Remove legacy Hive readers and writers

[electrum]

Description

This PR will be merged once we have confidence in the new readers and writers.

Release notes

(x) Release notes are required, with the following suggested text:

# Hive
* Remove legacy Hive readers. The `*_native_reader_enabled` session properties and the
  `*.native-reader.enabled` configuration properties have been removed. ({issue}`18241`)
* Remove legacy Hive writers. The `*_native_writer_enabled` session properties and the
  `*.native-writer.enabled` configuration properties have been removed. ({issue}`18241`)
* Remove support for S3 Select. The `s3_select_pushdown_enabled` session property
  and the `hive.s3select*` configuration properties have been removed. ({issue}`18241`)
* Optimized symlink listing is always enabled. The `optimize_symlink_listing` session property
  and `hive.optimize-symlink-listing` configuration property have been removed. ({issue}`18241`)

[awakeljw]

Ihave two questions， Will the s3 select pushdown feature come back ？ Will s3 select pushdown feature support parquet file pushdown?

[electrum]

@awakeljw We have no plans to reimplement S3 Select. For Parquet, the results are returned as CSV or JSON, which is why it was never implemented in Trino. The feature introduces significant complexity in Trino and only benefits textfile formats. Additionally, the pricing model for S3 Select seems to put it at a disadvantage compared to simply increasing compute on the Trino side. Thus, we don't believe it is worth the complexity and maintenance burden.