[SPARK-58021][CONNECT] Add local server pool storage model

[ericm-db]

What changes were proposed in this pull request?

This is layer 2 of the local Connect pool stack, following #57684.

It adds the filesystem-backed storage model for pool members:

• stable member identities and state-file paths;
• a per-pool cross-process file lock;
• secure JSON state reads, writes, renames, and cleanup;
• fingerprints for master, startup conf, working directory, and Python environment; and
• FIFO claiming of reachable, version-compatible members.

Process lifecycle, acquisition, SparkSession integration, and JIT warmup remain in later PRs.

Why are the changes needed?

The pool needs a small, independently reviewable state model before adding process supervision and background launching. Keeping this layer limited to storage and claiming makes the concurrency contract and on-disk state transitions reviewable without the attendant implementation.

Does this PR introduce any user-facing change?

No. The storage model is internal and is not wired into SparkSession in this layer.

How was this patch tested?

Added seven focused tests in pyspark.sql.tests.connect.test_connect_local_server_pool.LocalConnectServerPoolUnitTests covering directory selection, pool-size parsing, zombie detection, fingerprint identity, FIFO claiming, fingerprint matching, and unreachable members.

python -m unittest -v pyspark.sql.tests.connect.test_connect_local_server_pool.LocalConnectServerPoolUnitTests

All 7 tests passed. Ruff check, Ruff format check, and git diff --check also passed.

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

Generated-by: Claude Code (Fable 5) and OpenAI Codex (GPT-5)