acking-you commented on PR #11247: URL: https://github.com/apache/datafusion/pull/11247#issuecomment-2208533054
> > Thank you for this contribution @acking-you -- I am also running some benchmarks on this branch and will report back. This PR looks quite close to me > > I also think it would be good to move this optimization upstream into the arrow-rs binary kernels as well > > https://docs.rs/arrow/latest/arrow/compute/kernels/boolean/index.html > > I will file a ticket to do so later if no one beats me to it > > As @jayzhan211 pointed out [#11212 (comment)](https://github.com/apache/datafusion/issues/11212#issuecomment-2205933735) the optimization here is a bit more "advanced" than adding it to a kernel, as we would avoid evaluating the right side altogether (e.g. `ALL_FALSE AND (regexp_like(...))` stops after executing ALL_FALSE whereas changing the kernel still would execute the `regexp_like`) That's exactly what I was getting at. Unless datafusion significantly overhauls the calculation logic here, it's unlikely a solution could be achieved by modifying the kernel's code. -- 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: github-unsubscr...@datafusion.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org For additional commands, e-mail: github-h...@datafusion.apache.org
