add support for parallel run

[lk-naath]

@jerome3o-anthropic @jspahrsummers not able to add you guys as reviewer, please help me out here.

Ticket: #736

Added pytest configuration to enable parallel test execution and disable logging during tests for faster and cleaner test runs.

Motivation and Context

Running tests sequentially was causing longer test suite execution times, slowing down development and CI feedback. Enabling parallel test runs with multiple workers improves overall testing speed and efficiency.

How Has This Been Tested?

Tests were run locally with the new configuration enabled. Verified that:

• Tests execute in parallel across 4 processes.
• Logging output is disabled during test runs.

Breaking Changes

No breaking changes. This update only affects how tests are executed, not the application code or configurations.

Types of changes

• New feature (non-breaking change which adds functionality)

Checklist

• I have read the MCP Documentation
• My code follows the repository's style guidelines
• New and existing tests pass locally

Additional context

Uses pytest-xdist plugin for parallel test execution with --numprocesses 4
Disables logging during tests to reduce noise and improve test output clarity.

[mcp-shadow]

Positive aspects:

• pytest-xdist is already in the dev dependencies (line 55), so parallel execution is supported
• Parallel test execution can significantly speed up test runs
• The PR addresses issue Leverage Pytest-Xdist for Faster and Cleaner Test Runs #736

Issues and Concerns:

1. Hardcoded parallelism: Setting --numprocesses 4 hardcodes the parallelism level. This might not be optimal for:
   - CI environments with different CPU counts
   - Developer machines with fewer cores
   - Better to use --numprocesses auto or -n auto to automatically detect CPU count
2. Warnings suppression: --disable-warnings is problematic:
   - The project already carefully manages warnings with filterwarnings
   - Suppressing all warnings could hide important deprecation or other warnings
   - This conflicts with the existing warning configuration
3. Logging configuration: The [tool.pytest_env] section is not a standard pytest configuration
   - This might not work as intended
   - Better to use pytest's built-in logging configuration
4. Verbosity: -vv might be too verbose for regular test runs
   - Could clutter CI output
   - Better left to developer preference via command line

Recommendations:

1. Change --numprocesses 4 to --numprocesses auto
2. Remove --disable-warnings to respect existing warning filters
3. Remove -vv from default options
4. Remove the [tool.pytest_env] section and use proper pytest logging configuration if needed

[ihrpr]

Thank you so much for working on this.

Some concerns highlighted in the comments, please can those be addressed?

[lk-naath]

hey @ihrpr, I've implemented the suggested changes to improve the pytest configuration in pyproject.toml:

1. Changed --numprocesses 4 to --numprocesses auto to automatically detect and use optimal CPU cores
2. Removed --disable-warnings to respect the existing warning filters
3. Removed -vv from default options to reduce verbosity in regular test runs
4. Removed the non-standard [tool.pytest_env] section

Please review these changes and let me know if you'd like any further adjustments.

[ihrpr]

Thank you!