yiguolei commented on code in PR #52676:
URL: https://github.com/apache/doris/pull/52676#discussion_r2181617191
##########
be/src/vec/data_types/serde/data_type_object_serde.cpp:
##########
@@ -210,27 +210,50 @@ Status DataTypeVariantSerDe::write_column_to_orc(const
std::string& timezone, co
std::vector<StringRef>&
buffer_list) const {
const auto* var = check_and_get_column<ColumnVariant>(column);
orc::StringVectorBatch* cur_batch =
dynamic_cast<orc::StringVectorBatch*>(orc_col_batch);
-
- INIT_MEMORY_FOR_ORC_WRITER()
-
+ // First pass: calculate total memory needed and collect serialized values
+ std::vector<std::string> serialized_values;
+ std::vector<size_t> valid_row_indices;
+ size_t total_size = 0;
for (size_t row_id = start; row_id < end; row_id++) {
if (cur_batch->notNull[row_id] == 1) {
- auto serialized_value = std::make_unique<std::string>();
- if (!var->serialize_one_row_to_string(row_id,
serialized_value.get())) {
+ std::string* serialized_value;
+ if (!var->serialize_one_row_to_string(row_id, serialized_value)) {
throw doris::Exception(ErrorCode::INTERNAL_ERROR, "Failed to
serialize variant {}",
var->dump_structure());
}
- auto len = serialized_value->length();
-
- REALLOC_MEMORY_FOR_ORC_WRITER()
-
- memcpy(const_cast<char*>(bufferRef.data) + offset,
serialized_value->data(), len);
- cur_batch->data[row_id] = const_cast<char*>(bufferRef.data) +
offset;
- cur_batch->length[row_id] = len;
- offset += len;
+ serialized_values.push_back(*serialized_value);
Review Comment:
serialized_values.emplace_back(std::move(serialized_value);
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
