fix the 440->441 samples issue and pass a noop callback for release

[xianshijing-lk]

this PR does:

1. fix the issues that it will use 440 as number of 10ms samples when sample rate is 44100, which will split the 441 samples into 440 + 1 samples in two callbacks
2. fix the issue that when queue_size_ms is 0, audio_source.rs will passing null callback will crash the app in release mode.
   3

Summary by CodeRabbit

Bug Fixes

• Audio Processing Improvements: Fixed audio queue sizing calculation to ensure accurate rounding behavior and corrected initialization of audio parameters to prevent indeterminate states. Enhanced fast-path handling for scenarios with no audio buffering, improving stability and predictability of audio stream processing.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

📝 Walkthrough

Walkthrough

These changes modify audio source queue initialization across Rust and C++ components, adjusting queue size calculation order, introducing explicit zero-initialization of struct members, and adding a no-op callback for fast-path handling when no buffering is needed.

Changes

Cohort / File(s)	Summary
Audio Queue Calculation & Fast-Path Handling libwebrtc/src/native/audio_source.rs	Queue size calculation reordered from queue_size_ms * (sample_rate / 1000) * num_channels to (queue_size_ms * sample_rate * num_channels) / 1000 to alter evaluation order and rounding behavior; extern "C" no-op callback added for fast-path handling without buffering.
	webrtc-sys/include/livekit/audio_track.h
	webrtc-sys/src/audio_track.cpp

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐰 A queue so swift, now properly aligned,
From calculation reordered with care in mind,
With zero-paths clear and no-ops defined,
Audio buffers dance—synchronized and refined! 🎵

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title directly addresses the two main objectives of the pull request: fixing the 440→441 samples calculation issue and passing a noop callback for release mode stability.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

• 📝 Generate docstrings

---

📜 Recent review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between b092a2d and 3c06381.

📒 Files selected for processing (3)

• libwebrtc/src/native/audio_source.rs
• webrtc-sys/include/livekit/audio_track.h
• webrtc-sys/src/audio_track.cpp

🚧 Files skipped from review as they are similar to previous changes (3)

• webrtc-sys/src/audio_track.cpp
• libwebrtc/src/native/audio_source.rs
• webrtc-sys/include/livekit/audio_track.h

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (10)

• GitHub Check: Build (aarch64-unknown-linux-gnu)
• GitHub Check: Build (aarch64-pc-windows-msvc)
• GitHub Check: Build (aarch64-apple-ios)
• GitHub Check: Build (aarch64-apple-darwin)
• GitHub Check: Build (x86_64-pc-windows-msvc)
• GitHub Check: Build (aarch64-linux-android)
• GitHub Check: Build (x86_64-unknown-linux-gnu)
• GitHub Check: Build (aarch64-apple-ios-sim)
• GitHub Check: Test (x86_64-apple-darwin)
• GitHub Check: Build (x86_64-apple-darwin)

_{✏️ Tip: You can disable this entire section by setting review_details to false in your review settings.}

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[theomonnom]

TIL, makes sense

[ladvoc]

LGTM ✅

[coderabbitai]

Actionable comments posted: 1

🤖 Fix all issues with AI agents

In `@examples/local_video/src/list_devices.rs`:
- Around line 104-111: The formatter error is due to the multi-line expression
building fps_text from fps_list; reformat the block (or run cargo fmt --package
local_video) so the chained iterator/map/collect/join call meets rustfmt style:
locate the fps_text assignment that references fps_list, the iterator .map(|fps|
fps.to_string()).collect::<Vec<String>>().join(", "), and re-run rustfmt (cargo
fmt) or adjust line breaks to match rustfmt expectations.

🧹 Nitpick comments (1)

examples/local_video/Cargo.toml (1)

46-52: Consider removing the top-level objc2 dependency.

objc2 is only needed for macOS, but it's declared twice: once at the top level (line 46) and again under [target.'cfg(target_os = "macos")'.dependencies] (line 52). The top-level declaration will pull in objc2 on all platforms unnecessarily.

♻️ Suggested fix: Remove redundant top-level objc2

 bytemuck = { version = "1.16", features = ["derive"] }
 
-objc2 = { version = "0.6.0", features = ["relax-sign-encoding"] }
-
 nokhwa = { version = "0.10", default-features = false, features = ["output-threaded"] }
 
 [target.'cfg(target_os = "macos")'.dependencies]
 nokhwa = { version = "0.10", default-features = false, features = ["input-avfoundation", "output-threaded"] }
 objc2 = { version = "0.6.0", features = ["relax-sign-encoding"] }

📜 Review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 25f7738 and b092a2d.

⛔ Files ignored due to path filters (1)

• Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (2)

• examples/local_video/Cargo.toml
• examples/local_video/src/list_devices.rs

🧰 Additional context used

🪛 GitHub Actions: Rust Formatting

examples/local_video/src/list_devices.rs

[error] 104-108: cargo fmt -- --check failed: formatting difference detected in list_devices.rs; expected single-line join expression (currently multi-line) after refactor. Run 'cargo fmt' to fix formatting.

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (14)

• GitHub Check: Build (aarch64-pc-windows-msvc)
• GitHub Check: Build (armv7-linux-androideabi)
• GitHub Check: Build (aarch64-linux-android)
• GitHub Check: Build (aarch64-unknown-linux-gnu)
• GitHub Check: Build (x86_64-pc-windows-msvc)
• GitHub Check: Build (x86_64-linux-android)
• GitHub Check: Build (aarch64-apple-ios)
• GitHub Check: Build (x86_64-unknown-linux-gnu)
• GitHub Check: Build (aarch64-apple-ios-sim)
• GitHub Check: Build (x86_64-apple-darwin)
• GitHub Check: Build (aarch64-apple-darwin)
• GitHub Check: Test (x86_64-apple-darwin)
• GitHub Check: Test (x86_64-unknown-linux-gnu)
• GitHub Check: Test (x86_64-pc-windows-msvc)

🔇 Additional comments (4)

examples/local_video/src/list_devices.rs (2)

30-43: LGTM! Good portability improvement.

The refactoring to use only Nokhwa public APIs instead of platform-specific bindings like nokhwa_bindings_macos is a solid approach for cross-platform support. The defensive .clone() on the camera index is appropriate.

---

48-67: LGTM! FourCC enumeration with fallback is well-structured.

The logic correctly tries FourCC-based queries first and falls back to generic camera formats when unavailable. Sorting and deduping the FPS lists ensures consistent output.

examples/local_video/Cargo.toml (2)

7-8: LGTM! Feature gate for desktop binaries.

The empty desktop feature serves as a clean opt-in gate for the desktop-specific binaries. This is a reasonable pattern to prevent accidental compilation on unsupported platforms.

---

48-58: LGTM! Per-target nokhwa configuration is well-structured.

The pattern of declaring a base nokhwa dependency with output-threaded and then overriding with platform-specific input features (input-avfoundation, input-v4l, input-msmf) is correct. Cargo merges features from both declarations, ensuring each platform gets the appropriate input backend.

_{✏️ Tip: You can disable this entire section by setting review_details to false in your review settings.}