comphead opened a new pull request, #26688: URL: https://github.com/apache/datafusion/pull/26688
## Which issue does this PR close? - Closes #26687. ## Rationale for this change Since 55.0.0, reading a spill file validates every record batch again, including a UTF-8 check of every string. #15454 had turned that off, but `StreamDecoder::with_skip_validation`, which #21882 switched to, only applies to dictionary batches. ## What changes are included in this PR? `SpillReaderStream` already frames whole IPC messages. It now decodes them with `RecordBatchDecoder` and `read_dictionary_impl`, which take the `skip_validation` flag, instead of `StreamDecoder`. `spill_io` benchmark, criterion medians averaged over two alternating runs: | benchmark | main | this PR | change | |---|---:|---:|---:| | `spill_io/StreamReader/read_100` | 9.39 ms | 8.25 ms | -12% | | `spill_compression/q16/uncompressed` | 8.54 ms | 7.94 ms | -7% | | `spill_compression/wide/uncompressed` | 23.17 ms | 21.80 ms | -6% | | `spill_compression/q16/lz4_frame` | 22.65 ms | 21.54 ms | -5% | | `spill_compression/q16/zstd` | 40.31 ms | 39.46 ms | -2% | `q2` and `q20` have no strings and stay within noise. ## What is the testing strategy for this PR? `test_spill_read_skips_validation` spills a `Utf8` value that is not UTF-8. It fails on main with `Invalid UTF8 sequence at string index 0` and passes here. The existing spill tests cover decoding, including dictionaries and truncated files. ## Are there any user-facing changes? No. -- 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]
