CASSANDRA-20154: Fix BETWEEN bounds to preserve inverted bounds

[arvindksi274-ksolves]

Fix CASSANDRA-20154: BETWEEN operator returns wrong results when start bound is greater than end bound.

This PR fixes a bug where queries like SELECT * FROM table WHERE column BETWEEN 5 AND 3
incorrectly return rows instead of an empty result set. The issue was caused by Cassandra
automatically sorting BETWEEN bounds, which violates SQL standards.

Changes made:

• Modified SimpleRestriction.java to preserve user-specified bound order for BETWEEN queries
• Updated ClusteringColumnRestrictions.java to properly handle BETWEEN on first clustering column
• Added tests to verify inverted BETWEEN bounds now return empty results

Before: WHERE pk BETWEEN 5 AND 3 would return rows between 3 and 5
After: WHERE pk BETWEEN 5 AND 3 correctly returns no rows (as per SQL standard)

By Arvind Kandpal for CASSANDRA-20154

[arvindksi274-ksolves]

Hi team!

This PR has been open for 4 days and I'd appreciate some feedback on my fix for CASSANDRA-20154.

Quick Summary:

• Fixes BETWEEN operator to handle inverted bounds correctly per SQL standards
• WHERE pk BETWEEN 5 AND 3 now returns empty results (was incorrectly returning rows)
• Changes in SimpleRestriction.java and ClusteringColumnRestrictions.java
• Added comprehensive tests to verify the fix

Technical Details:
The core issue was Cassandra automatically sorting BETWEEN bounds, causing inverted bounds like BETWEEN 5 AND 3 to return rows instead of empty results. This PR preserves user-specified bound order to match SQL standard behavior.

Testing:

• Added tests for inverted BETWEEN bounds
• Verified existing functionality remains intact
• Covers both partition key and clustering column scenarios

Would really appreciate review from the community! Particularly looking for feedback on:

1. Any edge cases I might have missed
2. Performance implications
3. Additional test scenarios to consider

@dineshjoshi @maedhroz @smiklosovic @jmckenzie-dev @ifesdjeen - would any of you have time to take a look when convenient?

Thanks for your time!