xx

eldenmoon · Aug 6, 2024 · 5d2619b · 5d2619b
1 parent 9bd6883
commit 5d2619b
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 44 deletions.
diff --git a/be/src/olap/rowset/segment_v2/hierarchical_data_reader.h b/be/src/olap/rowset/segment_v2/hierarchical_data_reader.h
@@ -122,7 +122,7 @@ class HierarchicalDataReader : public ColumnIterator {
             auto type = root_var.get_root_type();
             container_variant.add_sub_column({}, std::move(column), type);
         }
-        // parent -> subcolumns
+        // parent path -> subcolumns
         std::map<PathInData, PathsWithColumnAndType> nested_subcolumns;
         PathsWithColumnAndType non_nested_subcolumns;
         RETURN_IF_ERROR(tranverse([&](SubstreamReaderTree::Node& node) {
@@ -182,44 +182,6 @@ class HierarchicalDataReader : public ColumnIterator {
             container_variant.add_sub_column(entry.first, array->assume_mutable(),
                                              ColumnObject::NESTED_TYPE);
         }
-        // if (has_nested) {
-        //     // rewrite nested nodes
-        //     container_variant.finalize_if_not();
-        //     MutableColumnPtr nested_object = ColumnObject::create(true, false);
-        //     MutableColumnPtr offset =
-        //             check_and_get_column<ColumnArray>(
-        //                     *remove_nullable(container_variant.get_subcolumns()
-        //                                              .get_leaves()[0]
-        //                                              ->data.get_finalized_column_ptr()))
-        //                     ->get_offsets_ptr()
-        //                     ->assume_mutable();
-        //     auto* nested_object_ptr = assert_cast<ColumnObject*>(nested_object.get());
-        //     // flatten nested arrays
-        //     for (const auto& entry : container_variant.get_subcolumns()) {
-        //         auto& column = entry->data.get_finalized_column_ptr();
-        //         const auto& type = entry->data.get_least_common_type();
-        //         if (!remove_nullable(column)->is_column_array()) {
-        //             return Status::InvalidArgument(
-        //                     "Meet none array column when flatten nested array, path {}, type {}",
-        //                     entry->path.get_path(), entry->data.get_finalized_column().get_name());
-        //         }
-        //         MutableColumnPtr flattend_column =
-        //                 check_and_get_column<ColumnArray>(
-        //                         remove_nullable(entry->data.get_finalized_column_ptr()).get())
-        //                         ->get_data_ptr()
-        //                         ->assume_mutable();
-        //         DataTypePtr flattend_type =
-        //                 check_and_get_data_type<DataTypeArray>(remove_nullable(type).get())
-        //                         ->get_nested_type();
-        //         nested_object_ptr->add_sub_column(entry->path, std::move(flattend_column),
-        //                                           std::move(flattend_type));
-        //     }
-        //     nested_object = make_nullable(nested_object->get_ptr())->assume_mutable();
-        //     auto array =
-        //             make_nullable(ColumnArray::create(std::move(nested_object), std::move(offset)));
-        //     container_variant.clear();
-        //     container_variant.create_root(ColumnObject::NESTED_TYPE, array->assume_mutable());
-        // }
 
         // TODO select v:b -> v.b / v.b.c but v.d maybe in v
         // copy container variant to dst variant, todo avoid copy

diff --git a/be/src/olap/rowset/segment_v2/segment_iterator.cpp b/be/src/olap/rowset/segment_v2/segment_iterator.cpp
@@ -2422,17 +2422,22 @@ Status SegmentIterator::_fill_missing_columns(vectorized::Block* block) {
         auto cid = _schema->column_id(i);
         const auto* column_desc = _schema->column(cid);
         int64_t current_size = block->get_by_position(i).column->size();
+        size_t max_rows = 0;
+        for (const auto& entry : *block) {
+            max_rows = std::max(max_rows, entry.column->size());
+        }
         if (column_desc->path() != nullptr && column_desc->path()->has_nested_part() &&
-            current_size < block->rows()) {
+            current_size < max_rows) {
+            // Find leaf node of the same nested parent which column size is greate than current_size
             const auto* leaf = offsets.get_leaf_of_the_same_nested(
                     *column_desc->path(), [](const auto& node) { return node.data->size(); },
                     current_size);
             if (!leaf) {
                 VLOG_DEBUG << "Not found any subcolumns column_desc: "
                            << column_desc->path()->get_path() << ", current_size: " << current_size
-                           << ", block_rows: " << block->rows();
+                           << ", block_rows: " << max_rows;
                 block->get_by_position(i).column->assume_mutable()->insert_many_defaults(
-                        block->rows() - current_size);
+                        max_rows - current_size);
                 continue;
             }
             LOG(INFO) << "fuck";
@@ -2444,8 +2449,8 @@ Status SegmentIterator::_fill_missing_columns(vectorized::Block* block) {
                             ->get_nested_type()
                             ->create_column_const_with_default_value(nested_padding_size);
             auto nested_new_offset = vectorized::ColumnArray::ColumnOffsets::create();
-            nested_new_offset->reserve(block->rows() - current_size);
-            for (size_t i = current_size; i < block->rows(); ++i) {
+            nested_new_offset->reserve(max_rows - current_size);
+            for (size_t i = current_size; i < max_rows; ++i) {
                 nested_new_offset->get_data().push_back_without_reserve(
                         (*offset)[i] - (*offset)[current_size - 1]);
             }