Gabriel39 commented on PR #66556: URL: https://github.com/apache/doris/pull/66556#issuecomment-5306903941
Follow-up on the previous P1: commit `4331bcc` does not fix the affected legacy reader path. The new commit changes only `be/src/format_v2/parquet/reader/native/column_chunk_reader.cpp` and adds `ParquetV2NativeDecoderTest.CompressedRequiredBoolWithoutValuesReturnsCorruption`. However, the crash reported by #66430 and the original fix in this PR use the legacy reader under `be/src/format/parquet`. Its `vparquet_column_chunk_reader.cpp` still contains: ```cpp _empty_value_section = _page_data.empty() && _max_def_level > 0; ``` Consequently, a required BOOLEAN/PLAIN page with a nonzero logical value count and an empty physical value section still bypasses `EmptyValueSectionDecoder` in the affected reader and can reach `BoolPlainDecoder::set_data()` with `Slice(nullptr, 0)`. The new native-reader test does not exercise or protect this path. Please apply the corresponding condition change to `be/src/format/parquet/vparquet_column_chunk_reader.cpp` and add the required-column regression case to `ParquetColumnChunkReaderTest`. -- 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]
