🐛 fix: Fix api

README.md

@@ -89,7 +89,7 @@ This project provides some additional configuration items set with environment v
 | Environment Variable | Description                                                                                                                                   | Default                     |
 | -------------------- | --------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------- |
 | `OPENAI_API_KEY`     | This is the API key you apply on the OpenAI account page                                                                                      | `sk-xxxxxx...xxxxxx`        |
-| `OPENAI_PROXY_URL`   | If you manually configure the OpenAI interface proxy, you can use this configuration item to override the default OpenAI API request base URL | `https://api.openai.com/v1` |
+| `OPENAI_BASE_URL`    | If you manually configure the OpenAI interface proxy, you can use this configuration item to override the default OpenAI API request base URL | `https://api.openai.com/v1` |
 
 <div align="right">
 

api/openai-stt.ts

@@ -12,13 +12,13 @@ export default async (req: Request) => {
   if (req.method !== 'POST') return new Response('Method Not Allowed', { status: 405 });
 
   const OPENAI_API_KEY = process.env.OPENAI_API_KEY;
-  const OPENAI_PROXY_URL = process.env.OPENAI_PROXY_URL;
+  const OPENAI_BASE_URL = process.env.OPENAI_BASE_URL;
 
   if (!OPENAI_API_KEY) return new Response('OPENAI_API_KEY is not set', { status: 500 });
 
   const payload = (await req.json()) as OpenAISTTPayload;
 
-  const openai = new OpenAI({ apiKey: OPENAI_API_KEY, baseURL: OPENAI_PROXY_URL });
+  const openai = new OpenAI({ apiKey: OPENAI_API_KEY, baseURL: OPENAI_BASE_URL });
   const res = await createOpenaiAudioTranscriptions({ openai, payload });
 
   return new Response(JSON.stringify(res), {

api/openai-tts.ts

@@ -11,13 +11,13 @@ export const config = {
 export default async (req: Request) => {
   if (req.method !== 'POST') return new Response('Method Not Allowed', { status: 405 });
   const OPENAI_API_KEY = process.env.OPENAI_API_KEY;
-  const OPENAI_PROXY_URL = process.env.OPENAI_PROXY_URL;
+  const OPENAI_BASE_URL = process.env.OPENAI_BASE_URL;
 
   if (!OPENAI_API_KEY) return new Response('OPENAI_API_KEY is not set', { status: 500 });
 
   const payload = (await req.json()) as OpenAITTSPayload;
 
-  const openai = new OpenAI({ apiKey: OPENAI_API_KEY, baseURL: OPENAI_PROXY_URL });
+  const openai = new OpenAI({ apiKey: OPENAI_API_KEY, baseURL: OPENAI_BASE_URL });
 
   return createOpenaiAudioSpeech({ openai, payload });
 };

src/core/EdgeSpeechTTS/createEdgeSpeech.ts

@@ -1,7 +1,7 @@
 import qs from 'query-string';
 import { v4 as uuidv4 } from 'uuid';
 
-import { SsmlOptions, genSSML } from '../utils/genSSML';
+import { type SsmlOptions, genSSML } from '../utils/genSSML';
 import { genSendContent } from '../utils/genSendContent';
 import { getHeadersAndData } from '../utils/getHeadersAndData';
 

src/core/EdgeSpeechTTS/index.ts

@@ -1,30 +1,24 @@
 import edgeVoiceList from '@/core/EdgeSpeechTTS/edgeVoiceList';
 import voiceName from '@/core/data/voiceList';
 import { arrayBufferConvert } from '@/core/utils/arrayBufferConvert';
-import { SsmlOptions } from '@/core/utils/genSSML';
 import { getVoiceLocaleOptions } from '@/core/utils/getVoiceList';
 
-import { createEdgeSpeech } from './createEdgeSpeech';
+import { type EdgeSpeechPayload, createEdgeSpeech } from './createEdgeSpeech';
 import { getEdgeVoiceOptions } from './options';
 
-export interface EdgeSpeechPayload {
-  /**
-   * @title 语音合成的文本
-   */
-  input: string;
-  /**
-   * @title SSML 语音合成的配置
-   */
-  options: Pick<SsmlOptions, 'voice'>;
+export type { EdgeSpeechPayload } from './createEdgeSpeech';
+
+export interface EdgeSpeechAPI {
+  backendURL?: string;
 }
 
 export class EdgeSpeechTTS {
   private locale?: string;
-  private BASE_URL: string | undefined;
+  private BACKEND_URL: string | undefined;
 
-  constructor({ baseURL, locale }: { baseURL?: string; locale?: string } = {}) {
+  constructor({ backendURL, locale }: EdgeSpeechAPI & { locale?: string } = {}) {
     this.locale = locale;
-    this.BASE_URL = baseURL;
+    this.BACKEND_URL = backendURL;
   }
 
   get voiceOptions() {
@@ -37,8 +31,8 @@ export class EdgeSpeechTTS {
   static createRequest = createEdgeSpeech;
 
   private fetch = async (payload: EdgeSpeechPayload) => {
-    const response = await (this.BASE_URL
-      ? fetch(this.BASE_URL, { body: JSON.stringify(payload), method: 'POST' })
+    const response = await (this.BACKEND_URL
+      ? fetch(this.BACKEND_URL, { body: JSON.stringify(payload), method: 'POST' })
       : createEdgeSpeech({ payload }));
 
     if (!response.ok) {

src/core/MicrosoftSpeechTTS/createMicrosoftSpeech.ts

@@ -1,6 +1,6 @@
 import { v4 as uuidv4 } from 'uuid';
 
-import { SsmlOptions, genSSML } from '../utils/genSSML';
+import { type SsmlOptions, genSSML } from '../utils/genSSML';
 
 const MICROSOFT_SPEECH_URL =
   'https://southeastasia.api.speech.microsoft.com/accfreetrial/texttospeech/acc/v3.0-beta1/vcg/speak';

src/core/MicrosoftSpeechTTS/index.ts

@@ -1,19 +1,24 @@
+import styleList from '@/core/data/styleList';
 import voiceName from '@/core/data/voiceList';
 import { arrayBufferConvert } from '@/core/utils/arrayBufferConvert';
 import { getVoiceLocaleOptions } from '@/core/utils/getVoiceList';
 
-import { MicrosoftSpeechPayload, createMicrosoftSpeech } from './createMicrosoftSpeech';
+import { type MicrosoftSpeechPayload, createMicrosoftSpeech } from './createMicrosoftSpeech';
 import azureVoiceList, { getAzureVoiceOptions } from './voiceList';
 
 export type { MicrosoftSpeechPayload } from './createMicrosoftSpeech';
 
+export interface MicrosoftSpeechAPI {
+  backendUrl?: string;
+}
+
 export class MicrosoftSpeechTTS {
   private locale?: string;
-  private BASE_URL: string | undefined;
+  private BACKEND_URL: string | undefined;
 
-  constructor({ baseURL, locale }: { baseURL?: string; locale?: string } = {}) {
+  constructor({ backendUrl, locale }: MicrosoftSpeechAPI & { locale?: string } = {}) {
     this.locale = locale;
-    this.BASE_URL = baseURL;
+    this.BACKEND_URL = backendUrl;
   }
   get voiceOptions() {
     return getAzureVoiceOptions(this.locale);
@@ -24,10 +29,11 @@ export class MicrosoftSpeechTTS {
 
   static voiceList = azureVoiceList;
   static voiceName = voiceName;
+  static styleList = styleList;
 
   private fetch = async (payload: MicrosoftSpeechPayload) => {
-    const response = await (this.BASE_URL
-      ? fetch(this.BASE_URL, { body: JSON.stringify(payload), method: 'POST' })
+    const response = await (this.BACKEND_URL
+      ? fetch(this.BACKEND_URL, { body: JSON.stringify(payload), method: 'POST' })
       : createMicrosoftSpeech({ payload }));
 
     if (!response.ok) {

src/core/OpenAISTT/index.ts

@@ -20,6 +20,12 @@ export interface OpenAISTTPayload {
   speech: Blob;
 }
 
+export interface OpenAISTTAPI {
+  apiKey?: string;
+  backendUrl?: string;
+  baseUrl?: string;
+}
+
 const genSTTBody = ({ speech, options }: OpenAISTTPayload) => {
   const mineType = options?.mineType || getRecordMineType();
   const filename = `${Date.now()}.${mineType.extension}`;
@@ -35,26 +41,32 @@ const genSTTBody = ({ speech, options }: OpenAISTTPayload) => {
 };
 
 export class OpenaiSTT {
-  private BASE_URL: string;
+  private OPENAI_BASE_URL: string;
   private OPENAI_API_KEY: string | undefined;
+  private BACKEND_URL: string | undefined;
 
-  constructor({ baseUrl, apiKey }: { apiKey?: string; baseUrl?: string } = {}) {
-    this.BASE_URL = baseUrl || OPENAI_BASE_URL;
+  constructor({ baseUrl, apiKey, backendUrl }: OpenAISTTAPI = {}) {
+    this.OPENAI_BASE_URL = baseUrl || OPENAI_BASE_URL;
     this.OPENAI_API_KEY = apiKey;
+    this.BACKEND_URL = backendUrl;
   }
 
   static safeRecordMineType = getRecordMineType;
 
+  fetch = async (payload: OpenAISTTPayload) => {
+    const url = urlJoin(this.OPENAI_BASE_URL, 'audio/speech');
+    return this.BACKEND_URL
+      ? fetch(this.BACKEND_URL, { body: JSON.stringify(payload), method: 'POST' })
+      : fetch(url, {
+          body: genSTTBody(payload),
+          headers: new Headers({
+            Authorization: `Bearer ${this.OPENAI_API_KEY}`,
+          }),
+          method: 'POST',
+        });
+  };
   create = async (payload: OpenAISTTPayload): Promise<string> => {
-    const url = urlJoin(this.BASE_URL, 'audio/speech');
-
-    const response = await fetch(url, {
-      body: genSTTBody(payload),
-      headers: new Headers({
-        Authorization: `Bearer ${this.OPENAI_API_KEY}`,
-      }),
-      method: 'POST',
-    });
+    const response = await this.fetch(payload);
 
     if (!response.ok) {
       throw new Error('Network response was not ok');

src/core/OpenAITTS/index.ts

@@ -2,9 +2,8 @@ import urlJoin from 'url-join';
 
 import { OPENAI_BASE_URL } from '@/core/const/api';
 import { arrayBufferConvert } from '@/core/utils/arrayBufferConvert';
-import { getVoiceLocaleOptions } from '@/core/utils/getVoiceList';
 
-import openaiVoiceList, { getOpenaiVoiceOptions } from './voiceList';
+import voiceList, { getOpenaiVoiceOptions } from './voiceList';
 
 export type OpenaiVoice = 'alloy' | 'echo' | 'fable' | 'onyx' | 'nova' | 'shimmer';
 
@@ -25,33 +24,49 @@ export interface OpenAITTSPayload {
   };
 }
 
+export interface OpenAITTSAPI {
+  apiKey?: string;
+  backendUrl?: string;
+  baseUrl?: string;
+}
+
 export class OpenAITTS {
-  static voiceList = openaiVoiceList;
-  private BASE_URL: string;
+  private OPENAI_BASE_URL: string;
   private OPENAI_API_KEY: string | undefined;
+  private BACKEND_URL: string | undefined;
 
-  constructor({ baseUrl, apiKey }: { apiKey?: string; baseUrl?: string } = {}) {
-    this.BASE_URL = baseUrl || OPENAI_BASE_URL;
+  constructor({ baseUrl, apiKey, backendUrl }: OpenAITTSAPI = {}) {
+    this.OPENAI_BASE_URL = baseUrl || OPENAI_BASE_URL;
     this.OPENAI_API_KEY = apiKey;
+    this.BACKEND_URL = backendUrl;
   }
 
   get voiceOptions() {
     return getOpenaiVoiceOptions();
   }
 
-  static localeOptions = getVoiceLocaleOptions();
+  static voiceList = voiceList;
 
-  create = async ({ input, options }: OpenAITTSPayload): Promise<AudioBuffer> => {
-    const url = urlJoin(this.BASE_URL, 'audio/speech');
+  fetch = async (payload: OpenAITTSPayload) => {
+    const url = urlJoin(this.OPENAI_BASE_URL, 'audio/speech');
+    return this.BACKEND_URL
+      ? fetch(this.BACKEND_URL, { body: JSON.stringify(payload), method: 'POST' })
+      : fetch(url, {
+          body: JSON.stringify({
+            input: payload.input,
+            model: payload.options?.model || 'tts-1',
+            voice: payload.options.voice,
+          }),
+          headers: new Headers({
+            'Authorization': `Bearer ${this.OPENAI_API_KEY}`,
+            'Content-Type': 'application/json',
+          }),
+          method: 'POST',
+        });
+  };
 
-    const response = await fetch(url, {
-      body: JSON.stringify({ input, model: options?.model || 'tts-1', voice: options.voice }),
-      headers: new Headers({
-        'Authorization': `Bearer ${this.OPENAI_API_KEY}`,
-        'Content-Type': 'application/json',
-      }),
-      method: 'POST',
-    });
+  create = async (payload: OpenAITTSPayload): Promise<AudioBuffer> => {
+    const response = await this.fetch(payload);
 
     if (!response.ok) {
       throw new Error('Network response was not ok');

src/core/VoiceList.ts

@@ -9,31 +9,22 @@ export class VoiceList {
   constructor(locale?: string) {
     this.locale = locale;
   }
-
   get speechSynthesVoiceOptions() {
     return getSpeechSynthesisVoiceOptions(this.locale);
   }
 
-  static allSpeechSynthesVoiceOptions = getSpeechSynthesisVoiceOptions();
-
   get azureVoiceOptions() {
     return getAzureVoiceOptions(this.locale);
   }
 
-  static allAzureVoiceOptions = getAzureVoiceOptions();
-
   get edgeVoiceOptions() {
     return getEdgeVoiceOptions(this.locale);
   }
 
-  static allEdgeVoiceOptions = getEdgeVoiceOptions();
-
   get microsoftVoiceOptions() {
     return getEdgeVoiceOptions(this.locale);
   }
 
-  static allMicrosoftVoiceOptions = getEdgeVoiceOptions();
-
   static openaiVoiceOptions = getOpenaiVoiceOptions();
 
   static localeOptions = getVoiceLocaleOptions();

src/core/const/api.ts

@@ -1 +1,5 @@
 export const OPENAI_BASE_URL = 'https://api.openai.com/v1';
+export const OPENAI_TTS_API = '/api/openai-tts';
+export const OPENAI_STT_API = '/api/openai-stt';
+export const EDGE_SPEECH_API = '/api/edge-speech';
+export const MICROSOFT_SPEECH_API = '/api/microsoft-speech';

src/core/data/styleList.ts

@@ -0,0 +1,13 @@
+export default [
+  'affectionate',
+  'angry',
+  'calm',
+  'cheerful',
+  'disgruntled',
+  'embarrassed',
+  'fearful',
+  'general',
+  'gentle',
+  'sad',
+  'serious',
+] as const;

src/core/index.ts

@@ -1,8 +1,16 @@
-export { type EdgeSpeechPayload, EdgeSpeechTTS } from '@/core/EdgeSpeechTTS';
-export { type MicrosoftSpeechPayload, MicrosoftSpeechTTS } from '@/core/MicrosoftSpeechTTS';
-export { OpenaiSTT, type OpenAISTTPayload } from '@/core/OpenAISTT';
-export type { OpenAITTSPayload, OpenaiVoice } from '@/core/OpenAITTS';
-export { OpenAITTS } from '@/core/OpenAITTS';
+export { type EdgeSpeechAPI, type EdgeSpeechPayload, EdgeSpeechTTS } from '@/core/EdgeSpeechTTS';
+export {
+  type MicrosoftSpeechAPI,
+  type MicrosoftSpeechPayload,
+  MicrosoftSpeechTTS,
+} from '@/core/MicrosoftSpeechTTS';
+export { OpenaiSTT, type OpenAISTTAPI, type OpenAISTTPayload } from '@/core/OpenAISTT';
+export {
+  OpenAITTS,
+  type OpenAITTSAPI,
+  type OpenAITTSPayload,
+  type OpenaiVoice,
+} from '@/core/OpenAITTS';
 export { SpeechSynthesisTTS } from '@/core/SpeechSynthesisTTS';
-export { type RecordMineType } from '@/core/utils/getRecordMineType';
+export { getRecordMineType, type RecordMineType } from '@/core/utils/getRecordMineType';
 export { VoiceList } from '@/core/VoiceList';

src/react/_util/api.ts

@@ -1,2 +1,2 @@
-export const MICROSOFT_SPEECH_API_URL = '/api/microsoft-speech';
-export const EDGE_SPEECH_API_URL = '/api/edge-speech';
+export const MICROSOFT_SPEECH_BACKEND_URL = '/api/microsoft-speech';
+export const EDGE_SPEECH_BACKEND_URL = '/api/edge-speech';

src/react/useEdgeSpeech/demos/index.tsx

@@ -5,7 +5,7 @@ import { Button, Input } from 'antd';
 import { Volume2 } from 'lucide-react';
 import { Flexbox } from 'react-layout-kit';
 
-import { EDGE_SPEECH_API_URL } from '../../_util/api';
+import { EDGE_SPEECH_BACKEND_URL } from '../../_util/api';
 import { genLevaOptions } from '../../_util/leva';
 
 const defaultText = '这是一段使用 Edge Speech 的语音演示';
@@ -15,7 +15,10 @@ export default () => {
 
   const api: any = useControls(
     {
-      url: EDGE_SPEECH_API_URL,
+      backendUrl: {
+        label: 'EDGE_SPEECH_BACKEND_URL',
+        value: EDGE_SPEECH_BACKEND_URL,
+      },
     },
     { store },
   );