edge-tts-node - npm Package Compare versions

Comparing version 1.3.1 to 1.3.5

dist/MsEdgeTTS.d.ts

@@ -1,3 +0,1 @@
-/// <reference types="node" />
-/// <reference types="node" />
 import { OUTPUT_FORMAT } from "./OUTPUT_FORMAT";
@@ -41,3 +39,3 @@ import { Readable } from "stream";
     static OUTPUT_FORMAT: typeof OUTPUT_FORMAT;
-    private static TRUSTED_CLIENT_TOKEN;
+    static TRUSTED_CLIENT_TOKEN: string;
     private static VOICES_URL;
@@ -56,2 +54,4 @@ private static SYNTH_URL;
     private readonly _agent;
+    private _arraybuffer;
+    private state;
     private _log;
@@ -67,55 +67,13 @@ /**
     private _initClient;
+    private _pushData;
     private _pushAudioData;
     private _SSMLTemplate;
-    /**
-     * Fetch the list of voices available in Microsoft Edge.
-     * These, however, are not all. The complete list of voices supported by this module [can be found here](https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/language-support) (neural, standard, and preview).
-     */
     getVoices(): Promise<Voice[]>;
-    /**
-     * Sets the required information for the speech to be synthesised and inits a new WebSocket connection.
-     * Must be called at least once before text can be synthesised.
-     * Saved in this instance. Can be called at any time times to update the metadata.
-     *
-     * @param voiceName a string with any `ShortName`. A list of all available neural voices can be found [here](https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/language-support#neural-voices). However, it is not limited to neural voices: standard voices can also be used. A list of standard voices can be found [here](https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/language-support#standard-voices)
-     * @param outputFormat any {@link OUTPUT_FORMAT}
-     * @param voiceLocale (optional) any voice locale that is supported by the voice. See the list of all voices for compatibility. If not provided, the locale will be inferred from the `voiceName`
-     */
+    setConfig(conf: any): void;
     setMetadata(voiceName: string, outputFormat: OUTPUT_FORMAT, voiceLocale?: string): Promise<void>;
     private _metadataCheck;
-    /**
-     * Close the WebSocket connection.
-     */
     close(): void;
-    /**
-     * Writes raw audio synthesised from text to a file. Uses a basic {@link _SSMLTemplate SML template}.
-     *
-     * @param path a valid output path, including a filename and file extension.
-     * @param input the input to synthesise
-     * @param options (optional) {@link ProsodyOptions}
-     * @returns {Promise<string>} - a `Promise` with the full filepath
-     */
     toFile(path: string, input: string, options?: ProsodyOptions): Promise<string>;
-    /**
-     * Writes raw audio synthesised from text in real-time to a {@link Readable}. Uses a basic {@link _SSMLTemplate SML template}.
-     *
-     * @param input the text to synthesise. Can include SSML elements.
-     * @param options (optional) {@link ProsodyOptions}
-     * @returns {Readable} - a `stream.Readable` with the audio data
-     */
     toStream(input: string, options?: ProsodyOptions): Readable;
-    /**
-     * Writes raw audio synthesised from text to a file. Has no SSML template. Basic SSML should be provided in the request.
-     *
-     * @param path a valid output path, including a filename and file extension.
-     * @param requestSSML the SSML to send. SSML elements required in order to work.
-     * @returns {Promise<string>} - a `Promise` with the full filepath
-     */
     rawToFile(path: string, requestSSML: string): Promise<string>;
-    /**
-     * Writes raw audio synthesised from a request in real-time to a {@link Readable}. Has no SSML template. Basic SSML should be provided in the request.
-     *
-     * @param requestSSML the SSML to send. SSML elements required in order to work.
-     * @returns {Readable} - a `stream.Readable` with the audio data
-     */
     rawToStream(requestSSML: string): Readable;
@@ -122,0 +80,0 @@ private _rawSSMLRequestToFile;

dist/MsEdgeTTS.js

@@ -33,6 +33,7 @@ "use strict";
 const buffer_1 = require("buffer");
-const crypto_1 = require("crypto");
+const randombytes_1 = __importDefault(require("randombytes"));
 const OUTPUT_FORMAT_1 = require("./OUTPUT_FORMAT");
 const stream_1 = require("stream");
 const fs = __importStar(require("fs"));
+const utils_1 = require("./utils");
 class ProsodyOptions {
@@ -60,3 +61,3 @@ /**
 class MsEdgeTTS {
-    static wordBoundaryEnabled = false;
+    static wordBoundaryEnabled = true;
     static OUTPUT_FORMAT = OUTPUT_FORMAT_1.OUTPUT_FORMAT;
@@ -77,4 +78,10 @@ static TRUSTED_CLIENT_TOKEN = "6A5AA1D4EAFF4E9FB37E23D68491D6F4";
     _agent;
+    _arraybuffer = false;
+    state = {
+        offsetCompensation: 0,
+        lastDurationOffset: 0
+    };
     _log(...o) {
         if (this._enableLogger) {
+            o.unshift('edgetts:');
             console.log(...o);
@@ -103,3 +110,3 @@ }
         this._ws.send(message, () => {
-            this._log("<- sent message: ", message);
+            //this._log("<- sent message: ", message);
         });
@@ -111,3 +118,4 @@ }
             : new isomorphic_ws_1.default(MsEdgeTTS.SYNTH_URL, { agent: this._agent });
-        this._ws.binaryType = "arraybuffer";
+        if (this._arraybuffer)
+            this._ws.binaryType = "arraybuffer";
         return new Promise((resolve, reject) => {
@@ -133,20 +141,77 @@ this._ws.onopen = () => {
             this._ws.onmessage = (m) => {
-                const buffer = buffer_1.Buffer.from(m.data);
-                const message = buffer.toString();
-                const requestId = /X-RequestId:(.*?)\r\n/gm.exec(message)[1];
-                if (message.includes("Path:turn.start")) {
-                    // start of turn, ignore
+                this._log("type:::::::: ", typeof m.data);
+                let mdata = m.data;
+                if (typeof mdata === 'string') {
+                    const encodedData = buffer_1.Buffer.from(mdata, 'utf8');
+                    const message = mdata;
+                    const requestId = /X-RequestId:(.*?)\r\n/gm.exec(message)[1];
+                    let [headers, data] = (0, utils_1.getHeadersAndData)(encodedData, encodedData.indexOf("\r\n\r\n"));
+                    const path = headers['Path'];
+                    if (path === "audio.metadata") {
+                        let parsedMetadata = (0, utils_1.parseMetadata)(data, this.state["offsetCompensation"]);
+                        this._pushData(parsedMetadata, requestId);
+                        // 更新上一次的持续时间偏移量，用于下一次 SSML 请求
+                        this.state["lastDurationOffset"] = parsedMetadata["offset"] + parsedMetadata["duration"];
+                    }
+                    else if (path === "turn.end") {
+                        this.state["offsetCompensation"] = this.state["lastDurationOffset"];
+                        this.state["offsetCompensation"] += 8750000;
+                    }
+                    else if (path !== "response" && path !== "turn.start") {
+                        // 如果路径不是 "response" 或 "turn.start"
+                        throw new Error("Unknown path received"); // 抛出未知响应错误
+                    }
                 }
-                else if (message.includes("Path:turn.end")) {
-                    // end of turn, close stream
-                    this._streams[requestId].push(null);
+                else if (buffer_1.Buffer.isBuffer(mdata)) {
+                    const message = mdata.toString();
+                    const requestId = /X-RequestId:(.*?)\r\n/gm.exec(message)[1];
+                    const headerLength = mdata.readUInt16BE(0);
+                    if (headerLength > mdata.length) {
+                        throw new Error("The header length is greater than the length of the data.");
+                    }
+                    // Parse the headers and data from the binary message.
+                    let [headers, data] = (0, utils_1.getHeadersAndData)(mdata, headerLength);
+                    if (headers['Path'] !== 'audio') {
+                        throw new Error("Received binary message, but the path is not audio.");
+                    }
+                    const contentType = headers['Content-Type'];
+                    if (contentType !== 'audio/mpeg' && contentType !== undefined) {
+                        throw new Error("Received binary message, but with an unexpected Content-Type.");
+                    }
+                    // We only allow no Content-Type if there is no data.
+                    if (contentType === undefined) {
+                        if (data.length === 0) {
+                            return;
+                        }
+                        // If the data is not empty, then we need to raise an exception.
+                        throw new Error("Received binary message with no Content-Type, but with data.");
+                    }
+                    // If the data is empty now, then we need to raise an exception.
+                    if (data.length === 0) {
+                        throw new Error("Received binary message, but it is missing the audio data.");
+                    }
+                    this._pushData({ type: "audio", data: data }, requestId);
                 }
-                else if (message.includes("Path:response")) {
-                    // context response, ignore
-                }
-                else if (message.includes("Path:audio") && m.data instanceof ArrayBuffer) {
-                    this._pushAudioData(buffer, requestId);
-                }
                 else {
-                    this._log("UNKNOWN MESSAGE", message);
+                    mdata = buffer_1.Buffer.isBuffer(mdata) ? mdata : mdata['data'];
+                    const buffer = buffer_1.Buffer.from(mdata);
+                    const message = buffer.toString();
+                    const requestId = /X-RequestId:(.*?)\r\n/gm.exec(message)[1];
+                    this._log(message.includes("Path:audio"), buffer_1.Buffer.isBuffer(mdata), mdata instanceof ArrayBuffer);
+                    if (message.includes("Path:turn.start")) {
+                        // start of turn, ignore
+                    }
+                    else if (message.includes("Path:turn.end")) {
+                        // end of turn, close stream
+                        this._streams[requestId].push(null);
+                    }
+                    else if (message.includes("Path:response")) {
+                        // context response, ignore
+                    }
+                    else if (message.includes("Path:audio") && buffer_1.Buffer.isBuffer(mdata)) {
+                        this._pushAudioData(buffer, requestId);
+                    }
+                    else {
+                        //this._log("UNKNOWN MESSAGE", message);
+                    }
                 }
@@ -165,2 +230,6 @@ };
     }
+    _pushData(data, requestId) {
+        data = typeof data == "string" ? data : JSON.stringify(data);
+        this._streams[requestId].push(data, 'utf8');
+    }
     _pushAudioData(audioBuffer, requestId) {
@@ -170,3 +239,3 @@ const audioStartIndex = audioBuffer.indexOf(MsEdgeTTS.BINARY_DELIM) + MsEdgeTTS.BINARY_DELIM.length;
         this._streams[requestId].push(audioData);
-        this._log("received audio chunk, size: ", audioData?.length);
+        this._log("_pushAudioData: received audio chunk, size: ", audioData?.length);
     }
@@ -184,6 +253,2 @@ _SSMLTemplate(input, options = {}) {
     }
-    /**
-     * Fetch the list of voices available in Microsoft Edge.
-     * These, however, are not all. The complete list of voices supported by this module [can be found here](https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/language-support) (neural, standard, and preview).
-     */
     getVoices() {
@@ -196,11 +261,5 @@ return new Promise((resolve, reject) => {
     }
-    /**
-     * Sets the required information for the speech to be synthesised and inits a new WebSocket connection.
-     * Must be called at least once before text can be synthesised.
-     * Saved in this instance. Can be called at any time times to update the metadata.
-     *
-     * @param voiceName a string with any `ShortName`. A list of all available neural voices can be found [here](https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/language-support#neural-voices). However, it is not limited to neural voices: standard voices can also be used. A list of standard voices can be found [here](https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/language-support#standard-voices)
-     * @param outputFormat any {@link OUTPUT_FORMAT}
-     * @param voiceLocale (optional) any voice locale that is supported by the voice. See the list of all voices for compatibility. If not provided, the locale will be inferred from the `voiceName`
-     */
+    setConfig(conf) {
+        this._arraybuffer = conf["arraybuffer"] ?? false;
+    }
     async setMetadata(voiceName, outputFormat, voiceLocale) {
@@ -232,26 +291,8 @@ const oldVoice = this._voice;
     }
-    /**
-     * Close the WebSocket connection.
-     */
     close() {
         this._ws.close();
     }
-    /**
-     * Writes raw audio synthesised from text to a file. Uses a basic {@link _SSMLTemplate SML template}.
-     *
-     * @param path a valid output path, including a filename and file extension.
-     * @param input the input to synthesise
-     * @param options (optional) {@link ProsodyOptions}
-     * @returns {Promise<string>} - a `Promise` with the full filepath
-     */
     toFile(path, input, options) {
         return this._rawSSMLRequestToFile(path, this._SSMLTemplate(input, options));
     }
-    /**
-     * Writes raw audio synthesised from text in real-time to a {@link Readable}. Uses a basic {@link _SSMLTemplate SML template}.
-     *
-     * @param input the text to synthesise. Can include SSML elements.
-     * @param options (optional) {@link ProsodyOptions}
-     * @returns {Readable} - a `stream.Readable` with the audio data
-     */
     toStream(input, options) {
@@ -261,18 +302,5 @@ const { stream } = this._rawSSMLRequest(this._SSMLTemplate(input, options));
     }
-    /**
-     * Writes raw audio synthesised from text to a file. Has no SSML template. Basic SSML should be provided in the request.
-     *
-     * @param path a valid output path, including a filename and file extension.
-     * @param requestSSML the SSML to send. SSML elements required in order to work.
-     * @returns {Promise<string>} - a `Promise` with the full filepath
-     */
     rawToFile(path, requestSSML) {
         return this._rawSSMLRequestToFile(path, requestSSML);
     }
-    /**
-     * Writes raw audio synthesised from a request in real-time to a {@link Readable}. Has no SSML template. Basic SSML should be provided in the request.
-     *
-     * @param requestSSML the SSML to send. SSML elements required in order to work.
-     * @returns {Readable} - a `stream.Readable` with the audio data
-     */
     rawToStream(requestSSML) {
@@ -303,3 +331,3 @@ const { stream } = this._rawSSMLRequest(requestSSML);
         this._metadataCheck();
-        const requestId = (0, crypto_1.randomBytes)(16).toString("hex");
+        const requestId = (0, randombytes_1.default)(16).toString("hex");
         const request = `X-RequestId:${requestId}\r\nContent-Type:application/ssml+xml\r\nPath:ssml\r\n\r\n
@@ -313,2 +341,3 @@ ` + requestSSML.trim();
             destroy(error, callback) {
+                self._log("+_+_+_+__+_", error);
                 delete self._streams[requestId];
@@ -315,0 +344,0 @@ callback(error);

package.json

 {
     "name": "edge-tts-node",
-    "version": "1.3.1",
+    "version": "1.3.5",
     "description": "An Azure Speech Service module that uses the Microsoft Edge Read Aloud API.",
@@ -40,2 +40,3 @@ "author": "Migushthe2nd <[e-mail]>",
         "process": "^0.11.10",
+        "randombytes": "^2.1.0",
         "stream-browserify": "^3.0.0",
@@ -42,0 +43,0 @@ "ws": "^8.14.1"

README.md

@@ -1,4 +0,5 @@
-# MsEdgeTTS
-[![npm version](https://badge.fury.io/js/msedge-tts.svg)](https://badge.fury.io/js/msedge-tts)
+# edge-tts-node
 
+[![npm version](https://badge.fury.io/js/edge-tts-node.svg)](https://badge.fury.io/js/edge-tts-node)
+
 An simple Azure Speech Service module that uses the Microsoft Edge Read Aloud API.
@@ -31,15 +32,18 @@
 ```js
-import {MsEdgeTTS, OUTPUT_FORMAT} from "msedge-tts";
+import { MsEdgeTTS, OUTPUT_FORMAT } from "edge-tts-node";
 
 const tts = new MsEdgeTTS();
-await tts.setMetadata("en-IE-ConnorNeural", OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS);
+await tts.setMetadata(
+  "en-IE-ConnorNeural",
+  OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS
+);
 const readable = tts.toStream("Hi, how are you?");
 
 readable.on("data", (data) => {
-    console.log("DATA RECEIVED", data);
-    // raw audio file data
+  console.log("DATA RECEIVED", data);
+  // raw audio file data
 });
 
 readable.on("close", () => {
-    console.log("STREAM CLOSED");
+  console.log("STREAM CLOSED");
 });
@@ -51,8 +55,11 @@ ```
 ```js
-import {MsEdgeTTS, OUTPUT_FORMAT} from "msedge-tts";
+import { MsEdgeTTS, OUTPUT_FORMAT } from "edge-tts-node";
 
 (async () => {
-    const tts = new MsEdgeTTS();
-    await tts.setMetadata("en-US-AriaNeural", OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS);
-    const filePath = await tts.toFile("./example_audio.webm", "Hi, how are you?");  
+  const tts = new MsEdgeTTS();
+  await tts.setMetadata(
+    "en-US-AriaNeural",
+    OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS
+  );
+  const filePath = await tts.toFile("./example_audio.webm", "Hi, how are you?");
 })();
@@ -62,9 +69,17 @@ ```
 ### Change voice rate, pitch and volume
+
 ```js
-import {MsEdgeTTS, OUTPUT_FORMAT} from "msedge-tts";
+import { MsEdgeTTS, OUTPUT_FORMAT } from "edge-tts-node";
 
 (async () => {
-    const tts = new MsEdgeTTS();
-    await tts.setMetadata("en-US-AriaNeural", OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS);
-    const filePath = await tts.toFile("./example_audio.webm", "Hi, how are you?", {rate: 0.5, pitch: "+200Hz"});
+  const tts = new MsEdgeTTS();
+  await tts.setMetadata(
+    "en-US-AriaNeural",
+    OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS
+  );
+  const filePath = await tts.toFile(
+    "./example_audio.webm",
+    "Hi, how are you?",
+    { rate: 0.5, pitch: "+200Hz" }
+  );
 })();
@@ -74,11 +89,18 @@ ```
 ### Use an alternative HTTP Agent
+
 Use a custom http.Agent implementation like [https-proxy-agent](https://github.com/TooTallNate/proxy-agents) or [socks-proxy-agent](https://github.com/TooTallNate/proxy-agents/tree/main/packages/socks-proxy-agent).
+
 ```js
-import {SocksProxyAgent} from 'socks-proxy-agent';
+import { SocksProxyAgent } from "socks-proxy-agent";
 
 (async () => {
-    const agent = new SocksProxyAgent("socks://your-name%40gmail.com:[e-mail]")
-    const tts = new MsEdgeTTS(agent);
-    await tts.setMetadata("en-US-AriaNeural", OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS);
-    const filePath = await tts.toFile("./example_audio.webm", "Hi, how are you?");
+  const agent = new SocksProxyAgent(
+    "socks://your-name%40gmail.com:[e-mail]"
+  );
+  const tts = new MsEdgeTTS(agent);
+  await tts.setMetadata(
+    "en-US-AriaNeural",
+    OUTPUT_FORMAT.WEBM_24KHZ_16BIT_MONO_OPUS
+  );
+  const filePath = await tts.toFile("./example_audio.webm", "Hi, how are you?");
 })();
@@ -85,0 +107,0 @@ ```

No alert changes

Improved metrics

Total package byte prevSize

34630

increased by2.7%

Number of package files

17

increased by13.33%

Lines of code

624

increased by6.67%

Number of lines in readme file

110

increased by25%

Worsened metrics

Dependency count

8

increased by14.29%

Dependency changes

• + Addedrandombytes@^2.1.0