Skip to content

support iceberg metadata columns _file, _pos, _spec, and _partition #5054

Description

@parthchandra

What is the problem the feature request solves?

Iceberg's metadata columns are currently not supported by Comet. A common use case for Iceberg (merge-on-read, copy-on-write) requires reading the metadata columns _file, _pos, _spec, and _partition. Iceberg scans automatically introduce these columns if required. Because these are not supported, these jobs simply fall back to Spark.

Describe the potential solution

Add support for these columns.

Additional context

_file, _pos, and, _spec are already supported by iceberg-rust. _partition is being reviewed.

Metadata

Metadata

Assignees

Type

No type

Projects

No projects

Milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions