Struct GstreamerAudioCapture 

pub struct GstreamerAudioCapture {
    child: Child,
    stdout: ChildStdout,
    format: AudioFormat,
    next_frame: u64,
    raw_buffer: Vec<u8>,
}

Streaming audio capture wrapping a gst-launch-1.0 child process.

Each Self::next_chunk call reads exactly frames frames of audio from the child’s stdout, packages them as a normalized-f32 AudioChunk, and assigns the appropriate PTS so the chunks form a contiguous timeline.

Fields

child: Child

stdout: ChildStdout

format: AudioFormat

next_frame: u64

raw_buffer: Vec<u8>

Pre-allocated scratch buffer for the raw bytes of one chunk. Sized lazily on first next_chunk; reused after.

Implementations

impl GstreamerAudioCapture

pub fn test_source( format: AudioFormat, frequency_hz: f32, ) -> Result<Self, Error>

Build a capture from audiotestsrc at the given sine frequency_hz. AUT-101’s “deterministic GStreamer source” path.

pub fn from_file(path: &Path, format: AudioFormat) -> Result<Self, Error>

Build a capture by decoding an audio file through GStreamer. Companion to Self::test_source for tests that want to exercise the histogram / waveform code against real audio.

pub fn from_microphone( mic_id: &str, native_id: &str, format: AudioFormat, ) -> Result<Self, Error>

Build a capture from a specific microphone via the per-OS gst element (M-MIC.1 / AUT-278 + M-MIC.3 / AUT-284).

• macOS: osxaudiosrc unique-id=<native_id>
• Linux: pulsesrc device=<native_id>
• Windows: wasapisrc device=<native_id>

When native_id is empty (the device didn’t expose unique-id in gst-device-monitor output, OR the caller wants the OS default), the pipeline falls back to autoaudiosrc which opens the OS default mic.

format must use SampleFormat::F32 — the pipeline caps to F32LE explicitly. Non-F32 rejects at construction (same shape as Self::test_source).

AUT-278 originally described the pipeline as
`…audio/x-raw,format=S16LE,…`. The capture infra around this
type is F32-only (see [`Self::reject_non_f32`] and
[`AudioChunk`] normalisation); reusing it required F32LE.
`audioresample` + `audioconvert` in the pipeline handle the
downstream conversion when a future encoder wants S16.

pub fn format(&self) -> AudioFormat

Format of the produced chunks (same as the format passed at construction).

pub fn frames_emitted(&self) -> u64

Cumulative frames emitted across next_chunk calls. Used by the sync harness (M-MEDIA.7) to assert against expected per-source counts.

pub fn next_chunk(&mut self, frames: u64) -> Result<AudioChunk, Error>

Read exactly frames frames of audio. Returns an AudioChunk with the PTS pointing at the first sample.

Errors

• Error::Io on read failures.
• Error::EndOfStream if the pipeline ends before frames are delivered (returns the partial count for diagnosis).

fn spawn(args: &[&str], format: AudioFormat) -> Result<Self, Error>

fn reject_non_f32(format: AudioFormat) -> Result<(), Error>

Trait Implementations

impl Debug for GstreamerAudioCapture

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Drop for GstreamerAudioCapture

fn drop(&mut self)

Executes the destructor for this type.

fn pin_drop(self: Pin<&mut Self>)

🔬This is a nightly-only experimental API. (pin_ergonomics)

Execute the destructor for this type, but different to Drop::drop, it requires self to be pinned.