rich7420 commented on PR #7156: URL: https://github.com/apache/datafusion-comet/pull/7156#issuecomment-6760199627
@andygrove thanks for the review! Addressed in e1afd07: - `promotable` now includes `timestamp` and `timestamp_ns`. The incomplete-history test adds a `timestamp` column and a `timestamptz` column. Only the `timestamp` column is treated as promoted. In Iceberg 1.11, `TypeUtil.isPromotionAllowed` doesn't yet allow promotion from date to timestamp, so Spark can't run that `ALTER`. Only the unit test covers this case. - The guide now lists `BIGINT`, `DOUBLE`, `DECIMAL` and `TIMESTAMP_NTZ`. - #5883 only covers partial conjuncts. I opened #7588 to track restoring float range pruning. The comment in `CometIcebergNativeScan.scala` and the guide link to it. The description now says these filters rely on the filter above the scan. - The promoted-column test now drops `id` and adds it back as a new `INT` column, then reads the last snapshot with `VERSION AS OF`. A plain drop isn't enough, because Iceberg can't plan a filter on a dropped column. Only the scan schema still shows that the old `id` was promoted. With the table schema alone, `id > 0` returns 2 of the 6 rows. -- 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]
