gf2121 commented on PR #12006: URL: https://github.com/apache/lucene/pull/12006#issuecomment-1345297296
Thanks @rmuir for feedback! This is the current implementation of `ArrayUtil#compareUnsigned4`: ``` public static int compareUnsigned4(byte[] a, int aOffset, byte[] b, int bOffset) { return Integer.compareUnsigned( (int) BitUtil.VH_BE_INT.get(a, aOffset), (int) BitUtil.VH_BE_INT.get(b, bOffset)); } ``` I think it's equivalent to ``` public static int compareUnsigned4(byte[] a, int aOffset, byte[] b, int bOffset) { return Integer.compare(NumericUtils.sortableBytesToInt(a, aOffset), NumericUtils.sortableBytesToInt(b, bOffset)); } ``` Because they are both doing a varhandle get and compare unsigned. The point of this PR is to avoid the decoding of one side as the query values are consistent during the intersect. I'll do some JMH benchmark to confirm if there is any speed up. -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org