Extended I/O Framework: Readers/Writers for Parquet

[sayedkeika]

Overview

This pull request adds support for the Parquet file format. The implementation includes both readers and writers that can handle Parquet files sequentially and concurrently.

Details

• Single-threaded reader and writer for smaller datasets
• Multi-threaded reader and writer for larger datasets
• Component tests for both sequential and parallel operations that cover different schemas
• Detailed documentation

[mboehm7]

LGTM - thanks for the patch @sayedkeika. During the merge I resolved the merge conflicts, fixed the warnings and formatting (tabs over spaces), added additional tests (sparse data), and left a FIXME (for removing the ExampleParquetWriter). Additionally, I fixed the parallel write task to call the sequential write instead of the parallel writer again (which worked because the number of part files was 1 due to the small size).