[CALCITE-5816] Query with LEFT SEMI JOIN should not refer to RHS columns

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 1.34.0
• Fix Version/s: 1.35.0
• Component/s: babel, core
• Labels: pull-request-available
• Assignee: hongyu guo
• Reporter: hongyu guo
• Created: 03/Jul/23 11:44
• Updated: 27/Jul/23 10:22
• Resolved: 05/Jul/23 11:26

Description

A LEFT SEMI JOIN can only return columns from the left-hand table.
But calcite also return both left-hand and right-hand table columns.

In select.iq

# LEFT SEMI JOIN (Hive only)
SELECT *
FROM emp LEFT SEMI JOIN dept ON emp.deptno = dept.deptno;

SELECT "EMP"."EMPNO", "EMP"."ENAME", "EMP"."JOB", "EMP"."MGR", "EMP"."HIREDATE", "EMP"."SAL", "EMP"."COMM", "EMP"."DEPTNO", "DEPT"."DEPTNO" AS "DEPTNO0", "DEPT"."DNAME", "DEPT"."LOC"
FROM "scott"."EMP" AS "EMP"
    LEFT SEMI JOIN "scott"."DEPT" AS "DEPT" ON "EMP"."DEPTNO" = "DEPT"."DEPTNO"
!explain-validated-on hive

Issue Links

• fixes: CALCITE-2304 In Babel parser, allow Hive-style syntax "LEFT SEMI JOIN"
• links to: GitHub Pull Request #3291