Consolidate serial and parallel runtimes

[davidjgoss]

🤔 What's changed?

Removes duplication of logic in serial and parallel runtimes by creating shared Coordinator and Worker classes that are used in both flavours, and keeping runtime-specific code to a minimum in adapters.

Also:

• Commands and events between coordinator and worker are now structured less ambiguously with a type discriminator
• Simplified usage of stopwatch functions
• Generally less passing around of things like eventDataCollector to get state that can be known upfront and passed in via arguments

⚡️ What's your motivation?

Part of #2419.

This is just the first step, and will next enable us to start plumbing in a plugins architecture for things that happen at the worker level, like retries.

Also, I wanted to take this further and have the "queue of work that is distributed among potentially multiple workers" concept pulled up to Coordinator to enable better testability, but that looked like a bridge too far so I may tackle it in a future PR.

🏷️ What kind of change is this?

• 🏦 Refactoring/debt/DX (improvement to code design, tooling, etc. without changing behaviour)

📋 Checklist:

• I agree to respect and uphold the Cucumber Community Code of Conduct
• I've changed the behaviour of the code
  • I have added/updated tests to cover my changes.
• My change requires a change to the documentation.
  • I have updated the documentation accordingly.
• Users should know about my change
  • I have added an entry to the "Unreleased" section of the CHANGELOG, linking to this pull request.

---

This text was originally generated from a template, then edited by hand. You can modify the template here.

[coveralls]

coverage: 97.838% (-0.09%) from 97.929%
when pulling e284770 on feat/unified-runner
into aa3ce7b on main.