Add structured logging across the codebase

[ajshedivy]

Summary

The codebase has inconsistent and minimal logging. The synchronous SQLJob path has zero logging — it uses print() on errors (e.g., sql_job.py:122). The async PoolJob has only 8 logger.debug() calls scoped to a misleading websockets.client logger name. No other modules (Connection, Cursor, Query, Pool management) emit any log messages.

This makes production debugging, performance analysis, and operational monitoring extremely difficult.

Current State

• Sync code (SQLJob, Query, Connection, Cursor): No logging at all
• SQLJob.connect(): Uses print(f"an error occured while loading connect result: {e}") — a bare print() in production code
• Async code (PoolJob): 8 logger.debug() calls using getLogger("websockets.client") — a misleading logger name
• Pool scaling decisions: Not logged (when connections are added/removed)
• Query lifecycle: Not logged (prepare, execute, fetch, close)
• Error paths: Errors are raised but not logged before propagation

Proposed Changes

1. Add a consistent logging hierarchy using Python's logging module:

   • mapepire_python (root)
   • mapepire_python.client (SQLJob, Query)
   • mapepire_python.pool (Pool, PoolJob, PoolQuery)
   • mapepire_python.core (Connection, Cursor)
   • mapepire_python.auth (Kerberos, authentication)

2. Replace all print() calls with proper logger.error() / logger.warning()

3. Add logging at key lifecycle points:

   • INFO: Connection opened/closed, pool scaled up/down
   • DEBUG: Query prepared/executed/fetched/closed, message sent/received
   • WARNING: Connection dropped, retry attempted, pool exhaustion approaching
   • ERROR: Authentication failure, query failure, connection failure

4. Ensure no sensitive data is logged (passwords, tokens, etc.)

Files to Modify

• mapepire_python/client/sql_job.py — Replace print(), add connection lifecycle logging
• mapepire_python/client/query.py — Add query lifecycle logging
• mapepire_python/pool/pool_client.py — Add pool scaling/management logging
• mapepire_python/pool/pool_job.py — Fix logger name, add consistent logging
• mapepire_python/pool/pool_query.py — Add async query logging
• mapepire_python/core/connection.py — Add connection open/close logging
• mapepire_python/core/cursor.py — Add cursor operation logging
• mapepire_python/authentication/kerberosTokenProvider.py — Add auth logging (no secrets)

Acceptance Criteria

• All print() calls in production code are replaced with proper logging
• Every module uses logging.getLogger(__name__) for its logger
• Connection open/close events logged at INFO level
• Query execution logged at DEBUG level
• Pool scaling events logged at INFO level
• Error paths logged at ERROR level before exception propagation
• No passwords, tokens, or credentials appear in log output
• Existing tests still pass