Struct SentenceChunker 

pub struct SentenceChunker { /* private fields */ }

Available on crate feature audio only.

Buffers LLM tokens and emits complete sentences at delimiter boundaries.

This reduces time-to-first-audio by sending each sentence to TTS as soon as it’s complete, rather than waiting for the full response.

Implementations

impl SentenceChunker

pub fn new() -> SentenceChunker

Create a new chunker with default delimiters (.!?;\n).

pub fn push(&mut self, token: &str) -> Vec<String>

Push a token and return any complete sentences.

pub fn flush(&mut self) -> Option<String>

Flush the remaining buffer as a final sentence.

Trait Implementations

impl Default for SentenceChunker

fn default() -> SentenceChunker

Returns the “default value” for a type.