sameerpandey17 commented on code in PR #75635: URL: https://github.com/apache/airflow/pull/75635#discussion_r4399472267
########## airflow-core/src/airflow/api_fastapi/common/parameters/search.py: ########## Review Comment: Thanks @yuseok89, this is very useful data. Keeping the original bounded range is fast, but it is wrong under some collations (contractions like Czech `ch`). Removing only the upper bound, as this PR does, fixes those but gives the slowdowns you measured on skewed data (39 ms → 10.6 s). I also found that the lower bound alone can drop valid matches under a few collations [Welsh, Igbo], so keeping it isn't fully safe either. Neither the original code nor the current PR is right on both correctness and speed. I'll look at an approach that is correct across collations and keeps range-scan speed, possibly the index-level change you suggested as a separate PR, and share findings here before changing the PR. -- 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]
