Override default implementation of skip(long) method

[findinpath]

Description

Delegate to stream the implementation of skip(long) method. By doing this, it is ensured that not the default implementation from java.io.InputStream is being used, because it calls read() causing actual reads from the underlying input stream although skip is just a logical operation.

Additional context and related issues

Fixes #18976

Release notes

( ) This is not user-visible or is docs only, and no release notes are required.
( ) Release notes are required. Please propose a release note for me.
(x) Release notes are required, with the following suggested text:

# Hive
* Avoid reading unrelated data to the text line reader splits. ({issue}`issuenumber`)

[electrum]

Where are we using skip() rather than seek() for large sizes? The new S3 implementation (and possibly Azure) will also need to implement this efficiently.

[electrum]

Changes requested for the S3InputStream change.

To answer my previous question: I found that TextLineReader and SequenceFileReader use skipNBytes(), which we implement for S3InputStream and AzureInputStream. For HdfsTrinoInputStream, it uses the default implementation, which calls skip(), hence why we need to make the change here to delegate correctly.

[findinpath]

I found that TextLineReader and SequenceFileReader use skipNBytes()

See here the hierarchy of calls

[findinpath]

Removed the changes in S3InputStream because they were unrelated to this change.
Thank you @electrum for raising this aspect.

[findepi]

/test-with-secrets sha=3af6109e7dc6a9f286f84c6c84b66bcf342c4967

[github-actions]

The CI workflow run with tests that require additional secrets has been started: https://github.com/trinodb/trino/actions/runs/6145626334