Struct FrameAdapter 

pub struct FrameAdapter { /* private fields */ }

Adapts audio frames to match a VAD backend’s requirements.

Buffers incoming samples and produces frames of the exact size required by the wrapped detector. Also handles sample rate validation.

Implementations

impl FrameAdapter

pub fn new(inner: Box<dyn VoiceActivityDetector>) -> Self

Create a new frame adapter wrapping a VAD detector.

pub fn capabilities(&self) -> &VadCapabilities

Returns the capabilities of the wrapped detector.

pub fn sample_rate(&self) -> u32

Returns the required sample rate.

pub fn frame_size(&self) -> usize

Returns the required frame size in samples.

pub fn process( &mut self, samples: &[i16], sample_rate: u32, ) -> Result<Option<f32>, VadError>

Process audio samples, buffering until a complete frame is available.

Returns Some(probability) when a complete frame was processed, or None if more samples are needed.

Arguments

• samples - Audio samples (any length)
• sample_rate - Sample rate of the input audio

Errors

Returns an error if the sample rate doesn’t match the detector’s requirements.

pub fn process_all( &mut self, samples: &[i16], sample_rate: u32, ) -> Result<Vec<f32>, VadError>

Process all complete frames in the buffer.

Returns a vector of probabilities, one for each complete frame processed. Useful when you want to process multiple frames at once.

pub fn process_latest( &mut self, samples: &[i16], sample_rate: u32, ) -> Result<f32, VadError>

Returns the last probability from processing, or 0.0 if no frame was complete.

This is a convenience method for real-time processing where you only care about the most recent result.

pub fn reset(&mut self)

Reset the adapter and the wrapped detector.

pub fn buffered_samples(&self) -> usize

Returns the number of samples currently buffered.

pub fn timings(&self) -> ProcessTimings

Returns accumulated processing timings from the inner detector.