js-tts-wrapper
Version:
A JavaScript/TypeScript library that provides a unified API for working with multiple cloud-based Text-to-Speech (TTS) services
323 lines (322 loc) • 12.8 kB
JavaScript
;
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
if (k2 === undefined) k2 = k;
var desc = Object.getOwnPropertyDescriptor(m, k);
if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
desc = { enumerable: true, get: function() { return m[k]; } };
}
Object.defineProperty(o, k2, desc);
}) : (function(o, m, k, k2) {
if (k2 === undefined) k2 = k;
o[k2] = m[k];
}));
var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
Object.defineProperty(o, "default", { enumerable: true, value: v });
}) : function(o, v) {
o["default"] = v;
});
var __importStar = (this && this.__importStar) || (function () {
var ownKeys = function(o) {
ownKeys = Object.getOwnPropertyNames || function (o) {
var ar = [];
for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
return ar;
};
return ownKeys(o);
};
return function (mod) {
if (mod && mod.__esModule) return mod;
var result = {};
if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
__setModuleDefault(result, mod);
return result;
};
})();
Object.defineProperty(exports, "__esModule", { value: true });
exports.ElevenLabsTTSClient = void 0;
const abstract_tts_1 = require("../core/abstract-tts");
const SpeechMarkdown = __importStar(require("../markdown/converter"));
const fetch_utils_1 = require("../utils/fetch-utils");
// Get the fetch implementation for the current environment
const fetch = (0, fetch_utils_1.getFetch)();
/**
* ElevenLabs TTS client
*/
class ElevenLabsTTSClient extends abstract_tts_1.AbstractTTSClient {
/**
* Create a new ElevenLabs TTS client
* @param credentials ElevenLabs credentials
*/
constructor(credentials = {}) {
super(credentials);
/**
* ElevenLabs API key
*/
Object.defineProperty(this, "apiKey", {
enumerable: true,
configurable: true,
writable: true,
value: void 0
});
/**
* Base URL for ElevenLabs API
*/
Object.defineProperty(this, "baseUrl", {
enumerable: true,
configurable: true,
writable: true,
value: "https://api.elevenlabs.io/v1"
});
this.apiKey = credentials.apiKey || process.env.ELEVENLABS_API_KEY || "";
}
/**
* Check if the credentials are valid
* @returns Promise resolving to true if credentials are valid, false otherwise
*/
async checkCredentials() {
if (!this.apiKey) {
console.error("ElevenLabs API key is required");
return false;
}
try {
// Try to list voices to check if the API key is valid
await this._getVoices();
return true;
}
catch (error) {
console.error("Error checking ElevenLabs credentials:", error);
return false;
}
}
/**
* Get available voices from the provider
* @returns Promise resolving to an array of voice objects
*/
async _getVoices() {
try {
const response = await fetch(`${this.baseUrl}/voices`, {
method: "GET",
headers: {
"xi-api-key": this.apiKey,
},
});
if (!response.ok) {
const errorText = await response.text();
console.error(`ElevenLabs API error: ${response.status} ${response.statusText}\nResponse: ${errorText}`);
throw new Error(`Failed to get voices: ${response.statusText}`);
}
const data = await response.json();
return data.voices;
}
catch (error) {
console.error("Error getting ElevenLabs voices:", error);
return [];
}
}
/**
* Prepare text for synthesis by stripping SSML tags
* @param text Text to prepare
* @param options Synthesis options
* @returns Prepared text
*/
async prepareText(text, options) {
let processedText = text;
// Convert from Speech Markdown if requested
if (options?.useSpeechMarkdown && SpeechMarkdown.isSpeechMarkdown(processedText)) {
// Convert to SSML first, then strip SSML tags
const ssml = await SpeechMarkdown.toSSML(processedText);
processedText = this._stripSSML(ssml);
}
// If text is SSML, strip the tags as ElevenLabs doesn't support SSML
// and has its own emotion analysis
if (this._isSSML(processedText)) {
processedText = this._stripSSML(processedText);
}
return processedText;
}
/**
* Convert text to audio bytes
* @param text Text to synthesize
* @param options Synthesis options
* @returns Promise resolving to audio bytes
*/
async synthToBytes(text, options) {
try {
// Use voice from options or the default voice
const voiceId = options?.voice || this.voiceId || "21m00Tcm4TlvDq8ikWAM"; // Default voice (Rachel)
// Prepare text for synthesis (strip SSML tags)
const preparedText = await this.prepareText(text, options);
// Prepare request options
const requestOptions = {
method: "POST",
headers: {
"Content-Type": "application/json",
"xi-api-key": this.apiKey,
},
body: JSON.stringify({
text: preparedText,
model_id: "eleven_monolingual_v1",
output_format: options?.format === "mp3" ? "mp3_44100_128" : "pcm_44100",
voice_settings: {
stability: 0.5,
similarity_boost: 0.75,
use_speaker_boost: true,
style: 0,
speed: typeof this.properties.rate === "number" ? this.properties.rate : 1.0,
},
}),
};
// Make API request
const response = await fetch(`${this.baseUrl}/text-to-speech/${voiceId}`, requestOptions);
if (!response.ok) {
const errorText = await response.text();
console.error(`ElevenLabs API error: ${response.status} ${response.statusText}\nResponse: ${errorText}`);
throw new Error(`Failed to synthesize speech: ${response.statusText}`);
}
// Get audio data as array buffer
const arrayBuffer = await response.arrayBuffer();
return new Uint8Array(arrayBuffer);
}
catch (error) {
console.error("Error synthesizing speech:", error);
throw error;
}
}
/**
* Synthesize text to a byte stream
* @param text Text to synthesize
* @param options Synthesis options
* @returns Promise resolving to an object containing the audio stream and an empty word boundaries array
*/
async synthToBytestream(text, options) {
try {
// Use voice from options or the default voice
const voiceId = options?.voice || this.voiceId || "21m00Tcm4TlvDq8ikWAM"; // Default voice (Rachel)
// Prepare text for synthesis (strip SSML tags)
const preparedText = await this.prepareText(text, options);
// Prepare request options
const requestOptions = {
method: "POST",
headers: {
"Content-Type": "application/json",
"xi-api-key": this.apiKey,
},
body: JSON.stringify({
text: preparedText,
model_id: "eleven_monolingual_v1",
output_format: options?.format === "mp3" ? "mp3_44100_128" : "pcm_44100",
voice_settings: {
stability: 0.5,
similarity_boost: 0.75,
use_speaker_boost: true,
style: 0,
speed: typeof this.properties.rate === "number" ? this.properties.rate : 1.0,
},
}),
};
// Make API request with streaming option
const response = await fetch(`${this.baseUrl}/text-to-speech/${voiceId}/stream`, requestOptions);
if (!response.ok) {
const errorText = await response.text();
console.error(`ElevenLabs API error: ${response.status} ${response.statusText}\nResponse: ${errorText}`);
throw new Error(`Failed to synthesize speech stream: ${response.statusText}`);
}
// Convert the response body to a proper ReadableStream
const responseArrayBuffer = await response.arrayBuffer();
const uint8Array = new Uint8Array(responseArrayBuffer);
// Create a ReadableStream from the Uint8Array
const readableStream = new ReadableStream({
start(controller) {
controller.enqueue(uint8Array);
controller.close();
},
});
return { audioStream: readableStream, wordBoundaries: [] };
}
catch (error) {
console.error("Error synthesizing speech stream:", error);
throw error;
}
}
/**
* Start playback with word boundary callbacks
* @param text Text to speak
* @param callback Callback function for word boundaries
* @param options Synthesis options
*/
async startPlaybackWithCallbacks(text, callback, options) {
// Register the callback
this.on("boundary", callback);
// Start playback
await this.speakStreamed(text, options);
}
/**
* Map ElevenLabs voice objects to unified format
* @param rawVoices Array of ElevenLabs voice objects
* @returns Promise resolving to an array of unified voice objects
*/
async _mapVoicesToUnified(rawVoices) {
// Map raw voices directly without language normalization for now
return rawVoices.map((voice) => ({
id: voice.voice_id,
name: voice.name,
gender: undefined, // ElevenLabs doesn't provide gender
languageCodes: [
{
bcp47: voice.labels?.accent || "en-US",
iso639_3: (voice.labels?.accent || "en-US").split("-")[0] || "eng",
display: voice.labels?.accent || "English",
},
],
provider: "elevenlabs",
}));
}
/**
* Get voice by ID
* @param voiceId Voice ID
* @returns Promise resolving to voice details
*/
async getVoice(voiceId) {
try {
const response = await fetch(`${this.baseUrl}/voices/${voiceId}`, {
method: "GET",
headers: {
"xi-api-key": this.apiKey,
},
});
if (!response.ok) {
if (response.status === 404) {
return null;
}
const errorText = await response.text();
console.error(`ElevenLabs API error: ${response.status} ${response.statusText}\nResponse: ${errorText}`);
throw new Error(`Failed to get voice: ${response.statusText}`);
}
const voice = await response.json();
// Map to unified format using the same logic as _mapVoicesToUnified
const unifiedVoice = {
id: voice.voice_id,
name: voice.name,
gender: voice.labels?.gender === "female"
? "Female"
: voice.labels?.gender === "male"
? "Male"
: "Unknown",
languageCodes: [
{
bcp47: voice.labels?.language || "en-US",
iso639_3: voice.labels?.language?.split("-")[0] || "eng",
display: voice.labels?.accent || "English",
},
],
provider: "elevenlabs",
};
return unifiedVoice;
}
catch (error) {
console.error("Error getting voice:", error);
throw error;
}
}
}
exports.ElevenLabsTTSClient = ElevenLabsTTSClient;