fix: align TrackNodeSend test with actual TrackNode API

[ChuxiJ]

Summary

• Fix TrackNodeSend.test.ts calling non-existent updateSendPrePost() method — use updateSendAmount() instead
• Fix all connectSend() calls to use boolean (true/false) instead of string literals ('pre'/'post')
• Test was written against a planned API that diverged from the final implementation in PR feat: wire aux sends to audio engine with pre/post fader toggle #1039

Test plan

• All 6 TrackNodeSend tests pass
• Full test suite passes (3476 tests, 0 failures)
• No production code changes (test-only fix)

Closes #1245

https://claude.ai/code/session_017CJfg9zq4GvRCuPaSWvHKW

[Copilot]

Pull request overview

Updates the TrackNodeSend unit tests to match the actual TrackNode aux-send API (as implemented in PR #1039), unblocking CI by removing calls to a non-existent method.

Changes:

• Replace updateSendPrePost() usage with updateSendAmount(id, amount, preFader) in TrackNodeSend.test.ts.
• Update connectSend() calls to pass preFader: boolean (true/false) instead of 'pre' | 'post'.
• Adjust one test description/comment to reflect the updated API intent.

Comments suppressed due to low confidence (1)

src/engine/tests/TrackNodeSend.test.ts:120

• The comment here says the pre-fader tap point should be sumGain, but TrackNode.connectSend() actually taps preFaderOutput (compressor or latency-comp node) before volumeGain. Also, the test currently has no assertions—updating the comment and asserting which node performed the connect() call would make this test meaningful.

    trackNode.connectSend('send-1', sendDest, 0.5, true);

    // The pre-fader tap point should be the sumGain (before volumeGain)
    // Verify by checking that sumGain.connect was called (for the send gain node)
    // This is an integration-level check — the send gain should connect from sumGain

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

This test case doesn't assert any observable behavior (it will pass as long as no exception is thrown), even though the name suggests it validates switching from post- to pre-fader. Consider asserting the internal send gain state (e.g., pre gain ramps to amount and post gain ramps to 0) similar to src/engine/__tests__/TrackNode.sends.test.ts so the test actually verifies the tap-point switch.

Suggested change

	trackNode.connectSend('send-1', sendDest, 0.5, false);
	trackNode.updateSendAmount('send-1', 0.5, true);
	// No throw = success; pre/post switching happens via gain crossfade
	const amount = 0.5;
	trackNode.connectSend('send-1', sendDest, amount, false);
	trackNode.updateSendAmount('send-1', amount, true);
	// Inspect all gain nodes created by the mock AudioContext
	const gainCreationResults = (ctx.createGain as unknown as ReturnType<typeof vi.fn>).mock
	.results;
	const gainNodes = gainCreationResults.map((r: { value: any }) => r.value);
	// Collect all linearRampToValueAtTime target values across all gain nodes
	const rampCalls = gainNodes.flatMap((node: any) =>
	(node.gain.linearRampToValueAtTime as ReturnType<typeof vi.fn>).mock.calls,
	);
	const rampTargets = rampCalls.map((args: any[]) => args[0]);
	// Verify that we actually scheduled ramps
	expect(rampTargets.length).toBeGreaterThan(0);
	// One gain should ramp up to the send amount (pre-fader), another down to 0 (post-fader)
	expect(rampTargets).toContain(amount);
	expect(rampTargets).toContain(0);