Right now, there are a few places that we are not very consistent.
- There are a few functions that are registered in FunctionRegistry, but not provided in functions.scala and functions.py. Examples are isnull and get_json_object.
- There are a few functions that we have different names in FunctionRegistry and method name in DataFrame API. spark_partition_id is an example. In FunctionRegistry, it is called spark_partition_id. But, DataFrame API, the method is called sparkPartitionId.