gaborkaszab commented on PR #16560: URL: https://github.com/apache/iceberg/pull/16560#issuecomment-4543826520
Thanks for pinging me on this, @ebyhr ! I wasn't aware of such a limitation from the Trino side. I checked the functions that you changed to public and IMO they are very low level functions meant for internal usage. I'm just throwing out ideas here, but would it make sense to split the current `computeAndWriteStatsFile` public function into multiple parts and make the part public that is independent from the Parquet read/write. In practice the public interface to stats does these steps: 1) read existing stats (Parquet reading involved) 2) calculate new stats either full or incremental (independent of file read/write, except Manifest reading) 3) write new stat file (Parquet writing involved) This is an oversimplified summary here, I know, but might worth exploring if we can split 2) into a separate function and make it public. Would. that be sufficient, @ebyhr ? -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
