This is an automated email from the ASF dual-hosted git repository.
hellostephen 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 5e603dde58c [check](column) Check that the values in ColumnBool are
only 0 or 1. (#59707)
5e603dde58c is described below
commit 5e603dde58ccf5fb662788fc93a7e538ff76e63c
Author: Mryange <[email protected]>
AuthorDate: Tue Jan 13 20:28:15 2026 +0800
[check](column) Check that the values in ColumnBool are only 0 or 1.
(#59707)
---
be/src/vec/columns/column.cpp | 47 +++++++++++++++++++++++++++++++
be/src/vec/columns/column.h | 3 ++
be/test/vec/columns/column_self_check.cpp | 27 ++++++++++++++++++
3 files changed, 77 insertions(+)
diff --git a/be/src/vec/columns/column.cpp b/be/src/vec/columns/column.cpp
index f6f595fc6cd..b206fc17034 100644
--- a/be/src/vec/columns/column.cpp
+++ b/be/src/vec/columns/column.cpp
@@ -62,6 +62,49 @@ bool IColumn::const_nested_check() const {
return const_cnt == 1 && is_column_const(*this);
}
+bool IColumn::column_boolean_check() const {
+ if (const auto* col_nullable =
check_and_get_column<ColumnNullable>(*this)) {
+ // for column nullable, we need to skip null values check
+ const auto& nested_col = col_nullable->get_nested_column();
+ const auto& null_map = col_nullable->get_null_map_data();
+ Filter not_null_filter;
+ not_null_filter.reserve(nested_col.size());
+ size_t result_size_hint = 0;
+ for (size_t i = 0; i < null_map.size(); ++i) {
+ not_null_filter.push_back(null_map[i] == 0);
+ if (null_map[i] == 0) {
+ ++result_size_hint;
+ }
+ }
+ auto nested_col_skip_null = nested_col.filter(not_null_filter,
result_size_hint);
+ return nested_col_skip_null->column_boolean_check();
+ }
+
+ auto check_boolean_is_zero_or_one = [&](const IColumn& subcolumn) {
+ if (const auto* column_boolean =
check_and_get_column<ColumnBool>(subcolumn)) {
+ for (size_t i = 0; i < column_boolean->size(); ++i) {
+ auto val = column_boolean->get_element(i);
+ if (val != 0 && val != 1) {
+ LOG_WARNING("column boolean check failed at index {} with
value {}", i, val)
+ .tag("column structure",
subcolumn.dump_structure());
+ return false;
+ }
+ }
+ }
+ return true;
+ };
+
+ bool is_valid = check_boolean_is_zero_or_one(*this);
+ ColumnCallback callback = [&](ColumnPtr& subcolumn) {
+ if (!subcolumn->column_boolean_check()) {
+ is_valid = false;
+ }
+ };
+ // simply read using for_each_subcolumn without modification; const_cast
can be used.
+ const_cast<IColumn*>(this)->for_each_subcolumn(callback);
+ return is_valid;
+}
+
bool IColumn::null_map_check() const {
auto check_null_map_is_zero_or_one = [&](const IColumn& subcolumn) {
if (is_column_nullable(subcolumn)) {
@@ -100,6 +143,10 @@ Status IColumn::column_self_check() const {
if (!null_map_check()) {
return Status::InternalError("null map check failed for column: {}",
get_name());
}
+ // check boolean column
+ if (!column_boolean_check()) {
+ return Status::InternalError("boolean column check failed for column:
{}", get_name());
+ }
#endif
return Status::OK();
}
diff --git a/be/src/vec/columns/column.h b/be/src/vec/columns/column.h
index 99653bd4acf..e162ee49516 100644
--- a/be/src/vec/columns/column.h
+++ b/be/src/vec/columns/column.h
@@ -679,6 +679,9 @@ public:
// const(array(const(...))) is not allowed
bool const_nested_check() const;
+ // column boolean check, only allow 0 and 1
+ bool column_boolean_check() const;
+
Status column_self_check() const;
// only used in agg value replace for column which is not variable length,
eg.BlockReader::_copy_value_data
diff --git a/be/test/vec/columns/column_self_check.cpp
b/be/test/vec/columns/column_self_check.cpp
index 13fcfcbf5f0..4a4de32c3f1 100644
--- a/be/test/vec/columns/column_self_check.cpp
+++ b/be/test/vec/columns/column_self_check.cpp
@@ -120,4 +120,31 @@ TEST(ColumnSelfCheckTest, nullmap_check_test) {
}
}
+TEST(ColumnSelfCheckTest, boolean_check) {
+ {
+ auto column_bool = ColumnHelper::create_column<DataTypeUInt8>({0, 1,
0, 1, 1});
+ EXPECT_EQ(column_bool->column_boolean_check(), true);
+ }
+ {
+ auto column_bool = ColumnHelper::create_column<DataTypeUInt8>({0, 1,
2, 1, 1});
+ EXPECT_EQ(column_bool->column_boolean_check(), false);
+ }
+
+ {
+ auto column_nullable_bool =
ColumnHelper::create_nullable_column<DataTypeUInt8>(
+ {0, 1, 0, 1, 1}, {0, 0, 0, 0, 0});
+ EXPECT_EQ(column_nullable_bool->column_boolean_check(), true);
+ }
+ {
+ auto column_nullable_bool =
ColumnHelper::create_nullable_column<DataTypeUInt8>(
+ {0, 1, 2, 1, 1}, {0, 0, 0, 0, 0});
+ EXPECT_EQ(column_nullable_bool->column_boolean_check(), false);
+ }
+
+ {
+ auto column_nullable_bool =
ColumnHelper::create_nullable_column<DataTypeUInt8>(
+ {0, 1, 2, 1, 1}, {0, 0, 1, 0, 0});
+ EXPECT_EQ(column_nullable_bool->column_boolean_check(), true);
+ }
+}
} // namespace doris::vectorized
\ No newline at end of file
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
