Improve XCM Debugging by Capturing Logs in Unit Tests

[raymondkfcheung]

Description

This PR introduces a lightweight log-capturing mechanism for XCM unit tests, making it easier to troubleshoot failures when needed. It partially addresses #6119 and #6125 by providing an optional way to verify log outputs in tests.

TODOs

• Introduces a structured way to capture logs for XCM debugging.
• Ensures logging remains unobtrusive in passing tests but is available when debugging.

Review Notes:

• This PR does not enforce logging in all tests but provides a reusable mechanism for capturing logs when necessary.
• The implementation is minimal and does not add complexity to existing test setups.
• Developers can still rely on standard logging, sp_tracing::init_for_tests(), when manual debugging is sufficient.

[paritytech-workflow-stopper]

All GitHub workflows were cancelled due to failure one of the required jobs.
Failed workflow url: https://github.com/paritytech/polkadot-sdk/actions/runs/13373075940
Failed job name: fmt

[raymondkfcheung]

Currently, there is no straightforward way to verify if log messages are correctly output during unit tests.

I initially explored using tracing-test, but it only captures logs at the global scope (see source). This means that when testing inside Relay::execute_with(|| { ... }), the logs are not captured in global_buf(), making it unsuitable for verifying log output in such cases.

For sp_tracing::try_init_simple() and sp_tracing::init_for_tests(), they are designed to enable log output but do not provide a way to capture logs for assertions in tests.

• try_init_simple()
• init_for_tests()

Given this, I've implemented a custom log capture mechanism within sp_tracing to support verifying logs in unit tests, particularly for XCM-related tests. Let me know if you have any thoughts on this approach!

[acatangiu]

Given this, I've implemented a custom log capture mechanism within sp_tracing to support verifying logs in unit tests, particularly for XCM-related tests. Let me know if you have any thoughts on this approach!

if this is simple, then sure, it's nice to have, otherwise I don't think we really need to verify this in tests - we have many devs acting as canaries 😆