AndreKurait opened a new pull request, #16746: URL: https://github.com/apache/lucene/pull/16746
### Description This change adds a performance improvement for some repetitive inputs, aligning with optimizations in the C LZ4 implementation, specifically [`LZ4_memcpy_using_offset()`](https://github.com/lz4/lz4/blob/0774d05537f9762f838f7ab541b7765f1a729cb5/lib/lz4.c#L540-L581). > AI-assisted context: C's offset-2 case expands a pattern from 2 to 4 to 8 bytes. Java generalizes pattern expansion with `Arrays.fill` and progressively larger, disjoint `System.arraycopy` calls. The loops differ; the encoder and stored format are unchanged. > > Updated-main local result: 4.55× repetitive-record decoding; ordinary-input regression bounds remain inconclusive. Full `tidy check` passed (18,310 tests, 811 skipped). [Benchmark evidence, tradeoffs, and standalone reproducer](https://github.com/AndreKurait/lucene/pull/1). -- 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]
