[GitHub] [lucene] wjp719 commented on pull request #687: LUCENE-10425：speed up IndexSortSortedNumericDocValuesRangeQuery#BoundedDocSetIdIterator construction using bkd binary search

[GitBox <gi...@apache.org>]

wjp719 commented on PR #687:
URL: https://github.com/apache/lucene/pull/687#issuecomment-1254416217

   > Thanks, this looks good to me! Can you add a CHANGES entry with your name under 9.5?
   
   Thanks a lot, I have added the Change entry.
   
   And This PR has a limitation that only index sorting by ascend order can use bkd binary search to get min/max docId. The main reason is that bkd now sorts point  by two dimension (point value, docId) both in ascend order. If doc is index-sorted by ascend order, then all the docId of all leaf point will be monotone increasing, so we can use bkd binay search.
   
   In our local work, if doc is index-sorted by descend order, we modify the bkd sorting logic by (point value in ascend order , docId in descend order), so that all the docId of all leaf point will be monotone decreasing, then we can use bkd binay search again. 
   
   So May I open another PR to add an option that BKD can sort by (point value in ascend order , docId in descend order)? then the bkd binary search can work in both ascend/descend index sorting


-- 
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-unsubscribe@lucene.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org