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 0f73bd6df2d [refactor](be) Use column check helper for type casts
(#63946)
0f73bd6df2d is described below
commit 0f73bd6df2d74e23acf1f5a980d239c102ec07a1
Author: Mryange <[email protected]>
AuthorDate: Wed Jun 3 12:02:46 2026 +0800
[refactor](be) Use column check helper for type casts (#63946)
### What problem does this PR solve?
Problem Summary: Replace direct typeid_cast usage for Doris column type
checks with the column-specific check_and_get_column helper. This keeps
column downcast checks consistent across core column code, expression
evaluation, storage segment code, and related table reader tests without
changing behavior.
### Release note
None
### Check List (For Author)
- Test <!-- At least one of them must be included. -->
- [ ] Regression test
- [ ] Unit Test
- [ ] Manual test (add detailed scripts or steps below)
- [ ] No need to test or manual test. Explain why:
- [ ] This is a refactor/code format and no logic has been changed.
- [ ] Previous test can cover this change.
- [ ] No code files have been changed.
- [ ] Other reason <!-- Add your reason? -->
- Behavior changed:
- [ ] No.
- [ ] Yes. <!-- Explain the behavior change -->
- Does this need documentation?
- [ ] No.
- [ ] Yes. <!-- Add document PR link here. eg:
https://github.com/apache/doris-website/pull/1214 -->
### Check List (For Reviewer who merge this PR)
- [ ] Confirm the release note
- [ ] Confirm test cases
- [ ] Confirm document
- [ ] Add branch pick label <!-- Add branch pick label that this PR
should merge into -->
---
be/src/core/column/column_array.cpp | 63 +++++++++++-----------
be/src/core/column/column_const.cpp | 2 +-
be/src/core/column/column_const.h | 2 +-
be/src/core/column/column_decimal.h | 2 +-
be/src/core/column/column_nullable.h | 2 +-
be/src/core/column/column_struct.cpp | 2 +-
be/src/exprs/function/if.cpp | 2 +-
be/src/exprs/vcondition_expr.cpp | 4 +-
.../storage/segment/row_binlog_segment_writer.cpp | 4 +-
be/src/storage/segment/segment_iterator.cpp | 6 +--
be/test/format/table/hive/hive_reader_test.cpp | 8 +--
.../format/table/iceberg/iceberg_reader_test.cpp | 8 +--
12 files changed, 55 insertions(+), 50 deletions(-)
diff --git a/be/src/core/column/column_array.cpp
b/be/src/core/column/column_array.cpp
index 534bdf8e4d6..604c4694853 100644
--- a/be/src/core/column/column_array.cpp
+++ b/be/src/core/column/column_array.cpp
@@ -50,7 +50,8 @@ namespace doris {
namespace {
const ColumnArray::ColumnOffsets& check_array_offsets_column(const IColumn&
offsets_column) {
- const auto* offsets_concrete = typeid_cast<const
ColumnArray::ColumnOffsets*>(&offsets_column);
+ const auto* offsets_concrete =
+ check_and_get_column<ColumnArray::ColumnOffsets>(&offsets_column);
if (!offsets_concrete) {
throw doris::Exception(ErrorCode::INTERNAL_ERROR, "offsets_column must
be a ColumnUInt64");
__builtin_unreachable();
@@ -851,35 +852,35 @@ size_t filter_generic_inplace(const Filter& filter,
IColumn& src_data, ColumnOff
ColumnArrayDataOffsets filter_return_new_dispatch(const Filter& filt, ssize_t
result_size_hint,
const ColumnPtr& data,
const ColumnOffsets*
offsets) {
- if (typeid_cast<const ColumnUInt8*>(data.get()))
+ if (check_and_get_column<ColumnUInt8>(data.get()))
return filter_number_return_new<TYPE_BOOLEAN>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnInt8*>(data.get()))
+ if (check_and_get_column<ColumnInt8>(data.get()))
return filter_number_return_new<TYPE_TINYINT>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnInt16*>(data.get()))
+ if (check_and_get_column<ColumnInt16>(data.get()))
return filter_number_return_new<TYPE_SMALLINT>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnInt32*>(data.get()))
+ if (check_and_get_column<ColumnInt32>(data.get()))
return filter_number_return_new<TYPE_INT>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnInt64*>(data.get()))
+ if (check_and_get_column<ColumnInt64>(data.get()))
return filter_number_return_new<TYPE_BIGINT>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnFloat32*>(data.get()))
+ if (check_and_get_column<ColumnFloat32>(data.get()))
return filter_number_return_new<TYPE_FLOAT>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnFloat64*>(data.get()))
+ if (check_and_get_column<ColumnFloat64>(data.get()))
return filter_number_return_new<TYPE_DOUBLE>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnDate*>(data.get()))
+ if (check_and_get_column<ColumnDate>(data.get()))
return filter_number_return_new<TYPE_DATE>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnDateV2*>(data.get()))
+ if (check_and_get_column<ColumnDateV2>(data.get()))
return filter_number_return_new<TYPE_DATEV2>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnDateTime*>(data.get()))
+ if (check_and_get_column<ColumnDateTime>(data.get()))
return filter_number_return_new<TYPE_DATETIME>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnDateTimeV2*>(data.get()))
+ if (check_and_get_column<ColumnDateTimeV2>(data.get()))
return filter_number_return_new<TYPE_DATETIMEV2>(filt,
result_size_hint, data, offsets);
- if (typeid_cast<const ColumnTimeStampTz*>(data.get()))
+ if (check_and_get_column<ColumnTimeStampTz>(data.get()))
return filter_number_return_new<TYPE_TIMESTAMPTZ>(filt,
result_size_hint, data, offsets);
- if (typeid_cast<const ColumnTimeV2*>(data.get()))
+ if (check_and_get_column<ColumnTimeV2>(data.get()))
return filter_number_return_new<TYPE_TIMEV2>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnIPv4*>(data.get()))
+ if (check_and_get_column<ColumnIPv4>(data.get()))
return filter_number_return_new<TYPE_IPV4>(filt, result_size_hint,
data, offsets);
- if (typeid_cast<const ColumnString*>(data.get()))
+ if (check_and_get_column<ColumnString>(data.get()))
return filter_string_return_new(filt, result_size_hint, data, offsets);
return filter_generic_return_new(filt, result_size_hint, data, offsets);
}
@@ -917,33 +918,33 @@ ColumnPtr ColumnArray::filter(const Filter& filt, ssize_t
result_size_hint) cons
// In order not to destroy the original logic, the original logic is still
retained here.
size_t filter_inplace_dispatch(const Filter& filter, IColumn& src_data,
ColumnOffsets& src_offsets) {
- if (typeid_cast<ColumnUInt8*>(&src_data))
+ if (check_and_get_column<ColumnUInt8>(&src_data))
return filter_number_inplace<TYPE_BOOLEAN>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnInt8*>(&src_data))
+ if (check_and_get_column<ColumnInt8>(&src_data))
return filter_number_inplace<TYPE_TINYINT>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnInt16*>(&src_data))
+ if (check_and_get_column<ColumnInt16>(&src_data))
return filter_number_inplace<TYPE_SMALLINT>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnInt32*>(&src_data))
+ if (check_and_get_column<ColumnInt32>(&src_data))
return filter_number_inplace<TYPE_INT>(filter, src_data, src_offsets);
- if (typeid_cast<ColumnInt64*>(&src_data))
+ if (check_and_get_column<ColumnInt64>(&src_data))
return filter_number_inplace<TYPE_BIGINT>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnFloat32*>(&src_data))
+ if (check_and_get_column<ColumnFloat32>(&src_data))
return filter_number_inplace<TYPE_FLOAT>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnFloat64*>(&src_data))
+ if (check_and_get_column<ColumnFloat64>(&src_data))
return filter_number_inplace<TYPE_DOUBLE>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnDate*>(&src_data))
+ if (check_and_get_column<ColumnDate>(&src_data))
return filter_number_inplace<TYPE_DATE>(filter, src_data, src_offsets);
- if (typeid_cast<ColumnDateV2*>(&src_data))
+ if (check_and_get_column<ColumnDateV2>(&src_data))
return filter_number_inplace<TYPE_DATEV2>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnDateTime*>(&src_data))
+ if (check_and_get_column<ColumnDateTime>(&src_data))
return filter_number_inplace<TYPE_DATETIME>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnDateTimeV2*>(&src_data))
+ if (check_and_get_column<ColumnDateTimeV2>(&src_data))
return filter_number_inplace<TYPE_DATETIMEV2>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnTimeStampTz*>(&src_data))
+ if (check_and_get_column<ColumnTimeStampTz>(&src_data))
return filter_number_inplace<TYPE_TIMESTAMPTZ>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnTimeV2*>(&src_data))
+ if (check_and_get_column<ColumnTimeV2>(&src_data))
return filter_number_inplace<TYPE_TIMEV2>(filter, src_data,
src_offsets);
- if (typeid_cast<ColumnIPv4*>(&src_data))
+ if (check_and_get_column<ColumnIPv4>(&src_data))
return filter_number_inplace<TYPE_IPV4>(filter, src_data, src_offsets);
return filter_generic_inplace(filter, src_data, src_offsets);
}
@@ -953,7 +954,7 @@ size_t ColumnArray::filter(const Filter& filter) {
return 0;
}
- if (auto* nullable_data_column = typeid_cast<ColumnNullable*>(data.get()))
{
+ if (auto* nullable_data_column =
check_and_get_column<ColumnNullable>(data.get())) {
const auto result_size = filter_arrays_impl_only_data(
nullable_data_column->get_null_map_data(), get_offsets(),
filter);
diff --git a/be/src/core/column/column_const.cpp
b/be/src/core/column/column_const.cpp
index 7688f5c7a72..be719ef7350 100644
--- a/be/src/core/column/column_const.cpp
+++ b/be/src/core/column/column_const.cpp
@@ -35,7 +35,7 @@ namespace doris {
ColumnPtr squash_const(const ColumnPtr& col) {
ColumnPtr res = col;
- while (const auto* c = typeid_cast<const ColumnConst*>(res.get())) {
+ while (const auto* c = check_and_get_column<ColumnConst>(res.get())) {
res = c->get_data_column_ptr();
}
return res;
diff --git a/be/src/core/column/column_const.h
b/be/src/core/column/column_const.h
index b213aeda0ff..bda7ea13fb1 100644
--- a/be/src/core/column/column_const.h
+++ b/be/src/core/column/column_const.h
@@ -264,7 +264,7 @@ public:
void for_each_subcolumn(ColumnCallback callback) override {
callback(data); }
bool structure_equals(const IColumn& rhs) const override {
- if (const auto* rhs_concrete = typeid_cast<const ColumnConst*>(&rhs)) {
+ if (const auto* rhs_concrete =
check_and_get_column<ColumnConst>(&rhs)) {
return data->structure_equals(*rhs_concrete->data);
}
return false;
diff --git a/be/src/core/column/column_decimal.h
b/be/src/core/column/column_decimal.h
index 29f44a55f07..3c89bd3d77b 100644
--- a/be/src/core/column/column_decimal.h
+++ b/be/src/core/column/column_decimal.h
@@ -204,7 +204,7 @@ public:
MutableColumnPtr permute(const IColumn::Permutation& perm, size_t limit)
const override;
bool structure_equals(const IColumn& rhs) const override {
- if (auto rhs_concrete = typeid_cast<const ColumnDecimal<T>*>(&rhs))
+ if (auto rhs_concrete = check_and_get_column<ColumnDecimal<T>>(&rhs))
return scale == rhs_concrete->scale;
return false;
}
diff --git a/be/src/core/column/column_nullable.h
b/be/src/core/column/column_nullable.h
index 73bab4ff156..bf924fcb2ec 100644
--- a/be/src/core/column/column_nullable.h
+++ b/be/src/core/column/column_nullable.h
@@ -261,7 +261,7 @@ public:
}
bool structure_equals(const IColumn& rhs) const override {
- if (const auto* rhs_nullable = typeid_cast<const
ColumnNullable*>(&rhs)) {
+ if (const auto* rhs_nullable =
check_and_get_column<ColumnNullable>(&rhs)) {
return
_nested_column->structure_equals(*rhs_nullable->_nested_column);
}
return false;
diff --git a/be/src/core/column/column_struct.cpp
b/be/src/core/column/column_struct.cpp
index 57463f6c435..e2a90432c56 100644
--- a/be/src/core/column/column_struct.cpp
+++ b/be/src/core/column/column_struct.cpp
@@ -386,7 +386,7 @@ void ColumnStruct::for_each_subcolumn(ColumnCallback
callback) {
}
bool ColumnStruct::structure_equals(const IColumn& rhs) const {
- if (const auto* rhs_tuple = typeid_cast<const ColumnStruct*>(&rhs)) {
+ if (const auto* rhs_tuple = check_and_get_column<ColumnStruct>(&rhs)) {
const size_t tuple_size = columns.size();
if (tuple_size != rhs_tuple->columns.size()) {
return false;
diff --git a/be/src/exprs/function/if.cpp b/be/src/exprs/function/if.cpp
index d571eae5a91..7e543499ada 100644
--- a/be/src/exprs/function/if.cpp
+++ b/be/src/exprs/function/if.cpp
@@ -240,7 +240,7 @@ public:
return Status::OK();
}
- const auto* cond_col = typeid_cast<const
ColumnUInt8*>(arg_cond.column.get());
+ const auto* cond_col =
check_and_get_column<ColumnUInt8>(arg_cond.column.get());
const ColumnConst* cond_const_col =
check_and_get_column_const<ColumnUInt8>(arg_cond.column.get());
diff --git a/be/src/exprs/vcondition_expr.cpp b/be/src/exprs/vcondition_expr.cpp
index 13e196a5e0b..7ef8ae98ae2 100644
--- a/be/src/exprs/vcondition_expr.cpp
+++ b/be/src/exprs/vcondition_expr.cpp
@@ -194,7 +194,7 @@ Status VectorizedIfExpr::execute_for_null_then_else(Block&
block,
return Status::OK();
}
- const auto* cond_col = typeid_cast<const
ColumnUInt8*>(arg_cond.column.get());
+ const auto* cond_col =
check_and_get_column<ColumnUInt8>(arg_cond.column.get());
const ColumnConst* cond_const_col =
check_and_get_column_const<ColumnUInt8>(arg_cond.column.get());
@@ -781,4 +781,4 @@ Status
VectorizedCoalesceExpr::execute_column_impl(VExprContext* context, const
return Status::OK();
}
-} // namespace doris
\ No newline at end of file
+} // namespace doris
diff --git a/be/src/storage/segment/row_binlog_segment_writer.cpp
b/be/src/storage/segment/row_binlog_segment_writer.cpp
index ccfc53d9fe9..2cc23fc884b 100644
--- a/be/src/storage/segment/row_binlog_segment_writer.cpp
+++ b/be/src/storage/segment/row_binlog_segment_writer.cpp
@@ -312,7 +312,7 @@ Status RowBinlogSegmentWriter::_fill_binlog_columns(size_t
num_rows,
// wrong op only happens when partial-update, it will be fixed by
delete bitmap when publish
const FieldType op_col_type =
_tablet_schema->column(binlog_cids[1]).type();
IColumn* op_col_ptr = binlog_prefix_columns[1].get();
- auto* op_nullable_column = typeid_cast<ColumnNullable*>(op_col_ptr);
+ auto* op_nullable_column =
check_and_get_column<ColumnNullable>(op_col_ptr);
IColumn* op_nested_column = op_nullable_column != nullptr
?
&op_nullable_column->get_nested_column()
: op_col_ptr;
@@ -329,7 +329,7 @@ Status RowBinlogSegmentWriter::_fill_binlog_columns(size_t
num_rows,
// we can't get correct timestamp when commit
IColumn* ts_col_ptr = binlog_prefix_columns[2].get();
auto timestamp = UnixMillis();
- auto* ts_nullable_column = typeid_cast<ColumnNullable*>(ts_col_ptr);
+ auto* ts_nullable_column =
check_and_get_column<ColumnNullable>(ts_col_ptr);
if (ts_nullable_column != nullptr) {
assert_cast<ColumnInt64*>(&ts_nullable_column->get_nested_column())
->insert_many_vals(timestamp, num_rows);
diff --git a/be/src/storage/segment/segment_iterator.cpp
b/be/src/storage/segment/segment_iterator.cpp
index e81b523a94c..dc6930777f5 100644
--- a/be/src/storage/segment/segment_iterator.cpp
+++ b/be/src/storage/segment/segment_iterator.cpp
@@ -2566,8 +2566,8 @@ void SegmentIterator::_update_tso_col_if_needed(const
std::vector<ColumnId>& col
if (_is_pred_column[tso_col_idx]) {
// Nullable predicate column is represented as
ColumnNullable(predicate_col)
- if (auto* tso_nullable =
-
typeid_cast<ColumnNullable*>(_current_return_columns[tso_col_idx].get())) {
+ if (auto* tso_nullable = check_and_get_column<ColumnNullable>(
+ _current_return_columns[tso_col_idx].get())) {
_current_return_columns[tso_col_idx]->clear();
auto value = commit_time;
for (size_t j = 0; j < num_rows; j++) {
@@ -2592,7 +2592,7 @@ void SegmentIterator::_update_tso_col_if_needed(const
std::vector<ColumnId>& col
auto column = Schema::get_data_type_ptr(*column_desc)->create_column();
DCHECK(column_desc->type() == FieldType::OLAP_FIELD_TYPE_BIGINT);
- if (auto* tso_nullable = typeid_cast<ColumnNullable*>(column.get())) {
+ if (auto* tso_nullable =
check_and_get_column<ColumnNullable>(column.get())) {
auto* col_ptr =
assert_cast<ColumnInt64*>(&tso_nullable->get_nested_column());
for (size_t j = 0; j < num_rows; j++) {
col_ptr->insert_value(commit_time);
diff --git a/be/test/format/table/hive/hive_reader_test.cpp
b/be/test/format/table/hive/hive_reader_test.cpp
index 7746cbffa60..31932c2e9c2 100644
--- a/be/test/format/table/hive/hive_reader_test.cpp
+++ b/be/test/format/table/hive/hive_reader_test.cpp
@@ -413,7 +413,8 @@ protected:
EXPECT_GT(col->size(), 0) << name << " column/subcolumn
size should be > 0";
// Check if it's a nullable column
- if (const auto* nullable_col = typeid_cast<const
ColumnNullable*>(col.get())) {
+ if (const auto* nullable_col =
+
check_and_get_column<ColumnNullable>(col.get())) {
auto nested_type =
assert_cast<const
DataTypeNullable*>(type.get())->get_nested_type();
@@ -430,7 +431,8 @@ protected:
nested_name, depth +
(is_complex_type ? 1 : 0));
}
// Check if it's a struct column
- else if (const auto* struct_col = typeid_cast<const
ColumnStruct*>(col.get())) {
+ else if (const auto* struct_col =
+
check_and_get_column<ColumnStruct>(col.get())) {
auto struct_type = assert_cast<const
DataTypeStruct*>(type.get());
for (size_t i = 0; i < struct_col->tuple_size(); ++i) {
std::string field_name =
struct_type->get_element_name(i);
@@ -440,7 +442,7 @@ protected:
}
}
// Check if it's an array column
- else if (const auto* array_col = typeid_cast<const
ColumnArray*>(col.get())) {
+ else if (const auto* array_col =
check_and_get_column<ColumnArray>(col.get())) {
auto array_type = assert_cast<const
DataTypeArray*>(type.get());
auto element_type = array_type->get_nested_type();
print_column_rows(array_col->get_data_ptr(),
element_type, name + ".data",
diff --git a/be/test/format/table/iceberg/iceberg_reader_test.cpp
b/be/test/format/table/iceberg/iceberg_reader_test.cpp
index fb35b78c617..c2618597312 100644
--- a/be/test/format/table/iceberg/iceberg_reader_test.cpp
+++ b/be/test/format/table/iceberg/iceberg_reader_test.cpp
@@ -473,7 +473,8 @@ protected:
EXPECT_GT(col->size(), 0) << name << " column/subcolumn
size should be > 0";
// Check if it's a nullable column
- if (const auto* nullable_col = typeid_cast<const
ColumnNullable*>(col.get())) {
+ if (const auto* nullable_col =
+
check_and_get_column<ColumnNullable>(col.get())) {
auto nested_type =
assert_cast<const
DataTypeNullable*>(type.get())->get_nested_type();
@@ -490,7 +491,8 @@ protected:
nested_name, depth +
(is_complex_type ? 1 : 0));
}
// Check if it's a struct column
- else if (const auto* struct_col = typeid_cast<const
ColumnStruct*>(col.get())) {
+ else if (const auto* struct_col =
+
check_and_get_column<ColumnStruct>(col.get())) {
auto struct_type = assert_cast<const
DataTypeStruct*>(type.get());
for (size_t i = 0; i < struct_col->tuple_size(); ++i) {
std::string field_name =
struct_type->get_element_name(i);
@@ -500,7 +502,7 @@ protected:
}
}
// Check if it's an array column
- else if (const auto* array_col = typeid_cast<const
ColumnArray*>(col.get())) {
+ else if (const auto* array_col =
check_and_get_column<ColumnArray>(col.get())) {
auto array_type = assert_cast<const
DataTypeArray*>(type.get());
auto element_type = array_type->get_nested_type();
print_column_rows(array_col->get_data_ptr(),
element_type, name + ".data",
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
