Skip the storage of repetition level and definition level for all-null column

I have a very sparse protobuf message in my project, with thousands of fields.
In practise, most of the fields are all null values in one page.
But the repetition level and definition level takes lots of storage space.
Can parquet skip the storage of r level and d level for such all-null columns to save storage space?

**Reporter**: [Li](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=lirui-tx)

_{**Note**: *This issue was originally created as [PARQUET-792](https://issues.apache.org/jira/browse/PARQUET-792). Please see the [migration documentation](https://issues.apache.org/jira/browse/PARQUET-2502) for further details.*}

No file, test, or entry point is named in the issue. Begin by locating the Java code that encodes repetition and definition levels, then review existing tests for all-null columns. Done means an agreed implementation and regression coverage for avoiding level storage in the described case.

Written by the indexing model from the issue text.