * $Id: pg_statistic.h,v 1.6 1999/02/13 23:21:15 momjian Exp $
+ * $Id: pg_statistic.h,v 1.7 1999/08/01 04:54:21 tgl Exp $
  *
  * NOTES
  *	  the genbki.sh script reads this file and generates .bki
@@ -32,11 +32,32 @@
  */
 CATALOG(pg_statistic)
 {
-	Oid			starelid;
-	int2		staattnum;
-	Oid			staop;
-	text		stalokey;		/* VARIABLE LENGTH FIELD */
-	text		stahikey;		/* VARIABLE LENGTH FIELD */
+	/* These fields form the unique key for the entry: */
+	Oid			starelid;		/* relation containing attribute */
+	int2		staattnum;		/* attribute (column) stats are for */
+	Oid			staop;			/* '<' comparison op used for lo/hi vals */
+	/* Note: the current VACUUM code will never produce more than one entry
+	 * per column, but in theory there could be multiple entries if a datatype
+	 * has more than one useful ordering operator.  Also, the current code
+	 * will not write an entry unless it found at least one non-NULL value
+	 * in the column; so the remaining fields will never be NULL.
+	 */
+
+	/* These fields contain the stats about the column indicated by the key */
+	float4		stanullfrac;	/* the fraction of the entries that are NULL */
+	float4		stacommonfrac;	/* the fraction that are the most common val */
+
+	/* THE REST OF THESE ARE VARIABLE LENGTH FIELDS.
+	 * They cannot be accessed as C struct entries; you have to use the
+	 * full field access machinery (heap_getattr) for them.
+	 *
+	 * All three of these are text representations of data values of the
+	 * column's data type.  To re-create the actual Datum, do
+	 * datatypein(textout(givenvalue)).
+	 */
+	text		stacommonval;	/* most common non-null value in column */
+	text		staloval;		/* smallest non-null value in column */
+	text		stahival;		/* largest non-null value in column */
 } FormData_pg_statistic;
 
 /* ----------------
@@ -50,11 +71,14 @@ typedef FormData_pg_statistic *Form_pg_statistic;
  *		compiler constants for pg_statistic
  * ----------------
  */
-#define Natts_pg_statistic				5
+#define Natts_pg_statistic				8
 #define Anum_pg_statistic_starelid		1
 #define Anum_pg_statistic_staattnum		2
 #define Anum_pg_statistic_staop			3
-#define Anum_pg_statistic_stalokey		4
-#define Anum_pg_statistic_stahikey		5
+#define Anum_pg_statistic_stanullfrac	4
+#define Anum_pg_statistic_stacommonfrac	5
+#define Anum_pg_statistic_stacommonval	6
+#define Anum_pg_statistic_staloval		7
+#define Anum_pg_statistic_stahival		8
 
 #endif	 /* PG_STATISTIC_H */
diff --git a/src/include/commands/vacuum.h b/src/include/commands/vacuum.h
index 59a72bbb15b..59a7fe4a506 100644
--- a/src/include/commands/vacuum.h
+++ b/src/include/commands/vacuum.h
@@ -6,7 +6,7 @@
  *
  * Copyright (c) 1994, Regents of the University of California
  *
- * $Id: vacuum.h,v 1.22 1999/07/15 15:21:03 momjian Exp $
+ * $Id: vacuum.h,v 1.23 1999/08/01 04:54:25 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -67,22 +67,23 @@ typedef struct
 				guess2,
 				max,
 				min;
-	int16		best_len,
+	int			best_len,
 				guess1_len,
 				guess2_len,
 				max_len,
 				min_len;
-	int32		best_cnt,
+	long		best_cnt,
 				guess1_cnt,
 				guess1_hits,
 				guess2_hits,
 				null_cnt,
-				nonnull_cnt;
-	int32		max_cnt,
+				nonnull_cnt,
+				max_cnt,
 				min_cnt;
 	FmgrInfo	f_cmpeq,
 				f_cmplt,
 				f_cmpgt;
+	Oid			op_cmplt;
 	regproc		outfunc;
 	bool		initialized;
 } VacAttrStats;
diff --git a/src/include/utils/builtins.h b/src/include/utils/builtins.h
index e6a0b4157d5..dfe1897cbe2 100644
--- a/src/include/utils/builtins.h
+++ b/src/include/utils/builtins.h
@@ -6,7 +6,7 @@
  *
  * Copyright (c) 1994, Regents of the University of California
  *
- * $Id: builtins.h,v 1.84 1999/07/16 17:07:39 momjian Exp $
+ * $Id: builtins.h,v 1.85 1999/08/01 04:54:20 tgl Exp $
  *
  * NOTES
  *	  This should normally only be included by fmgr.h.
@@ -372,10 +372,10 @@ extern Oid	regproctooid(RegProcedure rp);
 #define RegprocToOid(rp) regproctooid(rp)
 
 /* selfuncs.c */
-extern float64 eqsel(Oid opid, Oid relid, AttrNumber attno, char *value, int32 flag);
-extern float64 neqsel(Oid opid, Oid relid, AttrNumber attno, char *value, int32 flag);
-extern float64 intltsel(Oid opid, Oid relid, AttrNumber attno, int32 value, int32 flag);
-extern float64 intgtsel(Oid opid, Oid relid, AttrNumber attno, int32 value, int32 flag);
+extern float64 eqsel(Oid opid, Oid relid, AttrNumber attno, Datum value, int32 flag);
+extern float64 neqsel(Oid opid, Oid relid, AttrNumber attno, Datum value, int32 flag);
+extern float64 intltsel(Oid opid, Oid relid, AttrNumber attno, Datum value, int32 flag);
+extern float64 intgtsel(Oid opid, Oid relid, AttrNumber attno, Datum value, int32 flag);
 extern float64 eqjoinsel(Oid opid, Oid relid1, AttrNumber attno1, Oid relid2, AttrNumber attno2);
 extern float64 neqjoinsel(Oid opid, Oid relid1, AttrNumber attno1, Oid relid2, AttrNumber attno2);
 extern float64 intltjoinsel(Oid opid, Oid relid1, AttrNumber attno1, Oid relid2, AttrNumber attno2);