Simplify shutdown handling for graceful cleanup

[jssmith]

Summary

Simplify the shutdown handling by replacing complex signal-based shutdown with a cleaner try/finally approach.

Problem

The previous implementation used custom signal handlers that called sys.exit() from async code, which could cause:

• "asyncio.run() cannot be called from a running event loop" errors
• Messy tracebacks on Ctrl+C
• Complex shutdown logic that was hard to reason about

Solution

• Remove custom signal handlers that called sys.exit() from async code
• Use try/finally in main() to ensure cleanup() is always called
• Handle KeyboardInterrupt in __init__.py to suppress traceback on Ctrl+C
• Remove unused imports (asyncio, signal, sys) from server.py
• Rename shutdown() to cleanup() to reflect its simpler purpose

This approach:

• Lets Python's default signal handling work naturally
• Ensures database connections are closed on any exit path
• Is simpler and more maintainable

Attribution

Based on work by @ahmedmustahid in #78. This PR extracts and simplifies just the shutdown handling portion.

Test plan

• All existing tests pass (189 passed)
• New tests verify cleanup is called on normal exit and on exception
• Tests verify cleanup handles database close errors gracefully

🤖 Generated with Claude Code

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: d00d517313

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Ensure cleanup runs under cancellation

When the server is interrupted (SIGINT/SIGTERM), asyncio.run() cancels the main() task; that cancellation propagates into the finally block and will cancel the awaited cleanup() call unless it is shielded or the CancelledError is handled. In that case, db_connection.close() may never run, so database connections can remain open on Ctrl+C despite the new shutdown simplification. Consider shielding cleanup() or catching CancelledError so cleanup completes even under cancellation.

Useful? React with 👍 / 👎.