We'll implement basic full table scan first so that we can start to integrate with query engine such as datafusion, which would be quite useful for data driven tests.
Advanced features such as partition pruning, data file pruning, sort order preserving will be implemented later.