Github user rdblue commented on the issue: https://github.com/apache/spark/pull/21306 @felixcheung, we're waiting on more reviews and a community decision about how to pass partition transforms. For passing transforms, I think the most reasonable compromise is to go with a generic function application, so each transform would be passed as a function/transform name with one or more arguments, where each argument is either a column reference (by name) or a literal value. That's a fairly small public API addition but it supports a lot of different partitioning schemes to be expressed, including the one above for Kudu. We already have all of this implemented based on the current PR, but I can update this in the next week or so.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org