github-actions[bot] commented on code in PR #63199:
URL: https://github.com/apache/doris/pull/63199#discussion_r3231376266
##########
be/src/format/parquet/vparquet_group_reader.cpp:
##########
@@ -1189,11 +1189,24 @@ Status RowGroupReader::_rewrite_dict_predicates() {
// 4. Rewrite conjuncts.
RETURN_IF_ERROR(_rewrite_dict_conjuncts(
dict_codes, slot_id,
temp_block.get_by_position(dict_pos).column->is_nullable()));
+ _remove_slot_filter_conjuncts(slot_id);
Review Comment:
This fixes a BE crash path, but the PR only records manual testing and does
not add an automated regression. Please add coverage for the exact scenario:
Parquet dictionary filtering on a string `IN` predicate with lazy
materialization enabled and at least one lazy-read column, so the predicate
phase reads dictionary ids and would previously execute the original string
predicate against `ColumnInt32`. There is already related coverage in
`regression-test/suites/external_table_p0/hive/test_string_dict_filter.groovy`
and `test_parquet_lazy_mat_profile.groovy`; extending one of those would
prevent this crash from regressing.
--
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]
