This is an automated email from the ASF dual-hosted git repository.
comphead pushed a commit to branch branch-53
in repository https://gitbox.apache.org/repos/asf/datafusion.git
The following commit(s) were added to refs/heads/branch-53 by this push:
new c45c2ce4a9 [branch-53] Substrait join consumer should not merge
nullability of join keys (#21121) (#21162)
c45c2ce4a9 is described below
commit c45c2ce4a9da6fb5e487c9585b2fda3110c9067d
Author: Haresh Khanna <[email protected]>
AuthorDate: Thu Mar 26 03:41:22 2026 +0000
[branch-53] Substrait join consumer should not merge nullability of join
keys (#21121) (#21162)
## Which issue does this PR close?
<!--
We generally require a GitHub issue to be filed for all bug fixes and
enhancements and this helps us generate change logs for our releases.
You can link an issue to this PR using the GitHub syntax. For example
`Closes #123` indicates that this PR will close issue #123.
-->
- Related to #21124
- Related to #21079
## Rationale for this change
<!--
Why are you proposing this change? If this is already explained clearly
in the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand
your changes and offer better suggestions for fixes.
-->
When a Substrait join expression contains both equal and
is_not_distinct_from predicates (e.g. Spark pushes a null-safe filter
into a join that already has a regular equality key), the
`split_eq_and_noneq_join_predicate_with_nulls_equality` function uses a
single `nulls_equal_nulls` boolean that gets overwritten per-predicate.
Whichever operator is processed last determines the `NullEquality` for
all keys, silently dropping NULL-matching rows.
Since NullEquality is a join-level setting (not per-key) across all
physical join implementations (HashJoinExec, SortMergeJoinExec,
SymmetricHashJoinExec), the correct fix is to match DataFusion's own SQL
planner behavior: demote IS NOT DISTINCT FROM keys to the join filter
when mixed with Eq keys. This is already correctly handled for SQL as
shown in
[join_is_not_distinct_from.slt:L188](https://sourcegraph.com/r/github.com/apache/datafusion@2b7d4f9a5b005905b23128274ad37c3306ffcd15/-/blob/datafusion/sqllogictest/test_files/join_is_not_distinct_from.slt?L188)
```
# Test mixed equal and IS NOT DISTINCT FROM conditions
# The `IS NOT DISTINCT FROM` expr should NOT in HashJoin's `on` predicate
query TT
EXPLAIN SELECT t1.id AS t1_id, t2.id AS t2_id, t1.val, t2.val
FROM t1
JOIN t2 ON t1.id = t2.id AND t1.val IS NOT DISTINCT FROM t2.val
----
logical_plan
01)Projection: t1.id AS t1_id, t2.id AS t2_id, t1.val, t2.val
02)--Inner Join: t1.id = t2.id Filter: t1.val IS NOT DISTINCT FROM t2.val
03)----TableScan: t1 projection=[id, val]
04)----TableScan: t2 projection=[id, val]
```
## What changes are included in this PR?
<!--
There is no need to duplicate the description in the issue here but it
is sometimes worth providing a summary of the individual changes in this
PR.
-->
`datafusion/substrait/src/logical_plan/consumer/rel/join_rel.rs`:
- Collect eq_keys and indistinct_keys separately instead of using a
single vec with an overwritable boolean
- When both are present (mixed case), use eq_keys as equijoin keys with
NullEqualsNothing and reconstruct the IsNotDistinctFrom expressions into
the join filter
- Return NullEquality directly instead of converting from bool
## Are these changes tested?
<!--
We typically require tests for all PRs in order to:
1. Prevent the code from being accidentally broken by subsequent changes
2. Serve as another way to document the expected behavior of the code
If tests are not included in your PR, please explain why (for example,
are they covered by existing tests)?
-->
Yes, three levels of coverage:
1. Unit tests (join_rel.rs) — directly assert the output of
split_eq_and_noneq_join_predicate_with_nulls_equality for eq-only,
indistinct-only, mixed, and non-column-operand cases
2. Integration test (consumer_integration.rs) — loads a JSON-encoded
Substrait plan with a JoinRel containing both operators through
from_substrait_plan, executes it, and asserts 6 rows (including
NULL=NULL matches)
3. Existing SLT (join_is_not_distinct_from.slt:179-205) — confirms the
SQL planner already exhibits the same demotion behavior that this PR
adds to the Substrait consumer
## Are there any user-facing changes?
<!--
If there are user-facing changes then we may require documentation to be
updated before approving the PR.
-->
<!--
If there are any breaking changes to public APIs, please add the `api
change` label.
-->
No API changes. Substrait plans with mixed equal/is_not_distinct_from
join predicates now correctly preserve null-safe semantics instead of
silently dropping NULL-matching rows.
---
.../test_files/join_is_not_distinct_from.slt | 30 ++++
.../src/logical_plan/consumer/rel/join_rel.rs | 186 ++++++++++++++++-----
.../substrait/tests/cases/consumer_integration.rs | 106 ++++++++++++
.../mixed_join_equal_and_indistinct.json | 102 +++++++++++
.../mixed_join_equal_and_indistinct_left.json | 102 +++++++++++
5 files changed, 486 insertions(+), 40 deletions(-)
diff --git a/datafusion/sqllogictest/test_files/join_is_not_distinct_from.slt
b/datafusion/sqllogictest/test_files/join_is_not_distinct_from.slt
index 8246f489c4..2bab89c99e 100644
--- a/datafusion/sqllogictest/test_files/join_is_not_distinct_from.slt
+++ b/datafusion/sqllogictest/test_files/join_is_not_distinct_from.slt
@@ -291,6 +291,36 @@ JOIN t4 ON (t3.val1 IS NOT DISTINCT FROM t4.val1) AND
(t3.val2 IS NOT DISTINCT F
2 2 NULL NULL 200 200
3 3 30 30 NULL NULL
+# Test mixed: 1 Eq key + multiple IS NOT DISTINCT FROM keys.
+# The optimizer unconditionally favours Eq keys (see
extract_equijoin_predicate.rs,
+# "Only convert when there are NO equijoin predicates, to be conservative").
+# All IS NOT DISTINCT FROM predicates should be demoted to filter, even when
they outnumber the Eq key.
+query TT
+EXPLAIN SELECT t3.id AS t3_id, t4.id AS t4_id, t3.val1, t4.val1, t3.val2,
t4.val2
+FROM t3
+JOIN t4 ON (t3.id = t4.id) AND (t3.val1 IS NOT DISTINCT FROM t4.val1) AND
(t3.val2 IS NOT DISTINCT FROM t4.val2)
+----
+logical_plan
+01)Projection: t3.id AS t3_id, t4.id AS t4_id, t3.val1, t4.val1, t3.val2,
t4.val2
+02)--Inner Join: t3.id = t4.id Filter: t3.val1 IS NOT DISTINCT FROM t4.val1
AND t3.val2 IS NOT DISTINCT FROM t4.val2
+03)----TableScan: t3 projection=[id, val1, val2]
+04)----TableScan: t4 projection=[id, val1, val2]
+physical_plan
+01)ProjectionExec: expr=[id@0 as t3_id, id@3 as t4_id, val1@1 as val1, val1@4
as val1, val2@2 as val2, val2@5 as val2]
+02)--HashJoinExec: mode=CollectLeft, join_type=Inner, on=[(id@0, id@0)],
filter=val1@0 IS NOT DISTINCT FROM val1@2 AND val2@1 IS NOT DISTINCT FROM val2@3
+03)----DataSourceExec: partitions=1, partition_sizes=[1]
+04)----DataSourceExec: partitions=1, partition_sizes=[1]
+
+# Verify correct results: all 3 rows should match (including NULL=NULL via IS
NOT DISTINCT FROM in filter)
+query IIIIII rowsort
+SELECT t3.id AS t3_id, t4.id AS t4_id, t3.val1, t4.val1, t3.val2, t4.val2
+FROM t3
+JOIN t4 ON (t3.id = t4.id) AND (t3.val1 IS NOT DISTINCT FROM t4.val1) AND
(t3.val2 IS NOT DISTINCT FROM t4.val2)
+----
+1 1 10 10 100 100
+2 2 NULL NULL 200 200
+3 3 30 30 NULL NULL
+
statement ok
drop table t0;
diff --git a/datafusion/substrait/src/logical_plan/consumer/rel/join_rel.rs
b/datafusion/substrait/src/logical_plan/consumer/rel/join_rel.rs
index 3604630d6f..7850dbea79 100644
--- a/datafusion/substrait/src/logical_plan/consumer/rel/join_rel.rs
+++ b/datafusion/substrait/src/logical_plan/consumer/rel/join_rel.rs
@@ -18,7 +18,7 @@
use crate::logical_plan::consumer::SubstraitConsumer;
use datafusion::common::{Column, JoinType, NullEquality, not_impl_err,
plan_err};
use datafusion::logical_expr::requalify_sides_if_needed;
-use datafusion::logical_expr::utils::split_conjunction;
+use datafusion::logical_expr::utils::split_conjunction_owned;
use datafusion::logical_expr::{
BinaryExpr, Expr, LogicalPlan, LogicalPlanBuilder, Operator,
};
@@ -56,15 +56,10 @@ pub async fn from_join_rel(
// So we extract each part as follows:
// - If an Eq or IsNotDistinctFrom op is encountered, add the left
column, right column and is_null_equal_nulls to `join_ons` vector
// - Otherwise we add the expression to join_filter (use
conjunction if filter already exists)
- let (join_ons, nulls_equal_nulls, join_filter) =
- split_eq_and_noneq_join_predicate_with_nulls_equality(&on);
+ let (join_ons, null_equality, join_filter) =
+ split_eq_and_noneq_join_predicate_with_nulls_equality(on);
let (left_cols, right_cols): (Vec<_>, Vec<_>) =
itertools::multiunzip(join_ons);
- let null_equality = if nulls_equal_nulls {
- NullEquality::NullEqualsNull
- } else {
- NullEquality::NullEqualsNothing
- };
left.join_detailed(
right.build()?,
join_type,
@@ -89,49 +84,61 @@ pub async fn from_join_rel(
}
fn split_eq_and_noneq_join_predicate_with_nulls_equality(
- filter: &Expr,
-) -> (Vec<(Column, Column)>, bool, Option<Expr>) {
- let exprs = split_conjunction(filter);
+ filter: Expr,
+) -> (Vec<(Column, Column)>, NullEquality, Option<Expr>) {
+ let exprs = split_conjunction_owned(filter);
- let mut accum_join_keys: Vec<(Column, Column)> = vec![];
+ let mut eq_keys: Vec<(Column, Column)> = vec![];
+ let mut indistinct_keys: Vec<(Column, Column)> = vec![];
let mut accum_filters: Vec<Expr> = vec![];
- let mut nulls_equal_nulls = false;
for expr in exprs {
- #[expect(clippy::collapsible_match)]
match expr {
- Expr::BinaryExpr(binary_expr) => match binary_expr {
- x @ (BinaryExpr {
- left,
- op: Operator::Eq,
- right,
+ Expr::BinaryExpr(BinaryExpr {
+ left,
+ op: op @ (Operator::Eq | Operator::IsNotDistinctFrom),
+ right,
+ }) => match (*left, *right) {
+ (Expr::Column(l), Expr::Column(r)) => match op {
+ Operator::Eq => eq_keys.push((l, r)),
+ Operator::IsNotDistinctFrom => indistinct_keys.push((l,
r)),
+ _ => unreachable!(),
+ },
+ (left, right) => {
+ accum_filters.push(Expr::BinaryExpr(BinaryExpr {
+ left: Box::new(left),
+ op,
+ right: Box::new(right),
+ }));
}
- | BinaryExpr {
- left,
- op: Operator::IsNotDistinctFrom,
- right,
- }) => {
- nulls_equal_nulls = match x.op {
- Operator::Eq => false,
- Operator::IsNotDistinctFrom => true,
- _ => unreachable!(),
- };
-
- match (left.as_ref(), right.as_ref()) {
- (Expr::Column(l), Expr::Column(r)) => {
- accum_join_keys.push((l.clone(), r.clone()));
- }
- _ => accum_filters.push(expr.clone()),
- }
- }
- _ => accum_filters.push(expr.clone()),
},
- _ => accum_filters.push(expr.clone()),
+ _ => accum_filters.push(expr),
}
}
+ let (join_keys, null_equality) =
+ match (eq_keys.is_empty(), indistinct_keys.is_empty()) {
+ // Mixed: use eq_keys as equijoin keys, demote indistinct keys to
filter
+ (false, false) => {
+ for (l, r) in indistinct_keys {
+ accum_filters.push(Expr::BinaryExpr(BinaryExpr {
+ left: Box::new(Expr::Column(l)),
+ op: Operator::IsNotDistinctFrom,
+ right: Box::new(Expr::Column(r)),
+ }));
+ }
+ (eq_keys, NullEquality::NullEqualsNothing)
+ }
+ // Only eq keys
+ (false, true) => (eq_keys, NullEquality::NullEqualsNothing),
+ // Only indistinct keys
+ (true, false) => (indistinct_keys, NullEquality::NullEqualsNull),
+ // No keys at all
+ (true, true) => (vec![], NullEquality::NullEqualsNothing),
+ };
+
let join_filter = accum_filters.into_iter().reduce(Expr::and);
- (accum_join_keys, nulls_equal_nulls, join_filter)
+ (join_keys, null_equality, join_filter)
}
fn from_substrait_jointype(join_type: i32) ->
datafusion::common::Result<JoinType> {
@@ -153,3 +160,102 @@ fn from_substrait_jointype(join_type: i32) ->
datafusion::common::Result<JoinTyp
plan_err!("invalid join type variant {join_type}")
}
}
+
+#[cfg(test)]
+mod tests {
+ use super::*;
+
+ fn col(name: &str) -> Expr {
+ Expr::Column(Column::from_name(name))
+ }
+
+ fn indistinct(left: Expr, right: Expr) -> Expr {
+ Expr::BinaryExpr(BinaryExpr {
+ left: Box::new(left),
+ op: Operator::IsNotDistinctFrom,
+ right: Box::new(right),
+ })
+ }
+
+ fn fmt_keys(keys: &[(Column, Column)]) -> String {
+ keys.iter()
+ .map(|(l, r)| format!("{l} = {r}"))
+ .collect::<Vec<_>>()
+ .join(", ")
+ }
+
+ #[test]
+ fn split_only_eq_keys() {
+ let expr = col("a").eq(col("b"));
+ let (keys, null_eq, filter) =
+ split_eq_and_noneq_join_predicate_with_nulls_equality(expr);
+
+ assert_eq!(fmt_keys(&keys), "a = b");
+ assert_eq!(null_eq, NullEquality::NullEqualsNothing);
+ assert!(filter.is_none());
+ }
+
+ #[test]
+ fn split_only_indistinct_keys() {
+ let expr = indistinct(col("a"), col("b"));
+ let (keys, null_eq, filter) =
+ split_eq_and_noneq_join_predicate_with_nulls_equality(expr);
+
+ assert_eq!(fmt_keys(&keys), "a = b");
+ assert_eq!(null_eq, NullEquality::NullEqualsNull);
+ assert!(filter.is_none());
+ }
+
+ /// Regression: mixed `equal` + `is_not_distinct_from` must demote
+ /// the indistinct key to the join filter so the single NullEquality
+ /// flag stays consistent (NullEqualsNothing for the eq keys).
+ #[test]
+ fn split_mixed_eq_and_indistinct_demotes_indistinct_to_filter() {
+ let expr =
+ indistinct(col("val_l"),
col("val_r")).and(col("id_l").eq(col("id_r")));
+
+ let (keys, null_eq, filter) =
+ split_eq_and_noneq_join_predicate_with_nulls_equality(expr);
+
+ assert_eq!(fmt_keys(&keys), "id_l = id_r");
+ assert_eq!(null_eq, NullEquality::NullEqualsNothing);
+ assert_eq!(
+ filter.unwrap().to_string(),
+ "val_l IS NOT DISTINCT FROM val_r"
+ );
+ }
+
+ /// Multiple IS NOT DISTINCT FROM keys with a single Eq key should demote
+ /// all indistinct keys to the filter.
+ #[test]
+ fn split_mixed_multiple_indistinct_demoted() {
+ let expr = indistinct(col("a_l"), col("a_r"))
+ .and(indistinct(col("b_l"), col("b_r")))
+ .and(col("id_l").eq(col("id_r")));
+
+ let (keys, null_eq, filter) =
+ split_eq_and_noneq_join_predicate_with_nulls_equality(expr);
+
+ assert_eq!(fmt_keys(&keys), "id_l = id_r");
+ assert_eq!(null_eq, NullEquality::NullEqualsNothing);
+ assert_eq!(
+ filter.unwrap().to_string(),
+ "a_l IS NOT DISTINCT FROM a_r AND b_l IS NOT DISTINCT FROM b_r"
+ );
+ }
+
+ #[test]
+ fn split_non_column_eq_goes_to_filter() {
+ let expr = Expr::Literal(
+ datafusion::common::ScalarValue::Utf8(Some("x".into())),
+ None,
+ )
+ .eq(col("b"));
+
+ let (keys, _, filter) =
+ split_eq_and_noneq_join_predicate_with_nulls_equality(expr);
+
+ assert!(keys.is_empty());
+ assert_eq!(filter.unwrap().to_string(), "Utf8(\"x\") = b");
+ }
+}
diff --git a/datafusion/substrait/tests/cases/consumer_integration.rs
b/datafusion/substrait/tests/cases/consumer_integration.rs
index 88c4eb6678..b5d9f36620 100644
--- a/datafusion/substrait/tests/cases/consumer_integration.rs
+++ b/datafusion/substrait/tests/cases/consumer_integration.rs
@@ -25,6 +25,8 @@
#[cfg(test)]
mod tests {
use crate::utils::test::add_plan_schemas_to_ctx;
+ use datafusion::arrow::record_batch::RecordBatch;
+ use datafusion::arrow::util::pretty::pretty_format_batches;
use datafusion::common::Result;
use datafusion::prelude::SessionContext;
use datafusion_substrait::logical_plan::consumer::from_substrait_plan;
@@ -33,6 +35,34 @@ mod tests {
use std::io::BufReader;
use substrait::proto::Plan;
+ async fn execute_plan(name: &str) -> Result<Vec<RecordBatch>> {
+ let path = format!("tests/testdata/test_plans/{name}");
+ let proto = serde_json::from_reader::<_, Plan>(BufReader::new(
+ File::open(path).expect("file not found"),
+ ))
+ .expect("failed to parse json");
+ let ctx = SessionContext::new();
+ let plan = from_substrait_plan(&ctx.state(), &proto).await?;
+ ctx.execute_logical_plan(plan).await?.collect().await
+ }
+
+ /// Pretty-print batches as a table with header on top and data rows
sorted.
+ fn pretty_sorted(batches: &[RecordBatch]) -> String {
+ let pretty = pretty_format_batches(batches).unwrap().to_string();
+ let all_lines: Vec<&str> = pretty.trim().lines().collect();
+ let header = &all_lines[..3];
+ let mut data: Vec<&str> = all_lines[3..all_lines.len() - 1].to_vec();
+ data.sort();
+ let footer = &all_lines[all_lines.len() - 1..];
+ header
+ .iter()
+ .copied()
+ .chain(data)
+ .chain(footer.iter().copied())
+ .collect::<Vec<_>>()
+ .join("\n")
+ }
+
async fn tpch_plan_to_string(query_id: i32) -> Result<String> {
let path =
format!("tests/testdata/tpch_substrait_plans/query_{query_id:02}_plan.json");
@@ -762,4 +792,80 @@ mod tests {
Ok(())
}
+
+ /// Substrait join with both `equal` and `is_not_distinct_from` must demote
+ /// `IS NOT DISTINCT FROM` to the join filter.
+ #[tokio::test]
+ async fn test_mixed_join_equal_and_indistinct_inner_join() -> Result<()> {
+ let plan_str =
+ test_plan_to_string("mixed_join_equal_and_indistinct.json").await?;
+ // Eq becomes the equijoin key; IS NOT DISTINCT FROM is demoted to
filter.
+ assert_snapshot!(
+ plan_str,
+ @r#"
+ Projection: left.id, left.val, left.comment, right.id AS id0,
right.val AS val0, right.comment AS comment0
+ Inner Join: left.id = right.id Filter: left.val IS NOT DISTINCT FROM
right.val
+ SubqueryAlias: left
+ Values: (Utf8("1"), Utf8("a"), Utf8("c1")), (Utf8("2"),
Utf8("b"), Utf8("c2")), (Utf8("3"), Utf8(NULL), Utf8("c3")), (Utf8("4"),
Utf8(NULL), Utf8("c4")), (Utf8("5"), Utf8("e"), Utf8("c5"))...
+ SubqueryAlias: right
+ Values: (Utf8("1"), Utf8("a"), Utf8("c1")), (Utf8("2"),
Utf8("b"), Utf8("c2")), (Utf8("3"), Utf8(NULL), Utf8("c3")), (Utf8("4"),
Utf8(NULL), Utf8("c4")), (Utf8("5"), Utf8("e"), Utf8("c5"))...
+ "#
+ );
+
+ // Execute and verify actual rows, including NULL=NULL matches (ids
3,4).
+ let results =
execute_plan("mixed_join_equal_and_indistinct.json").await?;
+ assert_snapshot!(pretty_sorted(&results),
+ @r"
+ +----+-----+---------+-----+------+----------+
+ | id | val | comment | id0 | val0 | comment0 |
+ +----+-----+---------+-----+------+----------+
+ | 1 | a | c1 | 1 | a | c1 |
+ | 2 | b | c2 | 2 | b | c2 |
+ | 3 | | c3 | 3 | | c3 |
+ | 4 | | c4 | 4 | | c4 |
+ | 5 | e | c5 | 5 | e | c5 |
+ | 6 | f | c6 | 6 | f | c6 |
+ +----+-----+---------+-----+------+----------+
+ "
+ );
+
+ Ok(())
+ }
+
+ /// Substrait join with both `equal` and `is_not_distinct_from` must demote
+ /// `IS NOT DISTINCT FROM` to the join filter.
+ #[tokio::test]
+ async fn test_mixed_join_equal_and_indistinct_left_join() -> Result<()> {
+ let plan_str =
+
test_plan_to_string("mixed_join_equal_and_indistinct_left.json").await?;
+ assert_snapshot!(
+ plan_str,
+ @r#"
+ Projection: left.id, left.val, left.comment, right.id AS id0,
right.val AS val0, right.comment AS comment0
+ Left Join: left.id = right.id Filter: left.val IS NOT DISTINCT FROM
right.val
+ SubqueryAlias: left
+ Values: (Utf8("1"), Utf8("a"), Utf8("c1")), (Utf8("2"),
Utf8("b"), Utf8("c2")), (Utf8("3"), Utf8(NULL), Utf8("c3")), (Utf8("4"),
Utf8(NULL), Utf8("c4")), (Utf8("5"), Utf8("e"), Utf8("c5"))...
+ SubqueryAlias: right
+ Values: (Utf8("1"), Utf8("a"), Utf8("c1")), (Utf8("2"),
Utf8("b"), Utf8("c2")), (Utf8("3"), Utf8(NULL), Utf8("c3")), (Utf8("4"),
Utf8(NULL), Utf8("c4")), (Utf8("5"), Utf8("e"), Utf8("c5"))...
+ "#
+ );
+
+ let results =
execute_plan("mixed_join_equal_and_indistinct_left.json").await?;
+ assert_snapshot!(pretty_sorted(&results),
+ @r"
+ +----+-----+---------+-----+------+----------+
+ | id | val | comment | id0 | val0 | comment0 |
+ +----+-----+---------+-----+------+----------+
+ | 1 | a | c1 | 1 | a | c1 |
+ | 2 | b | c2 | 2 | b | c2 |
+ | 3 | | c3 | 3 | | c3 |
+ | 4 | | c4 | 4 | | c4 |
+ | 5 | e | c5 | 5 | e | c5 |
+ | 6 | f | c6 | 6 | f | c6 |
+ +----+-----+---------+-----+------+----------+
+ "
+ );
+
+ Ok(())
+ }
}
diff --git
a/datafusion/substrait/tests/testdata/test_plans/mixed_join_equal_and_indistinct.json
b/datafusion/substrait/tests/testdata/test_plans/mixed_join_equal_and_indistinct.json
new file mode 100644
index 0000000000..642256c562
--- /dev/null
+++
b/datafusion/substrait/tests/testdata/test_plans/mixed_join_equal_and_indistinct.json
@@ -0,0 +1,102 @@
+{
+ "extensions": [
+ { "extensionFunction": { "functionAnchor": 0, "name":
"is_not_distinct_from" } },
+ { "extensionFunction": { "functionAnchor": 1, "name": "equal" } },
+ { "extensionFunction": { "functionAnchor": 2, "name": "and" } }
+ ],
+ "relations": [{
+ "root": {
+ "input": {
+ "join": {
+ "common": { "direct": {} },
+ "left": {
+ "read": {
+ "common": { "direct": {} },
+ "baseSchema": {
+ "names": ["id", "val", "comment"],
+ "struct": {
+ "types": [
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } },
+ { "string": { "nullability": "NULLABILITY_NULLABLE" } },
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } }
+ ],
+ "nullability": "NULLABILITY_REQUIRED"
+ }
+ },
+ "virtualTable": {
+ "values": [
+ { "fields": [{ "string": "1", "nullable": false }, {
"string": "a", "nullable": true }, { "string": "c1", "nullable": false }] },
+ { "fields": [{ "string": "2", "nullable": false }, {
"string": "b", "nullable": true }, { "string": "c2", "nullable": false }] },
+ { "fields": [{ "string": "3", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c3", "nullable": false }] },
+ { "fields": [{ "string": "4", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c4", "nullable": false }] },
+ { "fields": [{ "string": "5", "nullable": false }, {
"string": "e", "nullable": true }, { "string": "c5", "nullable": false }] },
+ { "fields": [{ "string": "6", "nullable": false }, {
"string": "f", "nullable": true }, { "string": "c6", "nullable": false }] }
+ ]
+ }
+ }
+ },
+ "right": {
+ "read": {
+ "common": { "direct": {} },
+ "baseSchema": {
+ "names": ["id", "val", "comment"],
+ "struct": {
+ "types": [
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } },
+ { "string": { "nullability": "NULLABILITY_NULLABLE" } },
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } }
+ ],
+ "nullability": "NULLABILITY_REQUIRED"
+ }
+ },
+ "virtualTable": {
+ "values": [
+ { "fields": [{ "string": "1", "nullable": false }, {
"string": "a", "nullable": true }, { "string": "c1", "nullable": false }] },
+ { "fields": [{ "string": "2", "nullable": false }, {
"string": "b", "nullable": true }, { "string": "c2", "nullable": false }] },
+ { "fields": [{ "string": "3", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c3", "nullable": false }] },
+ { "fields": [{ "string": "4", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c4", "nullable": false }] },
+ { "fields": [{ "string": "5", "nullable": false }, {
"string": "e", "nullable": true }, { "string": "c5", "nullable": false }] },
+ { "fields": [{ "string": "6", "nullable": false }, {
"string": "f", "nullable": true }, { "string": "c6", "nullable": false }] }
+ ]
+ }
+ }
+ },
+ "expression": {
+ "scalarFunction": {
+ "functionReference": 2,
+ "outputType": { "bool": { "nullability": "NULLABILITY_NULLABLE"
} },
+ "arguments": [
+ {
+ "value": {
+ "scalarFunction": {
+ "functionReference": 0,
+ "outputType": { "bool": { "nullability":
"NULLABILITY_NULLABLE" } },
+ "arguments": [
+ { "value": { "selection": { "directReference": {
"structField": { "field": 1 } }, "rootReference": {} } } },
+ { "value": { "selection": { "directReference": {
"structField": { "field": 4 } }, "rootReference": {} } } }
+ ]
+ }
+ }
+ },
+ {
+ "value": {
+ "scalarFunction": {
+ "functionReference": 1,
+ "outputType": { "bool": { "nullability":
"NULLABILITY_NULLABLE" } },
+ "arguments": [
+ { "value": { "selection": { "directReference": {
"structField": { "field": 0 } }, "rootReference": {} } } },
+ { "value": { "selection": { "directReference": {
"structField": { "field": 3 } }, "rootReference": {} } } }
+ ]
+ }
+ }
+ }
+ ]
+ }
+ },
+ "type": "JOIN_TYPE_INNER"
+ }
+ },
+ "names": ["id", "val", "comment", "id0", "val0", "comment0"]
+ }
+ }]
+}
diff --git
a/datafusion/substrait/tests/testdata/test_plans/mixed_join_equal_and_indistinct_left.json
b/datafusion/substrait/tests/testdata/test_plans/mixed_join_equal_and_indistinct_left.json
new file mode 100644
index 0000000000..f16672947e
--- /dev/null
+++
b/datafusion/substrait/tests/testdata/test_plans/mixed_join_equal_and_indistinct_left.json
@@ -0,0 +1,102 @@
+{
+ "extensions": [
+ { "extensionFunction": { "functionAnchor": 0, "name":
"is_not_distinct_from" } },
+ { "extensionFunction": { "functionAnchor": 1, "name": "equal" } },
+ { "extensionFunction": { "functionAnchor": 2, "name": "and" } }
+ ],
+ "relations": [{
+ "root": {
+ "input": {
+ "join": {
+ "common": { "direct": {} },
+ "left": {
+ "read": {
+ "common": { "direct": {} },
+ "baseSchema": {
+ "names": ["id", "val", "comment"],
+ "struct": {
+ "types": [
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } },
+ { "string": { "nullability": "NULLABILITY_NULLABLE" } },
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } }
+ ],
+ "nullability": "NULLABILITY_REQUIRED"
+ }
+ },
+ "virtualTable": {
+ "values": [
+ { "fields": [{ "string": "1", "nullable": false }, {
"string": "a", "nullable": true }, { "string": "c1", "nullable": false }] },
+ { "fields": [{ "string": "2", "nullable": false }, {
"string": "b", "nullable": true }, { "string": "c2", "nullable": false }] },
+ { "fields": [{ "string": "3", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c3", "nullable": false }] },
+ { "fields": [{ "string": "4", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c4", "nullable": false }] },
+ { "fields": [{ "string": "5", "nullable": false }, {
"string": "e", "nullable": true }, { "string": "c5", "nullable": false }] },
+ { "fields": [{ "string": "6", "nullable": false }, {
"string": "f", "nullable": true }, { "string": "c6", "nullable": false }] }
+ ]
+ }
+ }
+ },
+ "right": {
+ "read": {
+ "common": { "direct": {} },
+ "baseSchema": {
+ "names": ["id", "val", "comment"],
+ "struct": {
+ "types": [
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } },
+ { "string": { "nullability": "NULLABILITY_NULLABLE" } },
+ { "string": { "nullability": "NULLABILITY_REQUIRED" } }
+ ],
+ "nullability": "NULLABILITY_REQUIRED"
+ }
+ },
+ "virtualTable": {
+ "values": [
+ { "fields": [{ "string": "1", "nullable": false }, {
"string": "a", "nullable": true }, { "string": "c1", "nullable": false }] },
+ { "fields": [{ "string": "2", "nullable": false }, {
"string": "b", "nullable": true }, { "string": "c2", "nullable": false }] },
+ { "fields": [{ "string": "3", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c3", "nullable": false }] },
+ { "fields": [{ "string": "4", "nullable": false }, { "null":
{ "string": { "nullability": "NULLABILITY_NULLABLE" } }, "nullable": true }, {
"string": "c4", "nullable": false }] },
+ { "fields": [{ "string": "5", "nullable": false }, {
"string": "e", "nullable": true }, { "string": "c5", "nullable": false }] },
+ { "fields": [{ "string": "6", "nullable": false }, {
"string": "f", "nullable": true }, { "string": "c6", "nullable": false }] }
+ ]
+ }
+ }
+ },
+ "expression": {
+ "scalarFunction": {
+ "functionReference": 2,
+ "outputType": { "bool": { "nullability": "NULLABILITY_NULLABLE"
} },
+ "arguments": [
+ {
+ "value": {
+ "scalarFunction": {
+ "functionReference": 0,
+ "outputType": { "bool": { "nullability":
"NULLABILITY_NULLABLE" } },
+ "arguments": [
+ { "value": { "selection": { "directReference": {
"structField": { "field": 1 } }, "rootReference": {} } } },
+ { "value": { "selection": { "directReference": {
"structField": { "field": 4 } }, "rootReference": {} } } }
+ ]
+ }
+ }
+ },
+ {
+ "value": {
+ "scalarFunction": {
+ "functionReference": 1,
+ "outputType": { "bool": { "nullability":
"NULLABILITY_NULLABLE" } },
+ "arguments": [
+ { "value": { "selection": { "directReference": {
"structField": { "field": 0 } }, "rootReference": {} } } },
+ { "value": { "selection": { "directReference": {
"structField": { "field": 3 } }, "rootReference": {} } } }
+ ]
+ }
+ }
+ }
+ ]
+ }
+ },
+ "type": "JOIN_TYPE_LEFT"
+ }
+ },
+ "names": ["id", "val", "comment", "id0", "val0", "comment0"]
+ }
+ }]
+}
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
