kothok_edge_tts/
event.rs

1//! Streaming events emitted during a single synthesis turn.
2
3/// One item in the event stream produced by [`crate::Engine::synthesize`].
4///
5/// Events arrive in order. A well-formed turn starts with zero or more
6/// [`Audio`](TtsEvent::Audio) / [`WordBoundary`](TtsEvent::WordBoundary)
7/// events and terminates with exactly one [`TurnEnd`](TtsEvent::TurnEnd).
8#[derive(Debug, Clone)]
9pub enum TtsEvent {
10    /// A chunk of raw MP3 audio bytes (`audio-24khz-48kbitrate-mono-mp3`).
11    /// Concatenate all `Audio` events in order to reconstruct the full
12    /// utterance.
13    Audio(Vec<u8>),
14
15    /// Word-boundary timing metadata for highlight / karaoke effects.
16    /// `offset` and `duration` are in **100-nanosecond ticks** relative to
17    /// the start of the audio stream.
18    WordBoundary {
19        offset: u64,
20        duration: u64,
21        text: String,
22    },
23
24    /// The server signalled `turn.end` - no more audio or metadata will
25    /// arrive for this request.
26    TurnEnd,
27}
kothok_edge_tts/event.rs

kothok_edge_tts/
event.rs
