feat(signals): improves aws sqs signal consumer plugin

[mvilanova]

1. Session Management Overhaul:

   • Original: Used a single long-lived session for all operations
   • New: Uses application's get_session() context manager for proper session lifecycle
   • Benefit: Better connection management, automatic cleanup, and session tracking

2. Transaction Isolation:

   • Original: Used begin_nested() without clear transaction boundaries
   • New: Explicit batch and message-level transaction management
   • Benefit: Better isolation between messages, clearer transaction scope

3. Graceful Shutdown:

   • Original: No shutdown handling
   • New: Added _shutdown flag and signal handlers for SIGTERM/SIGINT
   • Benefit: Graceful process termination, prevents data loss during shutdown

4. Error Handling:

   • Original: Basic error handling with potential for session leaks
   • New: Comprehensive error handling with proper cleanup
   • Added delay on errors to prevent tight loops
   • Benefit: More resilient, better resource cleanup

5. Resource Management:

   • Original: Potential for connection/session leaks
   • New: Proper session cleanup through context managers
   • Integration with application's SessionTracker
   • Benefit: Prevents resource leaks, better monitoring

6. Code Structure:

   • Original: Single large method handling everything
   • New: Split into smaller, focused methods
   • Added proper docstrings and type hints
   • Benefit: More maintainable, better documented

7. Message Processing:

   • Original: No clear separation between message processing and session management
   • New: Separated message processing logic into _process_message
   • Better handling of message-level transactions
   • Benefit: Clearer code organization, better error isolation

8. Batch Processing:

   • Original: No clear batch transaction boundaries
   • New: Explicit batch-level transaction with proper commit/rollback
   • Only deletes successfully processed messages
   • Benefit: More reliable message processing

The changes make the code more:

• Production-ready
• Maintainable
• Reliable
• Resource-efficient
• Integrated with application patterns
• Observable (through better logging and session tracking)

These improvements address potential issues with resource leaks, improve error handling, and make the code more robust for production use while maintaining the same core functionality.