alamb commented on issue #6692: URL: https://github.com/apache/arrow-rs/issues/6692#issuecomment-2464643766
> https://github.com/apache/arrow-rs/blob/e9bf8aa6bf67ec192fce1a6f3e7ab604c9689fef/arrow-select/src/filter.rs#L740 > > This line of code extends buffer (byte copied) regardless of the filtered result, I think it is the reason why we need gc. If we do `append_value` here, we have additional hash lookup and insert, but less buffer copied especially in low selectivity and no gc required later on. To be clear -- I think the copy is of a `Vec<Buffer>` (which is a Vec of pointers to the data) -- 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]
