Optimize rendered sample buffering in audio output path

[Solvely-Colin]

Summary

• replace RenderedSampleBuffer's front-removal queue operations with an offset-based queue
• compact backing storage periodically instead of shifting on every dequeue/overflow trim
• add tests covering FIFO dequeue order and bounded-capacity dropping behavior

Why

emoveFirst(...) on Array is O(n) and was used in the render-adjacent buffering path. This change avoids repeated full-array shifts in steady-state callback traffic.

Testing

• Added RenderedSampleBufferTests for queue ordering and capacity trimming
• Could not run swift test in this environment because the Swift toolchain is unavailable (swift command not found).

[Copilot]

Pull request overview

This PR optimizes the RenderedSampleBuffer used in the audio output path by replacing Array.removeFirst(...)-based dequeuing/trimming (O(n) per operation) with an offset-based queue and periodic compaction, reducing steady-state work in render-adjacent code. It also adds unit tests to validate FIFO ordering and bounded-capacity drop behavior.

Changes:

• Track a readOffset into the backing [Float] array to avoid front-removal shifts on every dequeue/trim.
• Periodically compact backing storage (and reset when empty) instead of shifting on each operation.
• Add RenderedSampleBuffer tests for dequeue ordering and capacity trimming.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
Sources/OpenFaderCore/AudioRenderLoop.swift	Reworks RenderedSampleBuffer to use an offset-based queue with periodic compaction to avoid O(n) removeFirst shifting.
Tests/OpenFaderCoreTests/RenderedSampleBufferTests.swift	Adds tests for FIFO dequeue order and dropping oldest samples when exceeding maximum capacity.

---

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