Windows port Phase 1: real cpal/WASAPI audio backend
windows-build / windows-build (push) Has been cancelled
cargo-deny / cargo-deny (pull_request) Has been cancelled
windows-build / windows-build (pull_request) Has been cancelled

Replace the Phase 0 no-op CpalBackend stub with a working cpal backend
(WASAPI on Windows), preserving the exact PipeWire AudioBackend contract
so the mixer/encoder/jitter pipeline is unchanged.

- Capture: input stream -> downmix to mono -> 960-sample (20ms) i16 frames
  -> tx, matching the encoder/jitter frame size.
- Playback: 200ms stereo ring prefilled to PLAYBACK_TARGET_SAMPLES; the
  output callback drains it (silence on underrun) while the owning thread
  feeds it from rx. ring_fill is the exact delta-maintained occupancy
  counter (fetch_add on push, fetch_sub on pop), preserving the clock-paced
  production design (not ringbuf's stale occupied_len).
- cpal::Stream is !Send, but AudioBackend is Send+Sync and shared via Arc,
  so each stream lives on its own owning thread (built/played/dropped
  there); the struct holds only the running flag + JoinHandle. stop()
  flips the flag and joins.
- Generic over F32/I16/U16 sample formats; device selected by name else
  default; requires a native 48kHz config (clear error otherwise, no
  resampling yet). Mirrors the PipeWire drain_loop and playout-health line.
- Cargo.toml: add cpal 0.15 under cfg(windows).

Verified by temporarily compiling cpal_impl against real cpal on Linux/ALSA:
build + clippy clean, 6/6 cpal_impl unit tests pass. Reverted to windows-only
gating; shipped Linux state green (316/316). Runtime/WASAPI end-to-end is
unverified and pending a Windows host (plan M2).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
2026-06-18 17:15:56 -04:00
co-authored by Claude Opus 4.8
parent 85b12a26c9
commit 47c58047ce
3 changed files with 838 additions and 59 deletions
+567 -32
View File
@@ -1,31 +1,80 @@
//! Windows audio backend (cpal/WASAPI) — **Phase 0 stub**.
//! Windows audio backend cpal / WASAPI (Phase 1).
//!
//! This is a compile-and-run placeholder so the Windows build links and the app
//! starts up (networking, UI, and text chat all functional) while the real
//! capture/playback implementation lands in Phase 1. Every method satisfies the
//! [`AudioBackend`] contract as a no-op: no microphone is captured and nothing is
//! played. It deliberately pulls in no extra dependency — `cpal` is added only
//! when the real implementation arrives.
//! Implements [`AudioBackend`] on top of [`cpal`], which wraps WASAPI on Windows.
//! It is the Windows counterpart to `pipewire_impl.rs` and deliberately preserves
//! the exact same contract so the rest of the app (mixer, encoder, jitter buffer)
//! is unchanged:
//!
//! Phase 1 will replace this with cpal streams on the WASAPI host, mapping:
//! - `start_capture` → input stream, f32→i16, mono 48 kHz, into `tx`;
//! - `start_playback` → output stream draining a `ringbuf`, keeping `ring_fill`
//! updated so the existing hardware-clock pacing in the mixer keeps working;
//! - `stop` → drop the streams.
//! - **Capture**: mono, 48 kHz, S16 PCM, emitted as `Vec<i16>` frames of
//! [`CAPTURE_FRAME`] (960 = 20 ms) samples — matching the encoder/jitter frame.
//! - **Playback**: stereo interleaved ([`PLAYBACK_CHANNELS`]) S16 PCM at 48 kHz,
//! drained from a ring buffer that is paced to the device's hardware clock via
//! `ring_fill` exactly as the PipeWire backend does.
//!
//! ## Threading and the `!Send` stream
//!
//! `cpal::Stream` is `!Send` (some backends require it to be created and dropped
//! on the same thread), but [`AudioBackend`] is `Send + Sync` and the backend is
//! shared through an `Arc`. So the stream never lives in the struct: each of
//! `start_capture`/`start_playback` spawns one owning thread that builds the
//! stream, plays it, and keeps it alive until the per-worker `running` flag flips
//! (set by `stop`). The struct holds only `Send` handles (the flag + the join
//! handle). The stream's RT callback does the actual audio work; the owning
//! thread additionally feeds the playback ring from the network mixer.
//!
//! ## Sample rate
//!
//! The whole pipeline assumes 48 kHz (Opus + the 960-sample frame). Phase 1 only
//! selects a native-48 kHz device config; if the device can't do 48 kHz we return
//! a clear error rather than silently producing pitch-shifted audio. Arbitrary
//! sample-rate support (resampling) is a Phase 1.1 follow-up.
use std::sync::Arc;
use std::sync::atomic::AtomicUsize;
use std::sync::mpsc::{Receiver, Sender};
use std::sync::atomic::{AtomicBool, AtomicU64, AtomicUsize, Ordering};
use std::sync::mpsc::{Receiver, RecvTimeoutError, Sender};
use std::sync::{Arc, Mutex};
use std::thread::{self, JoinHandle};
use std::time::Duration;
use super::{AudioBackend, AudioError};
use cpal::traits::{DeviceTrait, HostTrait, StreamTrait};
use cpal::{Device, FromSample, Sample, SampleFormat, SampleRate, SizedSample, Stream, StreamConfig};
use ringbuf::{
traits::{Consumer, Producer, Split},
HeapRb,
};
/// No-op Windows audio backend (Phase 0). See module docs.
pub struct CpalBackend;
use super::{AudioBackend, AudioError, PLAYBACK_CHANNELS, PLAYBACK_TARGET_SAMPLES};
/// The one sample rate the pipeline supports (Opus + the 20 ms frame).
const SAMPLE_RATE: u32 = 48_000;
/// Mono capture frame: 960 samples = 20 ms @ 48 kHz. Matches the PipeWire backend
/// and `core::jitter::FRAME_SAMPLES`.
const CAPTURE_FRAME: usize = 960;
/// Playback ring capacity in interleaved samples: 200 ms of stereo @ 48 kHz.
/// Comfortably above [`PLAYBACK_TARGET_SAMPLES`] so the clock-paced producer has
/// headroom and never has to drop frames in steady state.
const RING_CAPACITY: usize = 9600 * PLAYBACK_CHANNELS;
/// How often a blocked playback worker re-checks its `running` flag, bounding how
/// long `stop()` can take to join it (mirrors the PipeWire backend's `WORKER_POLL`).
const WORKER_POLL: Duration = Duration::from_millis(100);
/// Windows audio backend. See module docs.
pub struct CpalBackend {
capture: Mutex<Option<StreamWorker>>,
playback: Mutex<Option<StreamWorker>>,
}
/// A spawned owning thread plus the flag that tells it to drop its stream and exit.
struct StreamWorker {
running: Arc<AtomicBool>,
thread: JoinHandle<()>,
}
impl CpalBackend {
pub fn new() -> Self {
crate::log_msg("CpalBackend: Phase 0 stub active (no audio I/O yet)");
CpalBackend
Self {
capture: Mutex::new(None),
playback: Mutex::new(None),
}
}
}
@@ -38,30 +87,516 @@ impl Default for CpalBackend {
impl AudioBackend for CpalBackend {
fn start_capture(
&self,
_tx: Sender<Vec<i16>>,
_target_node: Option<String>,
tx: Sender<Vec<i16>>,
target_node: Option<String>,
) -> Result<(), AudioError> {
// No capture stream yet: dropping `_tx` simply means no samples are ever
// produced (silent mic), which is the intended Phase 0 behaviour.
crate::log_msg("CpalBackend::start_capture: not yet implemented (Phase 1) — capturing silence");
let mut guard = self.capture.lock().unwrap();
if guard.is_some() {
return Err(AudioError::Stream("Capture already started".to_string()));
}
let running = Arc::new(AtomicBool::new(true));
let running_thread = running.clone();
let thread = thread::Builder::new()
.name("peerspeak-cpal-capture".to_string())
.spawn(move || {
if let Err(e) = run_capture(tx, target_node, running_thread) {
crate::log_msg(&format!("cpal capture error: {e}"));
}
})
.map_err(|e| AudioError::Init(e.to_string()))?;
*guard = Some(StreamWorker { running, thread });
Ok(())
}
fn start_playback(
&self,
rx: Receiver<Vec<i16>>,
_target_node: Option<String>,
_ring_fill: Arc<AtomicUsize>,
target_node: Option<String>,
ring_fill: Arc<AtomicUsize>,
) -> Result<(), AudioError> {
// Drain and discard incoming audio on a detached thread so the mixer's
// producer never blocks or sees a closed channel. This keeps the rest of
// the pipeline running normally while output is silent.
std::thread::spawn(move || while rx.recv().is_ok() {});
crate::log_msg("CpalBackend::start_playback: not yet implemented (Phase 1) — discarding output");
let mut guard = self.playback.lock().unwrap();
if guard.is_some() {
return Err(AudioError::Stream("Playback already started".to_string()));
}
let running = Arc::new(AtomicBool::new(true));
let running_thread = running.clone();
let thread = thread::Builder::new()
.name("peerspeak-cpal-playback".to_string())
.spawn(move || {
if let Err(e) = run_playback(rx, target_node, ring_fill, running_thread) {
crate::log_msg(&format!("cpal playback error: {e}"));
}
})
.map_err(|e| AudioError::Init(e.to_string()))?;
*guard = Some(StreamWorker { running, thread });
Ok(())
}
fn stop(&self) -> Result<(), AudioError> {
for slot in [&self.capture, &self.playback] {
if let Some(worker) = slot.lock().unwrap().take() {
worker.running.store(false, Ordering::Relaxed);
let _ = worker.thread.join();
}
}
Ok(())
}
}
// ---------------------------------------------------------------------------
// Device / config selection
// ---------------------------------------------------------------------------
/// Resolve a device (by `target` name, else the system default) and a stream
/// config running natively at [`SAMPLE_RATE`].
///
/// For output we require [`PLAYBACK_CHANNELS`] (stereo) so the interleaved ring
/// maps 1:1 to the device buffer; for input we prefer mono but accept any channel
/// count and downmix. A device with no 48 kHz config is a hard error (no
/// resampling yet — see module docs).
fn resolve(
output: bool,
target: Option<String>,
) -> Result<(Device, StreamConfig, SampleFormat), AudioError> {
let host = cpal::default_host();
let default = || {
if output {
host.default_output_device()
} else {
host.default_input_device()
}
};
let device = match target {
Some(name) => find_device_by_name(&host, output, &name).or_else(default),
None => default(),
}
.ok_or_else(|| AudioError::Device("no audio device available".to_string()))?;
let supported = choose_config(&device, output)?;
let sample_format = supported.sample_format();
let config = supported.config();
Ok((device, config, sample_format))
}
fn find_device_by_name(host: &cpal::Host, output: bool, name: &str) -> Option<Device> {
let devices = if output {
host.output_devices().ok()?
} else {
host.input_devices().ok()?
};
devices.into_iter().find(|d| d.name().is_ok_and(|n| n == name))
}
/// Pick a supported config at exactly [`SAMPLE_RATE`]. Output must be stereo;
/// input prefers mono, then any channel count (downmixed later).
fn choose_config(
device: &Device,
output: bool,
) -> Result<cpal::SupportedStreamConfig, AudioError> {
let ranges: Vec<cpal::SupportedStreamConfigRange> = if output {
device
.supported_output_configs()
.map_err(|e| AudioError::Device(e.to_string()))?
.collect()
} else {
device
.supported_input_configs()
.map_err(|e| AudioError::Device(e.to_string()))?
.collect()
};
// A range covers a sample-rate span and a fixed channel count.
let supports_48k = |r: &cpal::SupportedStreamConfigRange| {
r.min_sample_rate().0 <= SAMPLE_RATE && SAMPLE_RATE <= r.max_sample_rate().0
};
let pick = |channels: Option<u16>| {
ranges
.iter()
.find(|r| supports_48k(r) && channels.is_none_or(|c| r.channels() == c))
.cloned()
};
let chosen = if output {
pick(Some(PLAYBACK_CHANNELS as u16))
} else {
pick(Some(1)).or_else(|| pick(None))
};
chosen
.map(|r| r.with_sample_rate(SampleRate(SAMPLE_RATE)))
.ok_or_else(|| {
AudioError::Device(format!(
"device '{}' has no {SAMPLE_RATE} Hz {} config; resampling not yet implemented (Phase 1.1)",
device.name().unwrap_or_else(|_| "<unknown>".to_string()),
if output { "stereo output" } else { "input" },
))
})
}
// ---------------------------------------------------------------------------
// Capture
// ---------------------------------------------------------------------------
fn run_capture(
tx: Sender<Vec<i16>>,
target: Option<String>,
running: Arc<AtomicBool>,
) -> Result<(), AudioError> {
let (device, config, sample_format) = resolve(false, target)?;
let channels = config.channels as usize;
let stream = match sample_format {
SampleFormat::F32 => build_input::<f32>(&device, &config, tx, channels),
SampleFormat::I16 => build_input::<i16>(&device, &config, tx, channels),
SampleFormat::U16 => build_input::<u16>(&device, &config, tx, channels),
other => Err(AudioError::Stream(format!(
"unsupported capture sample format: {other:?}"
))),
}?;
stream.play().map_err(|e| AudioError::Stream(e.to_string()))?;
// The RT callback does the work; this thread just keeps `stream` alive until
// `stop()` flips the flag, at which point the stream is dropped (= stopped).
while running.load(Ordering::Relaxed) {
thread::sleep(WORKER_POLL);
}
Ok(())
}
fn build_input<T>(
device: &Device,
config: &StreamConfig,
tx: Sender<Vec<i16>>,
channels: usize,
) -> Result<Stream, AudioError>
where
T: SizedSample + Send + 'static,
i16: FromSample<T>,
{
let mut acc = FrameAccumulator::new(CAPTURE_FRAME);
let err_fn = |e| crate::log_msg(&format!("cpal capture stream error: {e}"));
device
.build_input_stream::<T, _, _>(
config,
move |data: &[T], _| {
for frame in data.chunks_exact(channels) {
let mono = downmix_to_mono(frame);
if let Some(full) = acc.push(mono) {
// Consumer gone (call ended) → stop feeding; the owning
// thread will drop the stream on `stop()`.
if tx.send(full).is_err() {
return;
}
}
}
},
err_fn,
None,
)
.map_err(|e| AudioError::Stream(e.to_string()))
}
/// Average a device frame's channels down to a single mono i16. For a 1-channel
/// device this is just the converted sample.
fn downmix_to_mono<T>(frame: &[T]) -> i16
where
T: Copy,
i16: FromSample<T>,
{
if frame.is_empty() {
return 0;
}
let sum: i32 = frame.iter().map(|&s| i16::from_sample(s) as i32).sum();
(sum / frame.len() as i32) as i16
}
/// Accumulates mono samples into fixed-size [`CAPTURE_FRAME`] frames. Pulled out
/// of the RT callback so the framing is unit-testable.
struct FrameAccumulator {
buf: Vec<i16>,
frame_len: usize,
}
impl FrameAccumulator {
fn new(frame_len: usize) -> Self {
Self {
buf: Vec::with_capacity(frame_len),
frame_len,
}
}
/// Push one sample; returns a completed frame when the buffer fills.
fn push(&mut self, sample: i16) -> Option<Vec<i16>> {
self.buf.push(sample);
if self.buf.len() == self.frame_len {
Some(std::mem::replace(
&mut self.buf,
Vec::with_capacity(self.frame_len),
))
} else {
None
}
}
}
// ---------------------------------------------------------------------------
// Playback
// ---------------------------------------------------------------------------
fn run_playback(
rx: Receiver<Vec<i16>>,
target: Option<String>,
ring_fill: Arc<AtomicUsize>,
running: Arc<AtomicBool>,
) -> Result<(), AudioError> {
let (device, config, sample_format) = resolve(true, target)?;
let rb = HeapRb::<i16>::new(RING_CAPACITY);
let (mut producer, consumer) = rb.split();
// Prefill to the steady-state depth so playout starts at target. `ring_fill`
// is an EXACT occupancy counter maintained by deltas (worker fetch_add on
// push, RT callback fetch_sub on pop) — not ringbuf's cached `occupied_len`,
// which is stale across the split halves and would lie high and starve the
// ring. See pipewire_impl.rs for the full rationale.
for _ in 0..PLAYBACK_TARGET_SAMPLES {
let _ = producer.try_push(0);
}
ring_fill.store(PLAYBACK_TARGET_SAMPLES, Ordering::Relaxed);
// Diagnostics (mirrors the PipeWire backend's playout-health line).
let underrun = Arc::new(AtomicU64::new(0));
let dropped = Arc::new(AtomicU64::new(0));
let stream = match sample_format {
SampleFormat::F32 => {
build_output::<f32, _>(&device, &config, consumer, ring_fill.clone(), underrun.clone())
}
SampleFormat::I16 => {
build_output::<i16, _>(&device, &config, consumer, ring_fill.clone(), underrun.clone())
}
SampleFormat::U16 => {
build_output::<u16, _>(&device, &config, consumer, ring_fill.clone(), underrun.clone())
}
other => Err(AudioError::Stream(format!(
"unsupported playback sample format: {other:?}"
))),
}?;
stream.play().map_err(|e| AudioError::Stream(e.to_string()))?;
let logger = spawn_health_logger(
running.clone(),
ring_fill.clone(),
underrun.clone(),
dropped.clone(),
);
// Feed the ring from the network mixer until `stop()` flips `running` or the
// sender disconnects (call ended). Clock-paced production keeps the ring near
// target, so the drop path below should never fire in steady state.
drain_loop(&rx, &running, |frame| {
if ring_fill.load(Ordering::Relaxed) + frame.len() > RING_CAPACITY {
dropped.fetch_add(1, Ordering::Relaxed);
return;
}
for &sample in &frame {
let _ = producer.try_push(sample);
}
ring_fill.fetch_add(frame.len(), Ordering::Relaxed);
});
// We're shutting down (either stop() or disconnect). Ensure the logger sees it
// even on the disconnect path, then drop the stream.
running.store(false, Ordering::Relaxed);
let _ = logger.join();
drop(stream);
Ok(())
}
fn build_output<T, C>(
device: &Device,
config: &StreamConfig,
mut consumer: C,
ring_fill: Arc<AtomicUsize>,
underrun: Arc<AtomicU64>,
) -> Result<Stream, AudioError>
where
T: SizedSample + FromSample<i16> + Send + 'static,
C: Consumer<Item = i16> + Send + 'static,
{
let err_fn = |e| crate::log_msg(&format!("cpal playback stream error: {e}"));
device
.build_output_stream::<T, _, _>(
config,
move |data: &mut [T], _| {
let (popped, starved) = fill_output(&mut consumer, data);
if starved > 0 {
underrun.fetch_add(starved, Ordering::Relaxed);
}
if popped > 0 {
// Decrement the exact occupancy by what we actually pulled
// (underruns removed nothing) so the mixer paces against the
// true ring depth.
ring_fill.fetch_sub(popped, Ordering::Relaxed);
}
},
err_fn,
None,
)
.map_err(|e| AudioError::Stream(e.to_string()))
}
/// Drain the ring into the device buffer, substituting silence on underrun.
/// Returns `(samples_popped, samples_starved)`. RT-safe (wait-free `try_pop`).
fn fill_output<T, C>(consumer: &mut C, out: &mut [T]) -> (usize, u64)
where
T: Sample + FromSample<i16>,
C: Consumer<Item = i16>,
{
let mut popped = 0usize;
let mut starved = 0u64;
for slot in out.iter_mut() {
match consumer.try_pop() {
Some(v) => {
*slot = T::from_sample(v);
popped += 1;
}
None => {
*slot = T::from_sample(0i16);
starved += 1;
}
}
}
(popped, starved)
}
/// Once-per-second playout-health line (mirrors the PipeWire backend). Quiet
/// unless a second actually glitched, or `PEERSPEAK_AUDIO_VERBOSE` is set.
fn spawn_health_logger(
running: Arc<AtomicBool>,
ring_fill: Arc<AtomicUsize>,
underrun: Arc<AtomicU64>,
dropped: Arc<AtomicU64>,
) -> JoinHandle<()> {
let verbose = std::env::var_os("PEERSPEAK_AUDIO_VERBOSE").is_some();
thread::spawn(move || {
let (mut last_u, mut last_d) = (0u64, 0u64);
while running.load(Ordering::Relaxed) {
thread::sleep(Duration::from_secs(1));
let u = underrun.load(Ordering::Relaxed);
let d = dropped.load(Ordering::Relaxed);
let fill = ring_fill.load(Ordering::Relaxed);
let (du, dd) = (u - last_u, d - last_d);
last_u = u;
last_d = d;
if verbose || du > 0 || dd > 0 {
crate::log_msg(&format!(
"playout-health: fill={fill} samples (~{}ms) | underrun +{du} samples/s (total {u}) | dropped +{dd} frames/s (total {d})",
fill / (48 * PLAYBACK_CHANNELS),
));
}
}
})
}
/// Pump frames from `rx` to `on_frame` until `running` goes false or the sender
/// disconnects. The timed receive re-checks `running` at least every
/// [`WORKER_POLL`], so `stop()` can join the worker promptly instead of hanging
/// on a parked blocking `recv()` (same A7 fix as the PipeWire backend). Pure
/// w.r.t. its inputs, so it's unit-testable.
fn drain_loop(rx: &Receiver<Vec<i16>>, running: &AtomicBool, mut on_frame: impl FnMut(Vec<i16>)) {
while running.load(Ordering::Relaxed) {
match rx.recv_timeout(WORKER_POLL) {
Ok(frame) => on_frame(frame),
Err(RecvTimeoutError::Timeout) => continue,
Err(RecvTimeoutError::Disconnected) => return,
}
}
}
#[cfg(test)]
mod tests {
use super::*;
use std::sync::mpsc;
#[test]
fn downmix_averages_channels() {
assert_eq!(downmix_to_mono::<i16>(&[100, 100]), 100);
assert_eq!(downmix_to_mono::<i16>(&[100, -100]), 0);
assert_eq!(downmix_to_mono::<i16>(&[50]), 50);
assert_eq!(downmix_to_mono::<i16>(&[]), 0);
// 4-channel average rounds toward zero (integer division).
assert_eq!(downmix_to_mono::<i16>(&[10, 20, 30, 41]), 25);
}
#[test]
fn frame_accumulator_emits_full_frames() {
let mut acc = FrameAccumulator::new(3);
assert_eq!(acc.push(1), None);
assert_eq!(acc.push(2), None);
assert_eq!(acc.push(3), Some(vec![1, 2, 3]));
// Resets for the next frame.
assert_eq!(acc.push(4), None);
assert_eq!(acc.push(5), None);
assert_eq!(acc.push(6), Some(vec![4, 5, 6]));
}
#[test]
fn fill_output_pops_then_substitutes_silence() {
let rb = HeapRb::<i16>::new(8);
let (mut prod, mut cons) = rb.split();
for v in [1, 2, 3] {
prod.try_push(v).unwrap();
}
let mut out = [0i16; 5];
let (popped, starved) = fill_output(&mut cons, &mut out);
assert_eq!(popped, 3);
assert_eq!(starved, 2);
assert_eq!(out, [1, 2, 3, 0, 0]);
}
#[test]
fn drain_loop_exits_when_running_flips_even_with_sender_alive() {
let (tx, rx) = mpsc::channel::<Vec<i16>>();
let running = Arc::new(AtomicBool::new(true));
let r2 = running.clone();
let h = thread::spawn(move || drain_loop(&rx, &r2, |_| {}));
thread::sleep(Duration::from_millis(50));
running.store(false, Ordering::Relaxed);
thread::sleep(WORKER_POLL + Duration::from_millis(150));
assert!(
h.is_finished(),
"drain_loop must exit after running=false even while the sender is alive"
);
drop(tx);
h.join().unwrap();
}
#[test]
fn drain_loop_returns_on_disconnect() {
let (tx, rx) = mpsc::channel::<Vec<i16>>();
let running = Arc::new(AtomicBool::new(true));
drop(tx);
drain_loop(&rx, &running, |_| panic!("no frame should arrive"));
}
#[test]
fn drain_loop_delivers_frames() {
let (tx, rx) = mpsc::channel::<Vec<i16>>();
let running = Arc::new(AtomicBool::new(true));
let r2 = running.clone();
let got = Arc::new(Mutex::new(Vec::new()));
let g2 = got.clone();
let h = thread::spawn(move || drain_loop(&rx, &r2, |f| g2.lock().unwrap().push(f)));
tx.send(vec![1, 2, 3]).unwrap();
tx.send(vec![4, 5]).unwrap();
thread::sleep(Duration::from_millis(50));
running.store(false, Ordering::Relaxed);
drop(tx);
h.join().unwrap();
assert_eq!(*got.lock().unwrap(), vec![vec![1, 2, 3], vec![4, 5]]);
}
}