github-actions[bot] commented on code in PR #63192:
URL: https://github.com/apache/doris/pull/63192#discussion_r3253811118
##########
be/src/format/parquet/vparquet_reader.cpp:
##########
@@ -1339,8 +1469,7 @@ Status ParquetReader::_process_column_stat_filter(
}
const auto& file_col_name =
_table_info_node_ptr->children_file_column_name(slot->col_name());
- const FieldSchema* col_schema =
- _file_metadata->schema().get_column(file_col_name);
+ const FieldSchema* col_schema =
parquet_file_schema().get_column(file_col_name);
int parquet_col_id = col_schema->physical_column_index;
auto meta_data =
row_group.columns[parquet_col_id].meta_data;
stat->col_schema = col_schema;
Review Comment:
This row-group stats path now also sees top-level VARIANT predicates, but a
VARIANT root is a group field with `physical_column_index == -1`. A predicate
such as `v IS NULL` / `v IS NOT NULL` can enter
`_process_column_stat_filter()`, map the slot to the top-level VARIANT schema,
and index `row_group.columns[-1]` here before any guard. The page-index path
below already returns false for `parquet_col_id < 0`; please add the same guard
before reading min/max metadata and in the bloom-filter lambda as well, with
coverage for a pushed predicate on a Parquet VARIANT column.
--
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]
