Refactor PythonResultResolver: Separate state from processing logic

[Copilot]

Refactors PythonResultResolver to separate stateful concerns (ID mappings) from stateless processing logic (payload handling), following the design in resultHandlerRefactor.md.

New Components

• TestItemIndex - Per-workspace state management for ID mappings (runId ↔ TestItem, runId ↔ vsId). Provides O(1) lookups with fallback strategies and stale reference cleanup.

• TestDiscoveryHandler - Stateless singleton for discovery payload processing. Handles error nodes and test tree population.

• TestExecutionHandler - Stateless singleton for execution results. Routes outcomes (success/failure/error/skip/subtest) and returns subtest stats to caller.

• TestCoverageHandler - Stateless singleton for coverage payloads. Creates FileCoverage objects and returns detailed coverage map.

Updated Component

PythonResultResolver is now a pure facade:

// Shared singleton handlers (stateless, safe to share across workspaces)
private static discoveryHandler = new TestDiscoveryHandler();
private static executionHandler = new TestExecutionHandler();
private static coverageHandler = new TestCoverageHandler();

// Per-workspace state
private testItemIndex: TestItemIndex;

public _resolveExecution(payload: ExecutionTestPayload, runInstance: TestRun): void {
    this.subTestStats = PythonResultResolver.executionHandler.processExecution(
        payload, runInstance, this.testItemIndex, this.testController, this
    );
}

Other Changes

• SubtestStats interface moved to types.ts to avoid duplication
• Backward compatibility maintained via getter delegation to TestItemIndex

Original prompt

Fantastic! I love this plan! Now go an execute the switch to the new design

Created from VS Code via the GitHub Pull Request extension.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.