Change Description

Introduced mgo-metric-service, a Python metrics client library that decouples application code from metrics backends. Applications emit counters, gauges, and timings through a single MetricsClient API, with pluggable backends (StatsD, in-memory, no-op) and formatters for metric keys and tags.

Changes Made

• Added MetricsClient with incr, gauge, and timing methods, optional tag support, and debug logging
• Implemented backends: StatsDBackend (StatsD over UDP), InMemoryBackend (records calls for inspection), and NoopBackend (discards metrics)
• Implemented formatters: StatsDFormatter (dot-separated keys with tag validation), InMemoryFormatter (human-readable keys for tests), and NoopFormatter formatting nothing.
• Added FakeMetricsClient with assert_metric_call, assert_incr, assert_gauge, assert_timing, and reset for test assertions
• Updated CI workflow to run make check (lint, format, mypy, pytest with coverage)
• Updated README.md with installation, usage examples, and testing guidance

Testing Instructions for QA Team

This is a Python library, not a deployable HTTP service.
Verification is through automated checks or library usage.

Checklist

• The pull request is linked to the corresponding GitHub issue.
• The developer confirms that this pull request has been verified against the Pull Request Checklist.
• I've read the contributing document
• I've read and understand the Code of Conduct
• I've signed off all commits in this pull request in accordance with the Developer Certificate of Origin (DCO)