baibaichen commented on issue #37559: URL: https://github.com/apache/arrow/issues/37559#issuecomment-1711020341
> @wgtmac Another problem is that, for RowSelector, we need to implement a "lazy" logic. > > i.e. use different filter in `AND(Filter1, Filter2, ..)` to produce sub-rowrange. The interface might be around `parquet::ColumnReader` with a selector? What does **lazy mean**? Does it mean [late materilization](https://docs.google.com/document/d/1QFu_Zu9nHuMpu5Pqb3qe62MbZPA88j_o7NtpZ2a2zSA/edit)?? If so, that would be the required feature. From experience of implementing filter push down in clickhouse (not in arrow), we have found decoding unnecessary filter columns hurt performance for tpch q6. -- 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]
