Don't assume a subquery's output is unique if there's a SRF in its tlist

Author: Tom Lane <tgl@sss.pgh.pa.us>
Date:   Tue Jul 8 14:03:32 2014 -0400

    While the x output of "select x from t group by x" can be presumed unique,
    this does not hold for "select x, generate_series(1,10) from t group by x",
    because we may expand the set-returning function after the grouping step.
    (Perhaps that should be re-thought; but considering all the other oddities
    involved with SRFs in targetlists, it seems unlikely we'll change it.)
    Put a check in query_is_distinct_for() so it's not fooled by such cases.

    Back-patch to all supported branches.

    David Rowley

(cherry picked from commit 2e7469dc8b3bac4fe0f9bd042aaf802132efde85)

src/backend/optimizer/util/pathnode.c

@@ -23,6 +23,7 @@
 #include "executor/executor.h"
 #include "miscadmin.h"
 #include "optimizer/clauses.h"              /* contain_mutable_functions() */
+#include "nodes/nodeFuncs.h"
 #include "optimizer/cost.h"
 #include "optimizer/pathnode.h"
 #include "optimizer/paths.h"
@@ -2147,6 +2148,17 @@ query_is_distinct_for(Query *query, List *colnos, List *opids)
 
 	Assert(list_length(colnos) == list_length(opids));
 
+	/*
+	 * A set-returning function in the query's targetlist can result in
+	 * returning duplicate rows, if the SRF is evaluated after the
+	 * de-duplication step; so we play it safe and say "no" if there are any
+	 * SRFs.  (We could be certain that it's okay if SRFs appear only in the
+	 * specified columns, since those must be evaluated before de-duplication;
+	 * but it doesn't presently seem worth the complication to check that.)
+	 */
+	if (expression_returns_set((Node *) query->targetList))
+		return false;
+
 	/*
 	 * DISTINCT (including DISTINCT ON) guarantees uniqueness if all the
 	 * columns in the DISTINCT clause appear in colnos and operator semantics

src/test/regress/expected/subselect.out

@@ -640,3 +640,31 @@ select * from int4_tbl where
   0
 (1 row)
 
+--
+-- Check for incorrect optimization when IN subquery contains a SRF
+--
+set enable_hashjoin to 0;
+explain select * from int4_tbl o where (f1, f1) in
+  (select f1, generate_series(1,2) / 10 g from int4_tbl i group by f1);
+                                            QUERY PLAN                                             
+---------------------------------------------------------------------------------------------------
+ Gather Motion 3:1  (slice2; segments: 3)  (cost=1.02..2.03 rows=1 width=4)
+   ->  Seq Scan on int4_tbl o  (cost=1.02..2.03 rows=1 width=4)
+         Filter: (hashed SubPlan 1)
+         SubPlan 1
+           ->  Materialize  (cost=1.02..1.03 rows=1 width=4)
+                 ->  Broadcast Motion 3:3  (slice1; segments: 3)  (cost=0.00..1.01 rows=1 width=4)
+                       ->  Result  (cost=0.00..1.01 rows=1 width=4)
+                             ->  Seq Scan on int4_tbl i  (cost=0.00..1.01 rows=1 width=4)
+ Settings:  enable_hashjoin=off; optimizer=off
+ Optimizer status: legacy query optimizer
+(10 rows)
+
+select * from int4_tbl o where (f1, f1) in
+  (select f1, generate_series(1,2) / 10 g from int4_tbl i group by f1);
+ f1 
+----
+  0
+(1 row)
+
+reset enable_hashjoin;

src/test/regress/expected/subselect_optimizer.out

@@ -658,3 +658,31 @@ select * from int4_tbl where
   0
 (1 row)
 
+--
+-- Check for incorrect optimization when IN subquery contains a SRF
+--
+set enable_hashjoin to 0;
+explain select * from int4_tbl o where (f1, f1) in
+  (select f1, generate_series(1,2) / 10 g from int4_tbl i group by f1);
+                                            QUERY PLAN                                             
+---------------------------------------------------------------------------------------------------
+ Gather Motion 3:1  (slice2; segments: 3)  (cost=1.02..2.03 rows=1 width=4)
+   ->  Seq Scan on int4_tbl o  (cost=1.02..2.03 rows=1 width=4)
+         Filter: (hashed SubPlan 1)
+         SubPlan 1
+           ->  Materialize  (cost=1.02..1.03 rows=1 width=4)
+                 ->  Broadcast Motion 3:3  (slice1; segments: 3)  (cost=0.00..1.01 rows=1 width=4)
+                       ->  Result  (cost=0.00..1.01 rows=1 width=4)
+                             ->  Seq Scan on int4_tbl i  (cost=0.00..1.01 rows=1 width=4)
+ Settings:  enable_hashjoin=off; optimizer=on
+ Optimizer status: legacy query optimizer
+(10 rows)
+
+select * from int4_tbl o where (f1, f1) in
+  (select f1, generate_series(1,2) / 10 g from int4_tbl i group by f1);
+ f1 
+----
+  0
+(1 row)
+
+reset enable_hashjoin;

src/test/regress/sql/subselect.sql

@@ -383,3 +383,13 @@ select '1'::text in (select '1'::name union all select '1'::name);
 select * from int4_tbl where
   (case when f1 in (select unique1 from tenk1 a) then f1 else null end) in
   (select ten from tenk1 b);
+
+--
+-- Check for incorrect optimization when IN subquery contains a SRF
+--
+set enable_hashjoin to 0;
+explain select * from int4_tbl o where (f1, f1) in
+  (select f1, generate_series(1,2) / 10 g from int4_tbl i group by f1);
+select * from int4_tbl o where (f1, f1) in
+  (select f1, generate_series(1,2) / 10 g from int4_tbl i group by f1);
+reset enable_hashjoin;