peterxcli commented on issue #3983: URL: https://github.com/apache/datafusion-comet/issues/3983#issuecomment-5874196434
Status update (2026-09-29): #5868 merged on September 25 and completed #5546. Comet now reconstructs whole Variant values from shredded top-level columns in ordinary native Parquet scans with `spark.sql.variant.allowReadingShredded=true` and `spark.sql.variant.pushVariantIntoScan=false`. This issue should remain open for shredded writing, selective subfield reads, and predicate pushdown. #5433 tracks the whole-value writer foundation; #5424/#5426 track extraction semantics. #5519 covers Spark's pushed one-field full-value request, and #5435 covers Variant nested inside ordinary structs, arrays, and maps. Current main pins DataFusion 55.1.0 and Arrow/Parquet 59.3.0 and already uses Arrow Variant compute. #5477 now records the current upstream dependencies; #5978 is the next whole-value performance priority. The refreshed roadmap is #5438. -- 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]
