Simplify FK-to-partitioned regression test query

Avoid a join between relations having the FK to detect FK violation.
The planner might optimize this considering the PK must exist on the
referenced side at some point, effectively masking a bug this test
tries to detect.

Tom Lane and Jehan-Guillaume de Rorthais
Discussion: https://postgr.es/m/467.1581270529@sss.pgh.pa.us

src/test/regress/expected/foreign_key.out

@@ -2456,12 +2456,14 @@ CREATE SCHEMA fkpart9
 INSERT INTO fkpart9.pk VALUES (35);
 INSERT INTO fkpart9.fk VALUES (35);
 DELETE FROM fkpart9.pk WHERE a=35;
-SELECT fk.fk_a, pk.a
-FROM fkpart9.fk
-LEFT JOIN fkpart9.pk ON fk.fk_a = pk.a
-WHERE fk.fk_a=35;
- fk_a | a 
-------+---
+SELECT * FROM fkpart9.pk;
+ a 
+---
+(0 rows)
+
+SELECT * FROM fkpart9.fk;
+ fk_a 
+------
 (0 rows)
 
 DROP SCHEMA fkpart9 CASCADE;

src/test/regress/sql/foreign_key.sql

@@ -1735,8 +1735,6 @@ CREATE SCHEMA fkpart9
 INSERT INTO fkpart9.pk VALUES (35);
 INSERT INTO fkpart9.fk VALUES (35);
 DELETE FROM fkpart9.pk WHERE a=35;
-SELECT fk.fk_a, pk.a
-FROM fkpart9.fk
-LEFT JOIN fkpart9.pk ON fk.fk_a = pk.a
-WHERE fk.fk_a=35;
+SELECT * FROM fkpart9.pk;
+SELECT * FROM fkpart9.fk;
 DROP SCHEMA fkpart9 CASCADE;