jaideeppyne commented on issue #11121: URL: https://github.com/apache/arrow-rs/issues/11121#issuecomment-6646854185
@alamb, here is a real [malformed Parquet file (487 bytes)](https://raw.githubusercontent.com/jaideeppyne/arrow-rs/3f70bb88cb0c1029e84fcd297fe6a742dc13500b/malformed-bss.parquet), alongside its [valid control](https://raw.githubusercontent.com/jaideeppyne/arrow-rs/3f70bb88cb0c1029e84fcd297fe6a742dc13500b/valid-bss.parquet) and a [public-API reproducer with run instructions](https://github.com/jaideeppyne/arrow-rs/blob/3f70bb88cb0c1029e84fcd297fe6a742dc13500b/README.md). The writer creates eight doubles. Changing only zero-based byte 14 from 16 to 126 makes the page declare 63 values without changing its payload or footer. [Hosted verification](https://github.com/jaideeppyne/arrow-rs/actions/runs/41945965606) confirmed that the control reads eight rows and the malformed file panics on unchanged upstream at the indexing operation in `join_streams_const`: `index out of bounds: the len is 56 but the index is 56`. The identical malformed bytes return an ordinary insufficient-values error with #11136 at b280e76. Both files are synthetic, and the new regression constructs the file through the writer and reads it through `ParquetRecordBatchReaderBuilder`. Prepared and verified with OpenAI Codex assistance. -- 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]
