gene-bordegaray commented on code in PR #18521:
URL: https://github.com/apache/datafusion/pull/18521#discussion_r2507183195
##########
datafusion/sqllogictest/test_files/joins.slt:
##########
@@ -3269,22 +3262,17 @@ logical_plan
10)----------TableScan: annotated_data projection=[a0, a, b, c, d]
physical_plan
01)SortPreservingMergeExec: [a@1 ASC, b@2 ASC NULLS LAST, c@3 ASC NULLS LAST,
rn1@11 ASC NULLS LAST]
-02)--SortExec: expr=[a@1 ASC, b@2 ASC NULLS LAST, c@3 ASC NULLS LAST, rn1@11
ASC NULLS LAST], preserve_partitioning=[true]
-03)----SortMergeJoin: join_type=Inner, on=[(a@1, a@1)]
-04)------SortExec: expr=[a@1 ASC], preserve_partitioning=[true]
-05)--------CoalesceBatchesExec: target_batch_size=2
-06)----------RepartitionExec: partitioning=Hash([a@1], 2), input_partitions=2
-07)------------RepartitionExec: partitioning=RoundRobinBatch(2),
input_partitions=1
-08)--------------ProjectionExec: expr=[a0@0 as a0, a@1 as a, b@2 as b, c@3 as
c, d@4 as d, row_number() ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED
FOLLOWING@5 as rn1]
-09)----------------BoundedWindowAggExec: wdw=[row_number() ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING: Field { "row_number() ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING": UInt64 }, frame: ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING], mode=[Sorted]
-10)------------------DataSourceExec: file_groups={1 group:
[[WORKSPACE_ROOT/datafusion/core/tests/data/window_2.csv]]}, projection=[a0, a,
b, c, d], output_ordering=[a@1 ASC, b@2 ASC NULLS LAST, c@3 ASC NULLS LAST],
file_type=csv, has_header=true
-11)------SortExec: expr=[a@1 ASC], preserve_partitioning=[true]
-12)--------CoalesceBatchesExec: target_batch_size=2
-13)----------RepartitionExec: partitioning=Hash([a@1], 2), input_partitions=2
-14)------------RepartitionExec: partitioning=RoundRobinBatch(2),
input_partitions=1
-15)--------------ProjectionExec: expr=[a0@0 as a0, a@1 as a, b@2 as b, c@3 as
c, d@4 as d, row_number() ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED
FOLLOWING@5 as rn1]
-16)----------------BoundedWindowAggExec: wdw=[row_number() ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING: Field { "row_number() ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING": UInt64 }, frame: ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING], mode=[Sorted]
-17)------------------DataSourceExec: file_groups={1 group:
[[WORKSPACE_ROOT/datafusion/core/tests/data/window_2.csv]]}, projection=[a0, a,
b, c, d], output_ordering=[a@1 ASC, b@2 ASC NULLS LAST, c@3 ASC NULLS LAST],
file_type=csv, has_header=true
+02)--SortMergeJoin: join_type=Inner, on=[(a@1, a@1)]
+03)----CoalesceBatchesExec: target_batch_size=2
+04)------RepartitionExec: partitioning=Hash([a@1], 2), input_partitions=1
+05)--------ProjectionExec: expr=[a0@0 as a0, a@1 as a, b@2 as b, c@3 as c, d@4
as d, row_number() ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING@5
as rn1]
+06)----------BoundedWindowAggExec: wdw=[row_number() ROWS BETWEEN UNBOUNDED
PRECEDING AND UNBOUNDED FOLLOWING: Field { "row_number() ROWS BETWEEN UNBOUNDED
PRECEDING AND UNBOUNDED FOLLOWING": UInt64 }, frame: ROWS BETWEEN UNBOUNDED
PRECEDING AND UNBOUNDED FOLLOWING], mode=[Sorted]
+07)------------DataSourceExec: file_groups={1 group:
[[WORKSPACE_ROOT/datafusion/core/tests/data/window_2.csv]]}, projection=[a0, a,
b, c, d], output_ordering=[a@1 ASC, b@2 ASC NULLS LAST, c@3 ASC NULLS LAST],
file_type=csv, has_header=true
+08)----CoalesceBatchesExec: target_batch_size=2
+09)------RepartitionExec: partitioning=Hash([a@1], 2), input_partitions=1
Review Comment:
I read the repartition code and ran tests, here is what I found
- yes, hash partitioning preserves order
- metadata system tracks global ordering across all partitions, not
per-partition (this could e something that could be useful for optimizations on
a partition basis)
this manual test is enough to confirm this:
```
> CREATE TABLE test AS
SELECT column1, column2, column3
FROM values
(1, 10, 100),
(1, 20, 200),
(2, 10, 300),
(2, 20, 400),
(3, 10, 500),
(3, 20, 600)
ORDER BY column1, column2, column3;
0 row(s) fetched.
Elapsed 0.016 seconds.
> SELECT column1, column2, column3,
ROW_NUMBER() OVER (
PARTITION BY column1 ORDER BY column2, column3) as rn_within_repartition,
ROW_NUMBER() OVER (
ORDER BY column1, column2, column3) as rn_global
FROM test;
+---------+---------+---------+-----------------------+-----------+
| column1 | column2 | column3 | rn_within_repartition | rn_global |
+---------+---------+---------+-----------------------+-----------+
| 1 | 10 | 100 | 1 | 1 |
| 1 | 20 | 200 | 2 | 2 |
| 2 | 10 | 300 | 1 | 3 |
| 2 | 20 | 400 | 2 | 4 |
| 3 | 10 | 500 | 1 | 5 |
| 3 | 20 | 600 | 2 | 6 |
+---------+---------+---------+-----------------------+-----------+
6 row(s) fetched.
Elapsed 0.009 seconds.
```
I can add unit / sqllogictests if needed
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
