github-actions[bot] commented on code in PR #35628: URL: https://github.com/apache/doris/pull/35628#discussion_r1620524841
########## be/src/olap/rowset/segment_v2/column_writer.cpp: ########## @@ -88,267 +88,305 @@ RleEncoder<bool> _rle_encoder; }; +Status ColumnWriter::create_struct_writer(const ColumnWriterOptions& opts, + const TabletColumn* column, io::FileWriter* file_writer, + std::unique_ptr<ColumnWriter>* writer) { + // not support empty struct + DCHECK(column->get_subtype_count() >= 1); + std::vector<std::unique_ptr<ColumnWriter>> sub_column_writers; + sub_column_writers.reserve(column->get_subtype_count()); + for (uint32_t i = 0; i < column->get_subtype_count(); i++) { + const TabletColumn& sub_column = column->get_sub_column(i); + + // create sub writer + ColumnWriterOptions column_options; + column_options.meta = opts.meta->mutable_children_columns(i); + column_options.need_zone_map = false; + column_options.need_bloom_filter = sub_column.is_bf_column(); + column_options.need_bitmap_index = sub_column.has_bitmap_index(); + if (sub_column.type() == FieldType::OLAP_FIELD_TYPE_STRUCT) { + if (column_options.need_bloom_filter) { + return Status::NotSupported("Do not support bloom filter for struct type"); + } + if (column_options.need_bitmap_index) { + return Status::NotSupported("Do not support bitmap index for struct type"); + } + } + if (sub_column.type() == FieldType::OLAP_FIELD_TYPE_ARRAY) { + if (column_options.need_bloom_filter) { + return Status::NotSupported("Do not support bloom filter for array type"); + } + if (column_options.need_bitmap_index) { + return Status::NotSupported("Do not support bitmap index for array type"); + } + } + std::unique_ptr<ColumnWriter> sub_column_writer; + RETURN_IF_ERROR( + ColumnWriter::create(column_options, &sub_column, file_writer, &sub_column_writer)); + sub_column_writers.push_back(std::move(sub_column_writer)); + } + + // if nullable, create null writer + ScalarColumnWriter* null_writer = nullptr; + if (opts.meta->is_nullable()) { + FieldType null_type = FieldType::OLAP_FIELD_TYPE_TINYINT; + ColumnWriterOptions null_options; + null_options.meta = opts.meta->add_children_columns(); + null_options.meta->set_column_id(column->get_subtype_count() + 1); + null_options.meta->set_unique_id(column->get_subtype_count() + 1); + null_options.meta->set_type(int(null_type)); + null_options.meta->set_is_nullable(false); + null_options.meta->set_length( + get_scalar_type_info<FieldType::OLAP_FIELD_TYPE_TINYINT>()->size()); + null_options.meta->set_encoding(DEFAULT_ENCODING); + null_options.meta->set_compression(opts.meta->compression()); + + null_options.need_zone_map = false; + null_options.need_bloom_filter = false; + null_options.need_bitmap_index = false; + + TabletColumn null_column = + TabletColumn(FieldAggregationMethod::OLAP_FIELD_AGGREGATION_NONE, null_type, + null_options.meta->is_nullable(), null_options.meta->unique_id(), + null_options.meta->length()); + null_column.set_name("nullable"); + null_column.set_index_length(-1); // no short key index + std::unique_ptr<Field> null_field(FieldFactory::create(null_column)); + null_writer = new ScalarColumnWriter(null_options, std::move(null_field), file_writer); + } + + *writer = std::unique_ptr<ColumnWriter>( + new StructColumnWriter(opts, std::unique_ptr<Field>(FieldFactory::create(*column)), + null_writer, sub_column_writers)); + return Status::OK(); +} + +Status ColumnWriter::create_array_writer(const ColumnWriterOptions& opts, + const TabletColumn* column, io::FileWriter* file_writer, + std::unique_ptr<ColumnWriter>* writer) { + DCHECK(column->get_subtype_count() == 1); + const TabletColumn& item_column = column->get_sub_column(0); + + // create item writer + ColumnWriterOptions item_options; + item_options.meta = opts.meta->mutable_children_columns(0); + item_options.need_zone_map = false; + item_options.need_bloom_filter = item_column.is_bf_column(); + item_options.need_bitmap_index = item_column.has_bitmap_index(); + if (item_column.type() == FieldType::OLAP_FIELD_TYPE_ARRAY) { + if (item_options.need_bloom_filter) { + return Status::NotSupported("Do not support bloom filter for array type"); + } + if (item_options.need_bitmap_index) { + return Status::NotSupported("Do not support bitmap index for array type"); + } + } + std::unique_ptr<ColumnWriter> item_writer; + RETURN_IF_ERROR(ColumnWriter::create(item_options, &item_column, file_writer, &item_writer)); + + // create length writer + FieldType length_type = FieldType::OLAP_FIELD_TYPE_UNSIGNED_BIGINT; + + ColumnWriterOptions length_options; + length_options.meta = opts.meta->add_children_columns(); + length_options.meta->set_column_id(2); + length_options.meta->set_unique_id(2); + length_options.meta->set_type(int(length_type)); + length_options.meta->set_is_nullable(false); + length_options.meta->set_length( + get_scalar_type_info<FieldType::OLAP_FIELD_TYPE_UNSIGNED_BIGINT>()->size()); + length_options.meta->set_encoding(DEFAULT_ENCODING); + length_options.meta->set_compression(opts.meta->compression()); + + length_options.need_zone_map = false; + length_options.need_bloom_filter = false; + length_options.need_bitmap_index = false; + + TabletColumn length_column = + TabletColumn(FieldAggregationMethod::OLAP_FIELD_AGGREGATION_NONE, length_type, + length_options.meta->is_nullable(), length_options.meta->unique_id(), + length_options.meta->length()); + length_column.set_name("length"); + length_column.set_index_length(-1); // no short key index + std::unique_ptr<Field> bigint_field(FieldFactory::create(length_column)); + auto* length_writer = + new OffsetColumnWriter(length_options, std::move(bigint_field), file_writer); + + // if nullable, create null writer + ScalarColumnWriter* null_writer = nullptr; + if (opts.meta->is_nullable()) { + FieldType null_type = FieldType::OLAP_FIELD_TYPE_TINYINT; + ColumnWriterOptions null_options; + null_options.meta = opts.meta->add_children_columns(); + null_options.meta->set_column_id(3); + null_options.meta->set_unique_id(3); + null_options.meta->set_type(int(null_type)); + null_options.meta->set_is_nullable(false); + null_options.meta->set_length( + get_scalar_type_info<FieldType::OLAP_FIELD_TYPE_TINYINT>()->size()); + null_options.meta->set_encoding(DEFAULT_ENCODING); + null_options.meta->set_compression(opts.meta->compression()); + + null_options.need_zone_map = false; + null_options.need_bloom_filter = false; + null_options.need_bitmap_index = false; + + TabletColumn null_column = + TabletColumn(FieldAggregationMethod::OLAP_FIELD_AGGREGATION_NONE, null_type, + length_options.meta->is_nullable(), null_options.meta->unique_id(), + null_options.meta->length()); + null_column.set_name("nullable"); + null_column.set_index_length(-1); // no short key index + std::unique_ptr<Field> null_field(FieldFactory::create(null_column)); + null_writer = new ScalarColumnWriter(null_options, std::move(null_field), file_writer); + } + + *writer = std::unique_ptr<ColumnWriter>( + new ArrayColumnWriter(opts, std::unique_ptr<Field>(FieldFactory::create(*column)), + length_writer, null_writer, std::move(item_writer))); + return Status::OK(); +} + +Status ColumnWriter::create_map_writer(const ColumnWriterOptions& opts, const TabletColumn* column, Review Comment: warning: function 'create_map_writer' exceeds recommended size/complexity thresholds [readability-function-size] ```cpp Status ColumnWriter::create_map_writer(const ColumnWriterOptions& opts, const TabletColumn* column, ^ ``` <details> <summary>Additional context</summary> **be/src/olap/rowset/segment_v2/column_writer.cpp:249:** 88 lines including whitespace and comments (threshold 80) ```cpp Status ColumnWriter::create_map_writer(const ColumnWriterOptions& opts, const TabletColumn* column, ^ ``` </details> ########## be/src/olap/rowset/segment_v2/column_writer.cpp: ########## @@ -88,267 +88,305 @@ class NullBitmapBuilder { RleEncoder<bool> _rle_encoder; }; +Status ColumnWriter::create_struct_writer(const ColumnWriterOptions& opts, + const TabletColumn* column, io::FileWriter* file_writer, + std::unique_ptr<ColumnWriter>* writer) { + // not support empty struct + DCHECK(column->get_subtype_count() >= 1); + std::vector<std::unique_ptr<ColumnWriter>> sub_column_writers; + sub_column_writers.reserve(column->get_subtype_count()); + for (uint32_t i = 0; i < column->get_subtype_count(); i++) { + const TabletColumn& sub_column = column->get_sub_column(i); + + // create sub writer + ColumnWriterOptions column_options; + column_options.meta = opts.meta->mutable_children_columns(i); + column_options.need_zone_map = false; + column_options.need_bloom_filter = sub_column.is_bf_column(); + column_options.need_bitmap_index = sub_column.has_bitmap_index(); + if (sub_column.type() == FieldType::OLAP_FIELD_TYPE_STRUCT) { + if (column_options.need_bloom_filter) { + return Status::NotSupported("Do not support bloom filter for struct type"); + } + if (column_options.need_bitmap_index) { + return Status::NotSupported("Do not support bitmap index for struct type"); + } + } + if (sub_column.type() == FieldType::OLAP_FIELD_TYPE_ARRAY) { + if (column_options.need_bloom_filter) { + return Status::NotSupported("Do not support bloom filter for array type"); + } + if (column_options.need_bitmap_index) { + return Status::NotSupported("Do not support bitmap index for array type"); + } + } + std::unique_ptr<ColumnWriter> sub_column_writer; + RETURN_IF_ERROR( + ColumnWriter::create(column_options, &sub_column, file_writer, &sub_column_writer)); + sub_column_writers.push_back(std::move(sub_column_writer)); + } + + // if nullable, create null writer + ScalarColumnWriter* null_writer = nullptr; + if (opts.meta->is_nullable()) { + FieldType null_type = FieldType::OLAP_FIELD_TYPE_TINYINT; + ColumnWriterOptions null_options; + null_options.meta = opts.meta->add_children_columns(); + null_options.meta->set_column_id(column->get_subtype_count() + 1); + null_options.meta->set_unique_id(column->get_subtype_count() + 1); + null_options.meta->set_type(int(null_type)); + null_options.meta->set_is_nullable(false); + null_options.meta->set_length( + get_scalar_type_info<FieldType::OLAP_FIELD_TYPE_TINYINT>()->size()); + null_options.meta->set_encoding(DEFAULT_ENCODING); + null_options.meta->set_compression(opts.meta->compression()); + + null_options.need_zone_map = false; + null_options.need_bloom_filter = false; + null_options.need_bitmap_index = false; + + TabletColumn null_column = + TabletColumn(FieldAggregationMethod::OLAP_FIELD_AGGREGATION_NONE, null_type, + null_options.meta->is_nullable(), null_options.meta->unique_id(), + null_options.meta->length()); + null_column.set_name("nullable"); + null_column.set_index_length(-1); // no short key index + std::unique_ptr<Field> null_field(FieldFactory::create(null_column)); + null_writer = new ScalarColumnWriter(null_options, std::move(null_field), file_writer); + } + + *writer = std::unique_ptr<ColumnWriter>( + new StructColumnWriter(opts, std::unique_ptr<Field>(FieldFactory::create(*column)), + null_writer, sub_column_writers)); + return Status::OK(); +} + +Status ColumnWriter::create_array_writer(const ColumnWriterOptions& opts, Review Comment: warning: function 'create_array_writer' exceeds recommended size/complexity thresholds [readability-function-size] ```cpp Status ColumnWriter::create_array_writer(const ColumnWriterOptions& opts, ^ ``` <details> <summary>Additional context</summary> **be/src/olap/rowset/segment_v2/column_writer.cpp:163:** 82 lines including whitespace and comments (threshold 80) ```cpp Status ColumnWriter::create_array_writer(const ColumnWriterOptions& opts, ^ ``` </details> -- 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: commits-unsubscr...@doris.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@doris.apache.org For additional commands, e-mail: commits-h...@doris.apache.org