gf2121 commented on PR #14333:
URL: https://github.com/apache/lucene/pull/14333#issuecomment-2709371309

   Thanks for feedback!
   
   > Since you wrote this, I expected tip files to become bigger, but your data 
suggests the opposite, tip files are getting smaller? Am I reading it correctly?
   
   Sorry I didn't make that clear. This was my motivation for implementing a 
specialized trie instead of a specialized FST. From a data structure 
perspective, an FST should be smaller than a trie because it can share output 
prefixes (as evidenced by https://github.com/apache/lucene/pull/12631). But in 
practice, this trie is specialized for block tree index, so the tip is actually 
smaller because this trie does more bitwise encoding than vint/vlong and it 
doesn't need to store the length of the floor data since it's already 
self-contained to know the length during reading.


-- 
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

Reply via email to