Gabriel39 commented on code in PR #65674:
URL: https://github.com/apache/doris/pull/65674#discussion_r3609373498
##########
be/src/format_v2/parquet/parquet_reader.cpp:
##########
@@ -119,22 +119,23 @@ void collect_request_leaf_column_ids(
}
std::vector<ParquetPageCacheRange> build_page_cache_ranges(
- const ::parquet::FileMetaData& metadata,
+ const tparquet::FileMetaData& metadata,
const std::vector<std::unique_ptr<ParquetColumnSchema>>& file_schema,
const format::FileScanRequest& request, const RowGroupScanPlan&
row_group_plan) {
std::unordered_set<int> leaf_column_ids;
collect_request_leaf_column_ids(file_schema, request, &leaf_column_ids);
std::vector<ParquetPageCacheRange> ranges;
ranges.reserve(row_group_plan.row_groups.size() * leaf_column_ids.size());
for (const auto& row_group_plan_item : row_group_plan.row_groups) {
- auto row_group_metadata =
metadata.RowGroup(row_group_plan_item.row_group_id);
- DORIS_CHECK(row_group_metadata != nullptr);
+ const auto& row_group_metadata =
metadata.row_groups[row_group_plan_item.row_group_id];
for (const auto leaf_column_id : leaf_column_ids) {
- DORIS_CHECK(leaf_column_id >= 0 && leaf_column_id <
row_group_metadata->num_columns());
- auto column_metadata =
row_group_metadata->ColumnChunk(leaf_column_id);
- DORIS_CHECK(column_metadata != nullptr);
- const int64_t offset = column_chunk_start_offset(*column_metadata);
- const int64_t size = column_metadata->total_compressed_size();
+ DORIS_CHECK(leaf_column_id >= 0 &&
+ leaf_column_id <
static_cast<int>(row_group_metadata.columns.size()));
+ const auto& column_metadata =
row_group_metadata.columns[leaf_column_id].meta_data;
+ const int64_t offset =
column_metadata.__isset.dictionary_page_offset
Review Comment:
Removed the obsolete open-time page-cache range traversal and its no-op
registration hook. Native readers now remain the sole owner of validated
exact-page cache ranges.
--
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]