[native_datafusion] Add support for reading row index metadata columns #3432
Description
What is the problem the feature request solves?
The native_datafusion scan currently falls back to Spark for queries that read row index metadata columns. It would be nice to support them.
Describe the potential solution
No response
Additional context
No response