[YSQL] Re-evaluate the constant trimming optimization

[pao214]

Jira Link: DB-15447

Description

Re-evaluate the below optimziation. Do we gain much from the below optimization.

---

CREATE TABLE th(h1 INT, h2 INT, r1 INT, PRIMARY KEY(h1 HASH, h2 HASH, r1 ASC);

Query

EXPLAIN
SELECT DISTINCT h1, h2 FROM th WHERE r1 = 1;

has a prefix length of 3

                                  QUERY PLAN                                  
------------------------------------------------------------------------------
 Distinct Index Scan using th_pkey on th  (cost=0.00..23.80 rows=200 width=8)
   Distinct Keys: h1, h2, r1
   Storage Filter: (r1 = 1)
(3 rows)

because inserting values

INSERT INTO th VALUES (0, 0, 0), (0, 0, 1);

Should return 1 row. And the query returns 0 rows with prefix length 2.

---

After #17741 is resolved, we should push down a prefix length of 2 in the above case.

This also allows us to remove the need for using get_op_opfamily_strategy in yb_is_const_clause_for_distinct_pushdown function. Since get_op_opfamily_strategy can lead to negative caching issues, this removal is desirable.

Issue Type

kind/bug

Warning: Please confirm that this issue does not contain any sensitive information

• I confirm this issue does not contain any sensitive information.