Everything inside of this directory are copied from Apache Spark 3.5.6 with minor modifications.
This is for avoid depending on rapidly changing internal Spark APIs related to ParquetFileFormat.
GeoParquetFileFormat used to depend on the ParquetFileFormat class of Spark and it has caused lots
of problems on Databricks, so we have to copy the code here to avoid the dependency.

See related issues:

- https://github.com/apache/sedona/issues/2250
- https://github.com/apache/sedona/issues/1905
- https://github.com/apache/sedona/issues/1869
- https://github.com/apache/sedona/issues/996
