This is an automated email from the ASF dual-hosted git repository.
zclllyybb 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 87316004891 [fix](be) Fix timestamptz group_array state serde (#63827)
87316004891 is described below
commit 87316004891ab8d32f107b353c48bc2b65625425
Author: Mryange <[email protected]>
AuthorDate: Fri May 29 10:08:17 2026 +0800
[fix](be) Fix timestamptz group_array state serde (#63827)
Fix collect_list/group_array on nested TIMESTAMPTZ values when complex
aggregate state is serialized through JSON. This keeps the existing
state format for compatibility, provides a UTC timezone during serde,
and adds regression coverage for the nested group_array case.
---
.../exprs/aggregate/aggregate_function_collect.h | 7 ++++
.../timestamptz/test_timestamptz_agg_functions.out | 3 ++
.../test_timestamptz_agg_functions.groovy | 37 ++++++++++++++++++++++
3 files changed, 47 insertions(+)
diff --git a/be/src/exprs/aggregate/aggregate_function_collect.h
b/be/src/exprs/aggregate/aggregate_function_collect.h
index 7eef3cff85e..ef8bb0437f7 100644
--- a/be/src/exprs/aggregate/aggregate_function_collect.h
+++ b/be/src/exprs/aggregate/aggregate_function_collect.h
@@ -17,6 +17,7 @@
#pragma once
+#include <cctz/time_zone.h>
#include <glog/logging.h>
#include <cstddef>
@@ -342,6 +343,10 @@ struct AggregateFunctionCollectListData<T, HasLimit> {
buf.write_binary(size);
DataTypeSerDe::FormatOptions opt;
+ auto timezone = cctz::utc_time_zone();
+ opt.timezone = &timezone;
+ // TODO: Refactor this aggregate state serialization to avoid
+ // round-tripping through a human-readable string format.
auto tmp_str = ColumnString::create();
VectorBufferWriter tmp_buf(*tmp_str.get());
@@ -367,6 +372,8 @@ struct AggregateFunctionCollectListData<T, HasLimit> {
StringRef s;
DataTypeSerDe::FormatOptions opt;
+ auto timezone = cctz::utc_time_zone();
+ opt.timezone = &timezone;
for (size_t i = 0; i < size; i++) {
buf.read_binary(s);
Slice slice(s.data, s.size);
diff --git
a/regression-test/data/datatype_p0/timestamptz/test_timestamptz_agg_functions.out
b/regression-test/data/datatype_p0/timestamptz/test_timestamptz_agg_functions.out
index 850cbe14a98..f7ff2eb36d0 100644
---
a/regression-test/data/datatype_p0/timestamptz/test_timestamptz_agg_functions.out
+++
b/regression-test/data/datatype_p0/timestamptz/test_timestamptz_agg_functions.out
@@ -11,3 +11,6 @@ true
-- !group_array_union --
3
+-- !group_array_nested_timestamptz --
+[["2024-01-01 00:00:00.000000+00:00", "2024-01-01 00:00:00.000000+00:00",
"2024-01-02 00:00:00.000000+00:00"], ["2024-01-01 00:00:00.000000+00:00",
"2024-01-02 00:00:00.000000+00:00", "2024-01-03 00:00:00.000000+00:00"]]
+
diff --git
a/regression-test/suites/datatype_p0/timestamptz/test_timestamptz_agg_functions.groovy
b/regression-test/suites/datatype_p0/timestamptz/test_timestamptz_agg_functions.groovy
index 89126b5a284..e5bf945225e 100644
---
a/regression-test/suites/datatype_p0/timestamptz/test_timestamptz_agg_functions.groovy
+++
b/regression-test/suites/datatype_p0/timestamptz/test_timestamptz_agg_functions.groovy
@@ -56,4 +56,41 @@ suite("test_timestamptz_agg_functions", "datatype_p0") {
qt_group_array_union "SELECT size(group_array_union(arr)) FROM test_tz_agg"
sql "DROP TABLE IF EXISTS test_tz_agg"
+
+ sql "DROP TABLE IF EXISTS tz_group_array_crash"
+ sql """
+ CREATE TABLE tz_group_array_crash (
+ grp INT,
+ arr ARRAY<TIMESTAMPTZ(6)>
+ )
+ DUPLICATE KEY(grp)
+ DISTRIBUTED BY HASH(grp) BUCKETS 1
+ PROPERTIES('replication_num' = '1')
+ """
+
+ sql """
+ INSERT INTO tz_group_array_crash VALUES
+ (
+ 1,
+ ARRAY(
+ CAST('2024-01-01 00:00:00 +00:00' AS TIMESTAMPTZ(6)),
+ CAST('2024-01-01 08:00:00 +08:00' AS TIMESTAMPTZ(6)),
+ CAST('2024-01-02 00:00:00 +00:00' AS TIMESTAMPTZ(6))
+ )
+ ),
+ (
+ 1,
+ ARRAY(
+ CAST('2024-01-01 00:00:00 +00:00' AS TIMESTAMPTZ(6)),
+ CAST('2024-01-02 08:00:00 +08:00' AS TIMESTAMPTZ(6)),
+ CAST('2024-01-03 00:00:00 +00:00' AS TIMESTAMPTZ(6))
+ )
+ )
+ """
+
+ qt_group_array_nested_timestamptz """
+ SELECT CAST(array_sort(group_array(arr)) AS STRING)
+ FROM tz_group_array_crash
+ GROUP BY grp
+ """
}
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
