[SPARK-53776][CONNECT] Support Spark Context Manager Support in Connect Mode

[shujingyang-db]

What changes were proposed in this pull request?

This PR adds missing context manager support (__enter__ and __exit__ methods) to the Spark Connect SparkSession class to enable with statement usage for automatic resource management.

The regular PySpark SparkSession already supports context manager syntax like:

with SparkSession.builder.master("local").getOrCreate() as session:
    session.range(5).show()

However, this functionality was missing from the Spark Connect implementation, causing inconsistency between the two APIs.

Why are the changes needed?

1. API Consistency: The Spark Connect SparkSession should have feature parity with the regular PySpark SparkSession
2. Resource Management: Context manager support enables automatic cleanup of sessions, preventing resource leaks
3. User Experience: Provides a more Pythonic way to handle session lifecycle management
4. Documentation Compliance: The existing docstrings in regular SparkSession show examples using context manager syntax, but this didn't work in Spark Connect mode

Does this PR introduce any user-facing change?

Yes, this PR enables new functionality for Spark Connect users:

Before this PR:

# This would fail in Spark Connect mode
with SparkSession.builder.remote("sc://localhost").getOrCreate() as session:
    session.range(5).show()
# AttributeError: __enter__ method missing

After this PR:

# This now works in Spark Connect mode
with SparkSession.builder.remote("sc://localhost").getOrCreate() as session:
    session.range(5).show()
# Session is automatically stopped when exiting the with block

How was this patch tested?

UTs

Was this patch authored or co-authored using generative AI tooling?

Yes