@cartesia/cartesia-js 2.2.2 → 2.2.4

package/serialization/resources/tts/types/WebSocketTtsRequest.d.ts

@@ -6,6 +6,7 @@ import * as Cartesia from "../../../../api/index";
 import * as core from "../../../../core";
 import { OutputFormat } from "./OutputFormat";
 import { TtsRequestVoiceSpecifier } from "./TtsRequestVoiceSpecifier";
+import { ModelSpeed } from "./ModelSpeed";
 export declare const WebSocketTtsRequest: core.serialization.ObjectSchema<serializers.WebSocketTtsRequest.Raw, Cartesia.WebSocketTtsRequest>;
 export declare namespace WebSocketTtsRequest {
     interface Raw {
@@ -21,6 +22,6 @@ export declare namespace WebSocketTtsRequest {
         continue?: boolean | null;
         context_id?: string | null;
         max_buffer_delay_ms?: number | null;
-        text_cfg?: number | null;
+        speed?: ModelSpeed.Raw | null;
     }
 }

package/serialization/resources/tts/types/WebSocketTtsRequest.js

@@ -40,6 +40,7 @@ exports.WebSocketTtsRequest = void 0;
 const core = __importStar(require("../../../../core"));
 const OutputFormat_1 = require("./OutputFormat");
 const TtsRequestVoiceSpecifier_1 = require("./TtsRequestVoiceSpecifier");
+const ModelSpeed_1 = require("./ModelSpeed");
 exports.WebSocketTtsRequest = core.serialization.object({
     modelId: core.serialization.property("model_id", core.serialization.string()),
     outputFormat: core.serialization.property("output_format", OutputFormat_1.OutputFormat.optional()),
@@ -53,5 +54,5 @@ exports.WebSocketTtsRequest = core.serialization.object({
     continue: core.serialization.boolean().optional(),
     contextId: core.serialization.property("context_id", core.serialization.string().optional()),
     maxBufferDelayMs: core.serialization.property("max_buffer_delay_ms", core.serialization.number().optional()),
-    textCfg: core.serialization.property("text_cfg", core.serialization.number().optional()),
+    speed: ModelSpeed_1.ModelSpeed.optional(),
 });

package/serialization/resources/tts/types/index.d.ts

@@ -1,5 +1,6 @@
 export * from "./ContextId";
 export * from "./FlushId";
+export * from "./ModelSpeed";
 export * from "./WebSocketBaseResponse";
 export * from "./WebSocketResponse";
 export * from "./WebSocketErrorResponse";

package/serialization/resources/tts/types/index.js

@@ -16,6 +16,7 @@ var __exportStar = (this && this.__exportStar) || function(m, exports) {
 Object.defineProperty(exports, "__esModule", { value: true });
 __exportStar(require("./ContextId"), exports);
 __exportStar(require("./FlushId"), exports);
+__exportStar(require("./ModelSpeed"), exports);
 __exportStar(require("./WebSocketBaseResponse"), exports);
 __exportStar(require("./WebSocketResponse"), exports);
 __exportStar(require("./WebSocketErrorResponse"), exports);

package/version.d.ts

@@ -1 +1 @@
-export declare const SDK_VERSION = "2.2.2";
+export declare const SDK_VERSION = "2.2.4";

package/version.js

@@ -1,4 +1,4 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.SDK_VERSION = void 0;
-exports.SDK_VERSION = "2.2.2";
+exports.SDK_VERSION = "2.2.4";

package/wrapper/Client.d.ts

@@ -1,6 +1,9 @@
 import { CartesiaClient as FernCartesiaClient } from "../Client";
 import { StreamingTTSClient } from "./StreamingTTSClient";
+import { StreamingSTTClient } from "./StreamingSTTClient";
 export declare class CartesiaClient extends FernCartesiaClient {
+    protected _stt: StreamingSTTClient | undefined;
     protected _tts: StreamingTTSClient | undefined;
+    get stt(): StreamingSTTClient;
     get tts(): StreamingTTSClient;
 }

package/wrapper/Client.js

@@ -3,7 +3,12 @@ Object.defineProperty(exports, "__esModule", { value: true });
 exports.CartesiaClient = void 0;
 const Client_1 = require("../Client");
 const StreamingTTSClient_1 = require("./StreamingTTSClient");
+const StreamingSTTClient_1 = require("./StreamingSTTClient");
 class CartesiaClient extends Client_1.CartesiaClient {
+    get stt() {
+        var _a;
+        return ((_a = this._stt) !== null && _a !== void 0 ? _a : (this._stt = new StreamingSTTClient_1.StreamingSTTClient(this._options)));
+    }
     get tts() {
         var _a;
         return ((_a = this._tts) !== null && _a !== void 0 ? _a : (this._tts = new StreamingTTSClient_1.StreamingTTSClient(this._options)));

package/wrapper/StreamingSTTClient.d.ts

@@ -0,0 +1,6 @@
+import { Stt } from "../api/resources/stt/client/Client";
+import SttWebsocket, { SttWebSocketOptions } from "./SttWebsocket";
+export declare class StreamingSTTClient extends Stt {
+    constructor(options?: Stt.Options);
+    websocket(options?: SttWebSocketOptions): SttWebsocket;
+}

package/wrapper/StreamingSTTClient.js

@@ -0,0 +1,17 @@
+"use strict";
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.StreamingSTTClient = void 0;
+const Client_1 = require("../api/resources/stt/client/Client");
+const SttWebsocket_1 = __importDefault(require("./SttWebsocket"));
+class StreamingSTTClient extends Client_1.Stt {
+    constructor(options = {}) {
+        super(options);
+    }
+    websocket(options = {}) {
+        return new SttWebsocket_1.default(options, Object.assign({ cartesiaVersion: "2024-06-10" }, this._options));
+    }
+}
+exports.StreamingSTTClient = StreamingSTTClient;

package/wrapper/SttWebsocket.d.ts

@@ -0,0 +1,29 @@
+import { ReconnectingWebSocket, Options } from "../core/websocket";
+import { Stt } from "../api/resources/stt/client/Client";
+export interface SttWebSocketOptions {
+    model?: string;
+    language?: string;
+    encoding?: string;
+    sampleRate?: number;
+}
+export interface TranscriptionResult {
+    type: "transcript" | "flush_done" | "done" | "error";
+    requestId: string;
+    text?: string;
+    isFinal?: boolean;
+    duration?: number;
+    language?: string;
+    message?: string;
+}
+export default class SttWebsocket {
+    #private;
+    private readonly options;
+    socket?: ReconnectingWebSocket;
+    constructor({ model, language, encoding, sampleRate, }: SttWebSocketOptions, options: Stt.Options);
+    send(audioData: ArrayBuffer): Promise<void>;
+    finalize(): Promise<void>;
+    done(): Promise<void>;
+    onMessage(callback: (result: TranscriptionResult) => void): Promise<void>;
+    connect(connectOptions?: Options): Promise<void>;
+    disconnect(): void;
+}

package/wrapper/SttWebsocket.js

@@ -0,0 +1,198 @@
+"use strict";
+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;
+    };
+})();
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+var __classPrivateFieldSet = (this && this.__classPrivateFieldSet) || function (receiver, state, value, kind, f) {
+    if (kind === "m") throw new TypeError("Private method is not writable");
+    if (kind === "a" && !f) throw new TypeError("Private accessor was defined without a setter");
+    if (typeof state === "function" ? receiver !== state || !f : !state.has(receiver)) throw new TypeError("Cannot write private member to an object whose class did not declare it");
+    return (kind === "a" ? f.call(receiver, value) : f ? f.value = value : state.set(receiver, value)), value;
+};
+var __classPrivateFieldGet = (this && this.__classPrivateFieldGet) || function (receiver, state, kind, f) {
+    if (kind === "a" && !f) throw new TypeError("Private accessor was defined without a getter");
+    if (typeof state === "function" ? receiver !== state || !f : !state.has(receiver)) throw new TypeError("Cannot read private member from an object whose class did not declare it");
+    return kind === "m" ? f : kind === "a" ? f.call(receiver) : f ? f.value : state.get(receiver);
+};
+var _SttWebsocket_instances, _SttWebsocket_isConnected, _SttWebsocket_model, _SttWebsocket_language, _SttWebsocket_encoding, _SttWebsocket_sampleRate, _SttWebsocket_connectionPromise, _SttWebsocket_ensureConnected;
+Object.defineProperty(exports, "__esModule", { value: true });
+const core = __importStar(require("../core"));
+const environments = __importStar(require("../environments"));
+const websocket_1 = require("../core/websocket");
+class SttWebsocket {
+    constructor({ model = "ink-whisper", language = "en", encoding = "pcm_s16le", sampleRate = 16000, }, options) {
+        _SttWebsocket_instances.add(this);
+        this.options = options;
+        _SttWebsocket_isConnected.set(this, false);
+        _SttWebsocket_model.set(this, void 0);
+        _SttWebsocket_language.set(this, void 0);
+        _SttWebsocket_encoding.set(this, void 0);
+        _SttWebsocket_sampleRate.set(this, void 0);
+        _SttWebsocket_connectionPromise.set(this, void 0);
+        __classPrivateFieldSet(this, _SttWebsocket_model, model, "f");
+        __classPrivateFieldSet(this, _SttWebsocket_language, language, "f");
+        __classPrivateFieldSet(this, _SttWebsocket_encoding, encoding, "f");
+        __classPrivateFieldSet(this, _SttWebsocket_sampleRate, sampleRate, "f");
+    }
+    send(audioData) {
+        return __awaiter(this, void 0, void 0, function* () {
+            var _a;
+            yield __classPrivateFieldGet(this, _SttWebsocket_instances, "m", _SttWebsocket_ensureConnected).call(this);
+            (_a = this.socket) === null || _a === void 0 ? void 0 : _a.send(audioData);
+        });
+    }
+    finalize() {
+        return __awaiter(this, void 0, void 0, function* () {
+            var _a;
+            yield __classPrivateFieldGet(this, _SttWebsocket_instances, "m", _SttWebsocket_ensureConnected).call(this);
+            (_a = this.socket) === null || _a === void 0 ? void 0 : _a.send("finalize");
+        });
+    }
+    done() {
+        return __awaiter(this, void 0, void 0, function* () {
+            var _a;
+            yield __classPrivateFieldGet(this, _SttWebsocket_instances, "m", _SttWebsocket_ensureConnected).call(this);
+            (_a = this.socket) === null || _a === void 0 ? void 0 : _a.send("done");
+        });
+    }
+    onMessage(callback) {
+        return __awaiter(this, void 0, void 0, function* () {
+            yield __classPrivateFieldGet(this, _SttWebsocket_instances, "m", _SttWebsocket_ensureConnected).call(this);
+            if (!this.socket) {
+                throw new Error("WebSocket failed to initialize after connection attempt.");
+            }
+            this.socket.addEventListener("message", (event) => {
+                try {
+                    const data = JSON.parse(event.data);
+                    const result = {
+                        type: data.type,
+                        requestId: data.request_id || "",
+                    };
+                    // Handle different message types according to stt.yml spec
+                    if (data.type === "transcript") {
+                        result.text = data.text || "";
+                        result.isFinal = data.is_final || false;
+                        result.duration = data.duration;
+                        result.language = data.language;
+                    }
+                    else if (data.type === "flush_done") {
+                        // Acknowledgment for finalize command
+                        // Only requestId is needed for flush_done
+                    }
+                    else if (data.type === "done") {
+                        // Acknowledgment for done command - session complete
+                        // Only requestId is needed for done
+                    }
+                    else if (data.type === "error") {
+                        result.message = data.message;
+                    }
+                    callback(result);
+                }
+                catch (error) {
+                    callback({
+                        type: "error",
+                        requestId: "",
+                        message: `Failed to parse message: ${error}`,
+                    });
+                }
+            });
+        });
+    }
+    connect() {
+        return __awaiter(this, arguments, void 0, function* (connectOptions = {}) {
+            if (__classPrivateFieldGet(this, _SttWebsocket_isConnected, "f")) {
+                throw new Error("WebSocket is already connected.");
+            }
+            this.socket = new websocket_1.ReconnectingWebSocket(() => __awaiter(this, void 0, void 0, function* () {
+                var _a;
+                const baseUrl = ((_a = (yield core.Supplier.get(this.options.environment))) !== null && _a !== void 0 ? _a : environments.CartesiaEnvironment.Production).replace(/^https?:\/\//, "");
+                const params = {
+                    model: __classPrivateFieldGet(this, _SttWebsocket_model, "f"),
+                    cartesia_version: this.options.cartesiaVersion || "2024-06-10",
+                    encoding: __classPrivateFieldGet(this, _SttWebsocket_encoding, "f"),
+                    sample_rate: __classPrivateFieldGet(this, _SttWebsocket_sampleRate, "f").toString(),
+                };
+                if (__classPrivateFieldGet(this, _SttWebsocket_language, "f"))
+                    params.language = __classPrivateFieldGet(this, _SttWebsocket_language, "f");
+                const apiKey = yield core.Supplier.get(this.options.apiKey);
+                if (apiKey) {
+                    params.api_key = apiKey;
+                }
+                else if (connectOptions.accessToken) {
+                    params.access_token = connectOptions.accessToken;
+                }
+                const queryString = Object.keys(params)
+                    .map(key => `${encodeURIComponent(key)}=${encodeURIComponent(params[key])}`)
+                    .join('&');
+                return `wss://${baseUrl}/stt/websocket?${queryString}`;
+            }), undefined, connectOptions);
+            return new Promise((resolve, reject) => {
+                this.socket.onopen = () => {
+                    __classPrivateFieldSet(this, _SttWebsocket_isConnected, true, "f");
+                    resolve();
+                };
+                this.socket.onclose = () => {
+                    __classPrivateFieldSet(this, _SttWebsocket_isConnected, false, "f");
+                };
+                this.socket.onerror = () => {
+                    reject(new Error("WebSocket failed to connect."));
+                };
+                this.socket.reconnect();
+            });
+        });
+    }
+    disconnect() {
+        var _a;
+        (_a = this.socket) === null || _a === void 0 ? void 0 : _a.close();
+        __classPrivateFieldSet(this, _SttWebsocket_isConnected, false, "f");
+    }
+}
+_SttWebsocket_isConnected = new WeakMap(), _SttWebsocket_model = new WeakMap(), _SttWebsocket_language = new WeakMap(), _SttWebsocket_encoding = new WeakMap(), _SttWebsocket_sampleRate = new WeakMap(), _SttWebsocket_connectionPromise = new WeakMap(), _SttWebsocket_instances = new WeakSet(), _SttWebsocket_ensureConnected = function _SttWebsocket_ensureConnected() {
+    return __awaiter(this, void 0, void 0, function* () {
+        if (__classPrivateFieldGet(this, _SttWebsocket_isConnected, "f"))
+            return;
+        if (!__classPrivateFieldGet(this, _SttWebsocket_connectionPromise, "f")) {
+            __classPrivateFieldSet(this, _SttWebsocket_connectionPromise, this.connect(), "f");
+        }
+        yield __classPrivateFieldGet(this, _SttWebsocket_connectionPromise, "f");
+    });
+};
+exports.default = SttWebsocket;

package/wrapper/Websocket.d.ts

@@ -25,7 +25,7 @@ export default class Websocket {
         source: Source;
         stop: unknown;
     }>;
-    continue(inputs: WebSocketTtsRequest): void;
+    continue(inputs: WebSocketTtsRequest): Promise<void>;
     /**
      * Authenticate and connect to a Cartesia streaming WebSocket.
      *

package/wrapper/Websocket.js

@@ -55,7 +55,7 @@ var __classPrivateFieldGet = (this && this.__classPrivateFieldGet) || function (
 var __importDefault = (this && this.__importDefault) || function (mod) {
     return (mod && mod.__esModule) ? mod : { "default": mod };
 };
-var _Websocket_instances, _Websocket_isConnected, _Websocket_sampleRate, _Websocket_container, _Websocket_encoding, _Websocket_generateId;
+var _Websocket_instances, _Websocket_isConnected, _Websocket_sampleRate, _Websocket_container, _Websocket_encoding, _Websocket_connectionPromise, _Websocket_ensureConnected, _Websocket_generateId;
 Object.defineProperty(exports, "__esModule", { value: true });
 const core = __importStar(require("../core"));
 const environments = __importStar(require("../environments"));
@@ -74,6 +74,7 @@ class Websocket {
         _Websocket_sampleRate.set(this, void 0);
         _Websocket_container.set(this, void 0);
         _Websocket_encoding.set(this, void 0);
+        _Websocket_connectionPromise.set(this, void 0);
         __classPrivateFieldSet(this, _Websocket_sampleRate, sampleRate, "f");
         __classPrivateFieldSet(this, _Websocket_container, container !== null && container !== void 0 ? container : "raw", "f");
         __classPrivateFieldSet(this, _Websocket_encoding, encoding !== null && encoding !== void 0 ? encoding : "pcm_f32le", "f");
@@ -91,9 +92,7 @@ class Websocket {
     send(inputs_1) {
         return __awaiter(this, arguments, void 0, function* (inputs, { timeout = 0 } = {}) {
             var _a, _b, _c, _d;
-            if (!__classPrivateFieldGet(this, _Websocket_isConnected, "f")) {
-                throw new Error("Not connected to WebSocket. Call .connect() first.");
-            }
+            yield __classPrivateFieldGet(this, _Websocket_instances, "m", _Websocket_ensureConnected).call(this);
             if (!inputs.contextId) {
                 inputs.contextId = __classPrivateFieldGet(this, _Websocket_instances, "m", _Websocket_generateId).call(this);
             }
@@ -153,17 +152,17 @@ class Websocket {
         });
     }
     continue(inputs) {
-        var _a;
-        if (!__classPrivateFieldGet(this, _Websocket_isConnected, "f")) {
-            throw new Error("Not connected to WebSocket. Call .connect() first.");
-        }
-        if (!inputs.contextId) {
-            throw new Error("context_id is required to continue a context.");
-        }
-        if (!inputs.outputFormat) {
-            inputs.outputFormat = (0, utils_1.resolveOutputFormat)(__classPrivateFieldGet(this, _Websocket_container, "f"), __classPrivateFieldGet(this, _Websocket_encoding, "f"), __classPrivateFieldGet(this, _Websocket_sampleRate, "f"));
-        }
-        (_a = this.socket) === null || _a === void 0 ? void 0 : _a.send(JSON.stringify(Object.assign({ continue: true }, serializers.WebSocketTtsRequest.jsonOrThrow(inputs, { unrecognizedObjectKeys: "strip" }))));
+        return __awaiter(this, void 0, void 0, function* () {
+            var _a;
+            yield __classPrivateFieldGet(this, _Websocket_instances, "m", _Websocket_ensureConnected).call(this);
+            if (!inputs.contextId) {
+                throw new Error("context_id is required to continue a context.");
+            }
+            if (!inputs.outputFormat) {
+                inputs.outputFormat = (0, utils_1.resolveOutputFormat)(__classPrivateFieldGet(this, _Websocket_container, "f"), __classPrivateFieldGet(this, _Websocket_encoding, "f"), __classPrivateFieldGet(this, _Websocket_sampleRate, "f"));
+            }
+            (_a = this.socket) === null || _a === void 0 ? void 0 : _a.send(JSON.stringify(Object.assign({ continue: true }, serializers.WebSocketTtsRequest.jsonOrThrow(inputs, { unrecognizedObjectKeys: "strip" }))));
+        });
     }
     /**
      * Authenticate and connect to a Cartesia streaming WebSocket.
@@ -226,7 +225,16 @@ class Websocket {
         (_a = this.socket) === null || _a === void 0 ? void 0 : _a.close();
     }
 }
-_Websocket_isConnected = new WeakMap(), _Websocket_sampleRate = new WeakMap(), _Websocket_container = new WeakMap(), _Websocket_encoding = new WeakMap(), _Websocket_instances = new WeakSet(), _Websocket_generateId = function _Websocket_generateId() {
+_Websocket_isConnected = new WeakMap(), _Websocket_sampleRate = new WeakMap(), _Websocket_container = new WeakMap(), _Websocket_encoding = new WeakMap(), _Websocket_connectionPromise = new WeakMap(), _Websocket_instances = new WeakSet(), _Websocket_ensureConnected = function _Websocket_ensureConnected() {
+    return __awaiter(this, void 0, void 0, function* () {
+        if (__classPrivateFieldGet(this, _Websocket_isConnected, "f"))
+            return;
+        if (!__classPrivateFieldGet(this, _Websocket_connectionPromise, "f")) {
+            __classPrivateFieldSet(this, _Websocket_connectionPromise, this.connect(), "f");
+        }
+        yield __classPrivateFieldGet(this, _Websocket_connectionPromise, "f");
+    });
+}, _Websocket_generateId = function _Websocket_generateId() {
     return (0, human_id_1.humanId)({
         separator: "-",
         capitalize: false,