[
https://issues.apache.org/jira/browse/SPARK-35080?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Allison Wang updated SPARK-35080:
---------------------------------
Description:
Correlated subqueries with aggregate that pass CheckAnalysis (with only
correlated equality predicates) can still return wrong results. This is because
equality predicates do not guarantee one-to-one mappings between inner and
outer attributes, and the semantics of the plan will be changed when the inner
attributes are pulled up through an Aggregate, which gives us wrong results.
Currently, the decorrelation framework does not support these types of
correlated subqueries, and they should be blocked in CheckAnalysis.
Example 1:
{code:sql}
create or replace view t1(c) as values ('a'), ('b')
create or replace view t2(c) as values ('ab'), ('abc'), ('bc')
select c, (select count(*) from t2 where t1.c = substring(t2.c, 1, 1)) from t1
{code}
Correct results: [(a, 2), (b, 1)]
Spark results:
{code:java}
+---+-----------------+
|c |scalarsubquery(c)|
+---+-----------------+
|a |1 |
|a |1 |
|b |1 |
+---+-----------------+{code}
Example 2:
{code:sql}
create or replace view t1(a, b) as values (0, 6), (1, 5), (2, 4), (3, 3);
create or replace view t2(c) as values (6);
select c, (select count(*) from t1 where a + b = c) from t2;{code}
Correct results: [(6, 4)]
Spark results:
{code:java}
+---+-----------------+
|c |scalarsubquery(c)|
+---+-----------------+
|6 |1 |
|6 |1 |
|6 |1 |
|6 |1 |
+---+-----------------+
{code}
was:
Correlated subqueries with aggregate that passes CheckAnalysis (with only
correlated equality predicates) can return wrong results. Currently, the
decorrelation framework does not support these types of correlated subqueries,
and they should be blocked in CheckAnalysis.
Example 1:
{code:sql}
create or replace view t1(c) as values ('a'), ('b')
create or replace view t2(c) as values ('ab'), ('abc'), ('bc')
select c, (select count(*) from t2 where t1.c = substring(t2.c, 1, 1)) from t1
{code}
Correct results: [(a, 2), (b, 1)]
Spark results:
{code:java}
+---+-----------------+
|c |scalarsubquery(c)|
+---+-----------------+
|a |1 |
|a |1 |
|b |1 |
+---+-----------------+{code}
Example 2:
{code:sql}
create or replace view t1(a, b) as values (0, 6), (1, 5), (2, 4), (3, 3);
create or replace view t2(c) as values (6);
select c, (select count(*) from t1 where a + b = c) from t2;{code}
Correct results: [(6, 4)]
Spark results:
{code:java}
+---+-----------------+
|c |scalarsubquery(c)|
+---+-----------------+
|6 |1 |
|6 |1 |
|6 |1 |
|6 |1 |
+---+-----------------+
{code}
> Correlated subqueries with equality predicates can return wrong results
> -----------------------------------------------------------------------
>
> Key: SPARK-35080
> URL: https://issues.apache.org/jira/browse/SPARK-35080
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 3.2.0
> Reporter: Allison Wang
> Priority: Major
>
> Correlated subqueries with aggregate that pass CheckAnalysis (with only
> correlated equality predicates) can still return wrong results. This is
> because equality predicates do not guarantee one-to-one mappings between
> inner and outer attributes, and the semantics of the plan will be changed
> when the inner attributes are pulled up through an Aggregate, which gives us
> wrong results. Currently, the decorrelation framework does not support these
> types of correlated subqueries, and they should be blocked in CheckAnalysis.
> Example 1:
> {code:sql}
> create or replace view t1(c) as values ('a'), ('b')
> create or replace view t2(c) as values ('ab'), ('abc'), ('bc')
> select c, (select count(*) from t2 where t1.c = substring(t2.c, 1, 1)) from t1
> {code}
> Correct results: [(a, 2), (b, 1)]
> Spark results:
> {code:java}
> +---+-----------------+
> |c |scalarsubquery(c)|
> +---+-----------------+
> |a |1 |
> |a |1 |
> |b |1 |
> +---+-----------------+{code}
> Example 2:
> {code:sql}
> create or replace view t1(a, b) as values (0, 6), (1, 5), (2, 4), (3, 3);
> create or replace view t2(c) as values (6);
> select c, (select count(*) from t1 where a + b = c) from t2;{code}
> Correct results: [(6, 4)]
> Spark results:
> {code:java}
> +---+-----------------+
> |c |scalarsubquery(c)|
> +---+-----------------+
> |6 |1 |
> |6 |1 |
> |6 |1 |
> |6 |1 |
> +---+-----------------+
> {code}
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
