Skip to content

[CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree when using correlated sub-query in on clause of equi-join #4763

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
iwanttobepowerful wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree when using correlated sub-query in on clause of equi-join #4763

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
279 changes: 279 additions & 0 deletions core/src/test/resources/sql/sub-query.iq
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8489,5 +8489,284 @@ FROM dept;
+--------+--------+
(4 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 1: Correlated scalar subquery in ON clause of INNER JOIN
#
# before fix return wrong result 0 rows and the wrong plan after SubQueryRemoveRule is:
# LogicalProject(ID=[$0], ID0=[$1])
# LogicalProject(id=[$0], id0=[$1])
# LogicalJoin(condition=[=($2, $1)], joinType=[inner])
# LogicalTableScan(table=[[a]])
# LogicalCorrelate(correlation=[$cor0], joinType=[left], requiredColumns=[{0}])
# LogicalTableScan(table=[[c]])
# LogicalAggregate(group=[{}], EXPR$0=[MIN($0)])
# LogicalProject(fk_c=[$2])
# LogicalFilter(condition=[=($cor0.id, $1)])
# LogicalTableScan(table=[[b]])
#
# after fix the plan after SubQueryRemoveRule is:
# LogicalProject(ID=[$0], ID0=[$1])
# LogicalProject(id=[$0], id0=[$2])
# LogicalJoin(condition=[=($1, $2)], joinType=[inner])
# LogicalCorrelate(correlation=[$cor0], joinType=[left], requiredColumns=[{0}])
# LogicalTableScan(table=[[a]])
# LogicalAggregate(group=[{}], EXPR$0=[MIN($0)])
# LogicalProject(fk_c=[$2])
# LogicalFilter(condition=[=($cor0.id, $1)])
# LogicalTableScan(table=[[b]])
# LogicalTableScan(table=[[c]])
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM A
INNER JOIN C
ON (
SELECT min(B.fk_C)
FROM B
WHERE A.id = B.fk_A
) = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
+-----+-----+
(1 row)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 1 (Verification): Manual rewrite of Case 1 using a derived table
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM (
SELECT
*,
(SELECT min(B.fk_C) FROM B WHERE A.id = B.fk_A) AS fk_C
FROM A
) AS A
INNER JOIN C
ON fk_c = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
+-----+-----+
(1 row)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 2: Correlated scalar subquery in ON clause of LEFT JOIN
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM A
LEFT JOIN C
ON (
SELECT min(B.fk_C)
FROM B
WHERE A.id = B.fk_A
) = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| 102 | |
+-----+-----+
(2 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 2 (Verification): Manual rewrite of Case 2 using a derived table
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM (
SELECT
*,
(SELECT min(B.fk_C) FROM B WHERE A.id = B.fk_A) AS fk_C
FROM A
) AS A
LEFT JOIN C
ON fk_c = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| 102 | |
+-----+-----+
(2 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 3: Correlated scalar subquery in ON clause of RIGHT JOIN.
# The subquery correlates to the RHS table (A), which is the PRESERVED side.
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM C
RIGHT JOIN A
ON (
SELECT min(B.fk_C)
FROM B
WHERE A.id = B.fk_A
) = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| 102 | |
+-----+-----+
(2 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 3 (Verification): Manual rewrite of Case 3 using a derived table.
# Projects the scalar subquery on the RHS (preserved side).
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM C
RIGHT JOIN (
SELECT
*,
(SELECT min(B.fk_C) FROM B WHERE A.id = B.fk_A) AS fk_C
FROM A
) AS A
ON fk_c = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| 102 | |
+-----+-----+
(2 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 4: Correlated scalar subquery in ON clause of LEFT JOIN
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM C
LEFT JOIN A
ON (
SELECT min(B.fk_C)
FROM B
WHERE A.id = B.fk_A
) = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| | 302 |
+-----+-----+
(2 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 4 (Verification): Manual rewrite of Case 4 using a derived table
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM C
LEFT JOIN (
SELECT
*,
(SELECT min(B.fk_C) FROM B WHERE A.id = B.fk_A) AS fk_C
FROM A
) AS A
ON fk_c = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| | 302 |
+-----+-----+
(2 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 5: Correlated scalar subquery in ON clause of RIGHT JOIN
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM A
RIGHT JOIN C
ON (
SELECT min(B.fk_C)
FROM B
WHERE A.id = B.fk_A
) = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| | 302 |
+-----+-----+
(2 rows)

!ok

# [CALCITE-6504] JOIN_SUB_QUERY_TO_CORRELATE/Join SubQueryRemoveRule produces incorrect tree
# when using correlated sub-query in on clause of equi-join
# Case 5 (Verification): Manual rewrite of Case 5 using a derived table
WITH
A(id) AS (VALUES (101), (102)),
C(id) AS (VALUES (301), (302)),
B(id, fk_A, fk_C) AS (VALUES (201, 101, 301), (202, 101, 999), (203, 999, 301), (204, 999, 999))
SELECT A.id, C.id
FROM (
SELECT
*,
(SELECT min(B.fk_C) FROM B WHERE A.id = B.fk_A) AS fk_C
FROM A
) AS A
RIGHT JOIN C
ON fk_c = C.id;
+-----+-----+
| ID | ID |
+-----+-----+
| 101 | 301 |
| | 302 |
+-----+-----+
(2 rows)

!ok
# End sub-query.iq
Loading