Skip to content

Support wider range of Subquery, handle the Count bug #6457

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

Merged
merged 18 commits into from
Jun 13, 2023
Merged

Support wider range of Subquery, handle the Count bug #6457

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
18 commits
Select commit Hold shift + click to select a range
d05275d
init checkin
mingmwang May 26, 2023
2fb8bba
resolve conflicts
mingmwang May 26, 2023
ea27c70
fix comment
mingmwang May 26, 2023
d11a0ef
fix fmt
mingmwang May 26, 2023
8128659
q64 still overflow stack
mingmwang May 26, 2023
27c449b
fix count agg bug
mingmwang May 30, 2023
f64473e
resolve review comments
mingmwang May 30, 2023
df94f18
refine the count bug handling logic
mingmwang Jun 1, 2023
75eeb88
tiny fix
mingmwang Jun 1, 2023
40b93ac
merge with upstream
mingmwang Jun 1, 2023
717f51d
fix doc
mingmwang Jun 1, 2023
a752ee8
fix tpch planning change
mingmwang Jun 1, 2023
c549b6a
Avoid unnecessary alias in the InSubquery rewriting
mingmwang Jun 1, 2023
5796fa0
Merge remote-tracking branch 'apache/main' into subquery_count_agg
alamb Jun 6, 2023
ad81636
resolve conflicts
mingmwang Jun 13, 2023
7662a3a
fix joins.slt
mingmwang Jun 13, 2023
23900f5
Merge branch 'subquery_count_agg' of https://github.com/mingmwang/arr…
mingmwang Jun 13, 2023
2112d2b
fix fmt
mingmwang Jun 13, 2023
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
983 changes: 922 additions & 61 deletions datafusion/core/tests/sql/subqueries.rs
View file
Open in desktop

Large diffs are not rendered by default.

57 changes: 32 additions & 25 deletions datafusion/core/tests/sqllogictests/test_files/joins.slt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1700,10 +1700,10 @@ from join_t1
where join_t1.t1_id + 12 in (select join_t2.t2_id + 1 from join_t2)
----
logical_plan
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_5.CAST(t2_id AS Int64) + Int64(1)
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_5.join_t2.t2_id + Int64(1)
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_5
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1) AS CAST(t2_id AS Int64) + Int64(1)
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1)
------TableScan: join_t2 projection=[t2_id]

query ITI rowsort
Expand All @@ -1729,10 +1729,10 @@ where join_t1.t1_id + 12 in
)
----
logical_plan
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_7.CAST(t2_id AS Int64) + Int64(1) Filter: join_t1.t1_int <= __correlated_sq_7.t2_int
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_7.join_t2.t2_id + Int64(1) Filter: join_t1.t1_int <= __correlated_sq_7.t2_int
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_7
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1) AS CAST(t2_id AS Int64) + Int64(1), join_t2.t2_int
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1), join_t2.t2_int
------Filter: join_t2.t2_int > UInt32(0)
--------TableScan: join_t2 projection=[t2_id, t2_int]

Expand Down Expand Up @@ -1765,10 +1765,10 @@ where join_t1.t1_id + 12 in
)
----
logical_plan
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_9.CAST(t2_id AS Int64) + Int64(1) Filter: join_t1.t1_int <= __correlated_sq_9.t2_int AND join_t1.t1_name != __correlated_sq_9.t2_name
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_9.join_t2.t2_id + Int64(1) Filter: join_t1.t1_int <= __correlated_sq_9.t2_int AND join_t1.t1_name != __correlated_sq_9.t2_name
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_9
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1) AS CAST(t2_id AS Int64) + Int64(1), join_t2.t2_int, join_t2.t2_name
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1), join_t2.t2_int, join_t2.t2_name
------Filter: join_t2.t2_int > UInt32(0)
--------TableScan: join_t2 projection=[t2_id, t2_name, t2_int]

Expand Down Expand Up @@ -1797,11 +1797,11 @@ where join_t1.t1_id + 12 in
(select join_t2.t2_id + 1 from join_t2 where join_t1.t1_int > 0)
----
logical_plan
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_11.CAST(t2_id AS Int64) + Int64(1)
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_11.join_t2.t2_id + Int64(1)
--Filter: join_t1.t1_int > UInt32(0)
----TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_11
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1) AS CAST(t2_id AS Int64) + Int64(1)
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1)
------TableScan: join_t2 projection=[t2_id]

# Not in subquery to join with correlated outer filter
Expand All @@ -1814,10 +1814,10 @@ where join_t1.t1_id + 12 not in
(select join_t2.t2_id + 1 from join_t2 where join_t1.t1_int > 0)
----
logical_plan
LeftAnti Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_12.CAST(t2_id AS Int64) + Int64(1) Filter: join_t1.t1_int > UInt32(0)
LeftAnti Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_12.join_t2.t2_id + Int64(1) Filter: join_t1.t1_int > UInt32(0)
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_12
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1) AS CAST(t2_id AS Int64) + Int64(1)
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1)
------TableScan: join_t2 projection=[t2_id]

# In subquery to join with outer filter
Expand All @@ -1836,11 +1836,11 @@ where join_t1.t1_id + 12 in
and join_t1.t1_id > 0
----
logical_plan
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_13.CAST(t2_id AS Int64) + Int64(1) Filter: join_t1.t1_int <= __correlated_sq_13.t2_int AND join_t1.t1_name != __correlated_sq_13.t2_name
LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_13.join_t2.t2_id + Int64(1) Filter: join_t1.t1_int <= __correlated_sq_13.t2_int AND join_t1.t1_name != __correlated_sq_13.t2_name
--Filter: join_t1.t1_id > UInt32(0)
----TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_13
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1) AS CAST(t2_id AS Int64) + Int64(1), join_t2.t2_int, join_t2.t2_name
----Projection: CAST(join_t2.t2_id AS Int64) + Int64(1), join_t2.t2_int, join_t2.t2_name
------TableScan: join_t2 projection=[t2_id, t2_name, t2_int]

query ITI rowsort
Expand Down Expand Up @@ -1869,15 +1869,15 @@ where join_t1.t1_id + 12 in (select join_t2.t2_id + 1 from join_t2)
and join_t1.t1_id > 0
----
logical_plan
LeftSemi Join: CAST(join_t1.t1_int AS Int64) = __correlated_sq_16.CAST(t2_int AS Int64) + Int64(1)
--LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_15.CAST(t2_id AS Int64) + Int64(1)
LeftSemi Join: CAST(join_t1.t1_int AS Int64) = __correlated_sq_16.join_t2.t2_int + Int64(1)
--LeftSemi Join: CAST(join_t1.t1_id AS Int64) + Int64(12) = __correlated_sq_15.join_t2.t2_id + Int64(1)
----Filter: join_t1.t1_id > UInt32(0)
------TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
----SubqueryAlias: __correlated_sq_15
------Projection: CAST(join_t2.t2_id AS Int64) + Int64(1) AS CAST(t2_id AS Int64) + Int64(1)
------Projection: CAST(join_t2.t2_id AS Int64) + Int64(1)
--------TableScan: join_t2 projection=[t2_id]
--SubqueryAlias: __correlated_sq_16
----Projection: CAST(join_t2.t2_int AS Int64) + Int64(1) AS CAST(t2_int AS Int64) + Int64(1)
----Projection: CAST(join_t2.t2_int AS Int64) + Int64(1)
------TableScan: join_t2 projection=[t2_int]

query ITI
Expand Down Expand Up @@ -2197,8 +2197,10 @@ logical_plan
LeftAnti Join: Filter: CAST(join_t1.t1_id AS Int64) + Int64(1) > CAST(__correlated_sq_31.t2_id AS Int64) * Int64(2)
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_31
----Aggregate: groupBy=[[join_t2.t2_id]], aggr=[[]]
------TableScan: join_t2 projection=[t2_id]
----Projection: join_t2.t2_id
------Aggregate: groupBy=[[join_t2.t2_int, join_t2.t2_id]], aggr=[[]]
--------Projection: join_t2.t2_int, join_t2.t2_id
----------TableScan: join_t2 projection=[t2_id, t2_int]

statement ok
set datafusion.optimizer.repartition_joins = false;
Expand Down Expand Up @@ -2244,8 +2246,10 @@ logical_plan
LeftAnti Join: Filter: CAST(join_t1.t1_id AS Int64) + Int64(1) > CAST(__correlated_sq_34.t2_id AS Int64) * Int64(2)
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_34
----Aggregate: groupBy=[[join_t2.t2_id]], aggr=[[]]
------TableScan: join_t2 projection=[t2_id]
----Projection: join_t2.t2_id
------Aggregate: groupBy=[[join_t2.t2_id + join_t2.t2_int, join_t2.t2_int, join_t2.t2_id]], aggr=[[]]
--------Projection: join_t2.t2_id + join_t2.t2_int, join_t2.t2_int, join_t2.t2_id
----------TableScan: join_t2 projection=[t2_id, t2_int]

statement ok
set datafusion.optimizer.repartition_joins = false;
Expand Down Expand Up @@ -2293,8 +2297,10 @@ logical_plan
LeftAnti Join: Filter: CAST(join_t1.t1_id AS Int64) + Int64(1) > CAST(__correlated_sq_37.t2_id AS Int64) * Int64(2)
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_37
----Aggregate: groupBy=[[join_t2.t2_id]], aggr=[[]]
------TableScan: join_t2 projection=[t2_id]
----Projection: join_t2.t2_id
------Aggregate: groupBy=[[Int64(1), join_t2.t2_int, join_t2.t2_id]], aggr=[[]]
--------Projection: Int64(1), join_t2.t2_int, join_t2.t2_id
----------TableScan: join_t2 projection=[t2_id, t2_int]

query ITI
SELECT * FROM join_t1
Expand Down Expand Up @@ -2322,8 +2328,10 @@ logical_plan
LeftAnti Join: Filter: CAST(join_t1.t1_id AS Int64) + Int64(1) > CAST(__correlated_sq_39.t2_id AS Int64) * Int64(2)
--TableScan: join_t1 projection=[t1_id, t1_name, t1_int]
--SubqueryAlias: __correlated_sq_39
----Aggregate: groupBy=[[join_t2.t2_id]], aggr=[[]]
------TableScan: join_t2 projection=[t2_id]
----Projection: join_t2.t2_id
------Aggregate: groupBy=[[Int64(1), join_t2.t2_int, join_t2.t2_id]], aggr=[[]]
--------Projection: Int64(1), join_t2.t2_int, join_t2.t2_id
----------TableScan: join_t2 projection=[t2_id, t2_int]

query ITI
SELECT * FROM join_t1
Expand Down Expand Up @@ -2936,4 +2944,3 @@ set datafusion.execution.target_partitions = 2;

statement ok
set datafusion.execution.batch_size = 4096;

Loading