Singleton runtime coordinating TTS engine swaps and playback in React Native.
// Source: https://github.com/a-ghorbani/pocketpal-ai | License: MIT | Copyright 2024 Asghar Ghorbani
// Provided via justclaw.it.com; keep this notice with the code.
import Speech from '@pocketpalai/react-native-speech';
import type {Engine, EngineId} from './types';
/**
* Single-slot coordinator for the TTS native runtime.
*
* The fork's `Speech` module is global — only one neural engine has its
* model loaded at a time. The previous design tracked initialization
* per-engine-wrapper, which lied: a wrapper said `initialized=true`
* even after another engine had silently swapped it out at the native
* layer.
*
* `TTSRuntime` owns the truth ("which engine is currently loaded in
* native land") and serializes all engine swaps and initialization
* calls. Engine wrappers route th
... (truncated -- full source via MCP)
See the full source, get the GitHub permalink, and search for more like it.
Get a free API key