This is an automated email from the ASF dual-hosted git repository.
yiguolei pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris.git
The following commit(s) were added to refs/heads/master by this push:
new 47955f9109f [bug](intersect) fix intersect node nullable field should
depend on _row_descriptor (#26038)
47955f9109f is described below
commit 47955f9109fb6fdf955c86911c09d8746156e29e
Author: zhangstar333 <87313068+zhangstar...@users.noreply.github.com>
AuthorDate: Sat Oct 28 22:39:36 2023 +0800
[bug](intersect) fix intersect node nullable field should depend on
_row_descriptor (#26038)
---
be/src/pipeline/exec/set_source_operator.cpp | 16 +++---
be/src/vec/exec/vset_operation_node.cpp | 16 +++---
.../data/correctness_p0/test_set_operation.out | 8 +++
.../correctness_p0/test_set_operation.groovy | 58 ++++++++++++++++++++++
4 files changed, 86 insertions(+), 12 deletions(-)
diff --git a/be/src/pipeline/exec/set_source_operator.cpp
b/be/src/pipeline/exec/set_source_operator.cpp
index 31d372a002a..e679274d92c 100644
--- a/be/src/pipeline/exec/set_source_operator.cpp
+++ b/be/src/pipeline/exec/set_source_operator.cpp
@@ -67,13 +67,17 @@ Status
SetSourceLocalState<is_intersect>::open(RuntimeState* state) {
SCOPED_TIMER(_open_timer);
RETURN_IF_ERROR(PipelineXLocalState<SetDependency>::open(state));
auto& child_exprs_lists = _shared_state->child_exprs_lists;
- vector<bool> nullable_flags;
- nullable_flags.resize(child_exprs_lists[0].size(), false);
- for (int i = 0; i < child_exprs_lists.size(); ++i) {
- for (int j = 0; j < child_exprs_lists[i].size(); ++j) {
- nullable_flags[j] = nullable_flags[j] ||
child_exprs_lists[i][j]->root()->is_nullable();
- }
+ auto output_data_types = vectorized::VectorizedUtils::get_data_types(
+ _parent->cast<SetSourceOperatorX<is_intersect>>()._row_descriptor);
+ auto column_nums = child_exprs_lists[0].size();
+ DCHECK_EQ(output_data_types.size(), column_nums)
+ << output_data_types.size() << " " << column_nums;
+ // the nullable is not depend on child, it's should use _row_descriptor
from FE plan
+ // some case all not nullable column from children, but maybe need output
nullable.
+ vector<bool> nullable_flags(column_nums, false);
+ for (int i = 0; i < column_nums; ++i) {
+ nullable_flags[i] = output_data_types[i]->is_nullable();
}
_left_table_data_types.clear();
diff --git a/be/src/vec/exec/vset_operation_node.cpp
b/be/src/vec/exec/vset_operation_node.cpp
index f12669336cf..791e92679d0 100644
--- a/be/src/vec/exec/vset_operation_node.cpp
+++ b/be/src/vec/exec/vset_operation_node.cpp
@@ -160,14 +160,18 @@ Status
VSetOperationNode<is_intersect>::prepare(RuntimeState* state) {
_build_timer = ADD_TIMER(runtime_profile(), "BuildTime");
_probe_timer = ADD_TIMER(runtime_profile(), "ProbeTime");
_pull_timer = ADD_TIMER(runtime_profile(), "PullTime");
+ auto output_data_types = VectorizedUtils::get_data_types(_row_descriptor);
+ auto column_nums = _child_expr_lists[0].size();
+ DCHECK_EQ(output_data_types.size(), column_nums)
+ << output_data_types.size() << " " << column_nums;
+ // the nullable is not depend on child, it's should use _row_descriptor
from FE plan
+ // some case all not nullable column from children, but maybe need output
nullable.
+ vector<bool> nullable_flags(column_nums, false);
+ for (int i = 0; i < column_nums; ++i) {
+ nullable_flags[i] = output_data_types[i]->is_nullable();
+ }
- // Prepare result expr lists.
- vector<bool> nullable_flags;
- nullable_flags.resize(_child_expr_lists[0].size(), false);
for (int i = 0; i < _child_expr_lists.size(); ++i) {
- for (int j = 0; j < _child_expr_lists[i].size(); ++j) {
- nullable_flags[j] = nullable_flags[j] ||
_child_expr_lists[i][j]->root()->is_nullable();
- }
RETURN_IF_ERROR(VExpr::prepare(_child_expr_lists[i], state,
child(i)->row_desc()));
}
for (int i = 0; i < _child_expr_lists[0].size(); ++i) {
diff --git a/regression-test/data/correctness_p0/test_set_operation.out
b/regression-test/data/correctness_p0/test_set_operation.out
index 095c7b20356..f6d9f32d4cc 100644
--- a/regression-test/data/correctness_p0/test_set_operation.out
+++ b/regression-test/data/correctness_p0/test_set_operation.out
@@ -2,3 +2,11 @@
-- !sql --
1
+-- !select1 --
+aaaa
+bbbb
+
+-- !select1 --
+aaaa
+bbbb
+
diff --git a/regression-test/suites/correctness_p0/test_set_operation.groovy
b/regression-test/suites/correctness_p0/test_set_operation.groovy
index cd366f4d2bf..e3e28a9e3a4 100644
--- a/regression-test/suites/correctness_p0/test_set_operation.groovy
+++ b/regression-test/suites/correctness_p0/test_set_operation.groovy
@@ -67,4 +67,62 @@ suite("test_set_operation") {
WHERE a = 3 ) ) ) ) c
ORDER BY one_uid;
"""
+
+ sql """
+ drop table if exists test_A;
+ """
+
+ sql """
+ drop table if exists test_B;
+ """
+
+ sql """
+ CREATE TABLE `test_A` (
+ `stat_day` varchar(200) NOT NULL ,
+ `sku_code` varchar(200) NOT NULL ,
+ `site_code` varchar(200) NOT NULL )
+ ENGINE = OLAP DUPLICATE KEY(`stat_day`) COMMENT ''
+ DISTRIBUTED BY HASH(`stat_day`,`sku_code`,`site_code`) BUCKETS 16
+ PROPERTIES (
+ "replication_allocation" = "tag.location.default: 1",
+ "is_being_synced" = "false",
+ "storage_format" = "V2",
+ "disable_auto_compaction" = "false",
+ "enable_single_replica_compaction" = "false" );
+ """
+
+
+ sql """
+ CREATE TABLE `test_B` (
+ `stat_day` varchar(200) NULL,
+ `sku_code` varchar(200) NULL )
+ ENGINE = OLAP DUPLICATE KEY(`stat_day`) COMMENT '-'
+ DISTRIBUTED BY HASH(`stat_day`,`sku_code`) BUCKETS 16
+ PROPERTIES (
+ "replication_allocation" = "tag.location.default: 1",
+ "is_being_synced" = "false",
+ "storage_format" = "V2",
+ "disable_auto_compaction" = "false",
+ "enable_single_replica_compaction" = "false" );
+ """
+
+ sql """
+ insert into test_A values("aa","aaaa","aaaaaa");
+ """
+
+ sql """
+ insert into test_B values("bb","bbbb");
+ """
+
+ sql """
+ set experimental_enable_nereids_planner = false;
+ """
+
+ qt_select1 """ SELECT DISTINCT * FROM((SELECT sku_code FROM test_B)
INTERSECT (SELECT sku_code FROM test_B) UNION (SELECT sku_code FROM test_A)) as
t order by 1; """
+
+ sql """
+ set experimental_enable_nereids_planner = true;
+ """
+ qt_select1 """ SELECT DISTINCT * FROM((SELECT sku_code FROM test_B)
INTERSECT (SELECT sku_code FROM test_B) UNION (SELECT sku_code FROM test_A)) as
t order by 1; """
+
}
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@doris.apache.org
For additional commands, e-mail: commits-h...@doris.apache.org
