Northern_Lights/cherry-studio
1
0
Fork 0
mirror of https://github.com/CherryHQ/cherry-studio.git synced 2025-12-26 03:31:24 +08:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

修复部分问题

Browse Source
This commit is contained in:
1600822305 2025-04-11 03:50:12 +08:00
parent ff95670f25
commit 1f967765e4
26 changed files with 1321 additions and 962 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/main/index.ts
View File

@ -6,8 +6,8 @@ import installExtension, { REACT_DEVELOPER_TOOLS, REDUX_DEVTOOLS } from 'electro
import { registerIpc } from './ipc'
import { configManager } from './services/ConfigManager'
import { CHERRY_STUDIO_PROTOCOL, handleProtocolUrl, registerProtocolClient } from './services/ProtocolClient'
import { registerMsTTSIpcHandlers } from './services/MsTTSIpcHandler'
import { CHERRY_STUDIO_PROTOCOL, handleProtocolUrl, registerProtocolClient } from './services/ProtocolClient'
import { registerShortcuts } from './services/ShortcutService'
import { TrayService } from './services/TrayService'
import { windowService } from './services/WindowService'

8
src/main/ipc.ts
View File

@ -19,11 +19,11 @@ import FileStorage from './services/FileStorage'
import { GeminiService } from './services/GeminiService'
import KnowledgeService from './services/KnowledgeService'
import mcpService from './services/MCPService'
import * as MsTTSService from './services/MsTTSService'
import * as NutstoreService from './services/NutstoreService'
import ObsidianVaultService from './services/ObsidianVaultService'
import { ProxyConfig, proxyManager } from './services/ProxyManager'
import { searchService } from './services/SearchService'
import * as MsTTSService from './services/MsTTSService'
import { registerShortcuts, unregisterAllShortcuts } from './services/ShortcutService'
import { TrayService } from './services/TrayService'
import { windowService } from './services/WindowService'
@ -307,9 +307,7 @@ export function registerIpc(mainWindow: BrowserWindow, app: Electron.App) {
// 注册MsTTS IPC处理程序
ipcMain.handle(IpcChannel.MsTTS_GetVoices, MsTTSService.getVoices)
ipcMain.handle(
IpcChannel.MsTTS_Synthesize,
(_, text: string, voice: string, outputFormat: string) =>
MsTTSService.synthesize(text, voice, outputFormat)
ipcMain.handle(IpcChannel.MsTTS_Synthesize, (_, text: string, voice: string, outputFormat: string) =>
MsTTSService.synthesize(text, voice, outputFormat)
)
}

11
src/main/services/ASRServerService.ts
View File

@ -27,7 +27,7 @@ class ASRServerService {
* 启动ASR服务器
* @returns Promise<{success: boolean, pid?: number, error?: string}>
*/
private async startServer(): Promise<{success: boolean, pid?: number, error?: string}> {
private async startServer(): Promise<{ success: boolean; pid?: number; error?: string }> {
try {
if (this.asrServerProcess) {
return { success: true, pid: this.asrServerProcess.pid }
@ -90,7 +90,7 @@ class ASRServerService {
})
// 等待一段时间确保服务器启动
await new Promise(resolve => setTimeout(resolve, 1000))
await new Promise((resolve) => setTimeout(resolve, 1000))
return { success: true, pid: this.asrServerProcess.pid }
} catch (error) {
@ -105,7 +105,10 @@ class ASRServerService {
* @param pid 进程ID
* @returns Promise<{success: boolean, error?: string}>
*/
private async stopServer(_event: Electron.IpcMainInvokeEvent, pid?: number): Promise<{success: boolean, error?: string}> {
private async stopServer(
_event: Electron.IpcMainInvokeEvent,
pid?: number
): Promise<{ success: boolean; error?: string }> {
try {
if (!this.asrServerProcess) {
return { success: true }
@ -120,7 +123,7 @@ class ASRServerService {
this.asrServerProcess.kill()
// 等待一段时间确保进程已经退出
await new Promise(resolve => setTimeout(resolve, 500))
await new Promise((resolve) => setTimeout(resolve, 500))
this.asrServerProcess = null
return { success: true }

81
src/main/services/MsEdgeTTSService.ts
View File

@ -1,23 +1,24 @@
import { EdgeTTS } from 'node-edge-tts';
import fs from 'node:fs';
import path from 'node:path';
import { app } from 'electron';
import log from 'electron-log';
import fs from 'node:fs'
import path from 'node:path'
import { app } from 'electron'
import log from 'electron-log'
import { EdgeTTS } from 'node-edge-tts'
/**
* Microsoft Edge TTS服务
* 使用Microsoft Edge的在线TTS服务，不需要API密钥
*/
class MsEdgeTTSService {
private static instance: MsEdgeTTSService;
private tempDir: string;
private static instance: MsEdgeTTSService
private tempDir: string
private constructor() {
this.tempDir = path.join(app.getPath('temp'), 'cherry-tts');
this.tempDir = path.join(app.getPath('temp'), 'cherry-tts')
// 确保临时目录存在
if (!fs.existsSync(this.tempDir)) {
fs.mkdirSync(this.tempDir, { recursive: true });
fs.mkdirSync(this.tempDir, { recursive: true })
}
}
@ -26,9 +27,9 @@ class MsEdgeTTSService {
*/
public static getInstance(): MsEdgeTTSService {
if (!MsEdgeTTSService.instance) {
MsEdgeTTSService.instance = new MsEdgeTTSService();
MsEdgeTTSService.instance = new MsEdgeTTSService()
}
return MsEdgeTTSService.instance;
return MsEdgeTTSService.instance
}
/**
@ -46,11 +47,11 @@ class MsEdgeTTSService {
{ name: 'zh-CN-XiaomoNeural', locale: 'zh-CN', gender: 'Female' },
{ name: 'zh-CN-XiaoxuanNeural', locale: 'zh-CN', gender: 'Female' },
{ name: 'zh-CN-XiaoruiNeural', locale: 'zh-CN', gender: 'Female' },
{ name: 'zh-CN-YunfengNeural', locale: 'zh-CN', gender: 'Male' },
];
{ name: 'zh-CN-YunfengNeural', locale: 'zh-CN', gender: 'Male' }
]
} catch (error) {
log.error('获取Microsoft Edge TTS语音列表失败:', error);
throw error;
log.error('获取Microsoft Edge TTS语音列表失败:', error)
throw error
}
}
@ -63,15 +64,15 @@ class MsEdgeTTSService {
*/
public async synthesize(text: string, voice: string, outputFormat: string): Promise<string> {
try {
log.info(`Microsoft Edge TTS合成语音: 文本="${text.substring(0, 30)}...", 语音=${voice}, 格式=${outputFormat}`);
log.info(`Microsoft Edge TTS合成语音: 文本="${text.substring(0, 30)}...", 语音=${voice}, 格式=${outputFormat}`)
// 验证输入参数
if (!text || text.trim() === '') {
throw new Error('要合成的文本不能为空');
throw new Error('要合成的文本不能为空')
}
if (!voice || voice.trim() === '') {
throw new Error('语音名称不能为空');
throw new Error('语音名称不能为空')
}
// 创建一个新的EdgeTTS实例，并设置参数
@ -79,58 +80,58 @@ class MsEdgeTTSService {
voice: voice,
outputFormat: outputFormat,
timeout: 30000, // 30秒超时
rate: '+0%', // 正常语速
rate: '+0%', // 正常语速
pitch: '+0Hz', // 正常音调
volume: '+0%' // 正常音量
});
volume: '+0%' // 正常音量
})
// 生成临时文件路径
const timestamp = Date.now();
const fileExtension = outputFormat.includes('mp3') ? 'mp3' : outputFormat.split('-').pop() || 'audio';
const outputPath = path.join(this.tempDir, `tts_${timestamp}.${fileExtension}`);
const timestamp = Date.now()
const fileExtension = outputFormat.includes('mp3') ? 'mp3' : outputFormat.split('-').pop() || 'audio'
const outputPath = path.join(this.tempDir, `tts_${timestamp}.${fileExtension}`)
log.info(`开始生成语音文件: ${outputPath}`);
log.info(`开始生成语音文件: ${outputPath}`)
// 使用ttsPromise方法生成文件
await tts.ttsPromise(text, outputPath);
await tts.ttsPromise(text, outputPath)
// 验证生成的文件是否存在且大小大于0
if (!fs.existsSync(outputPath)) {
throw new Error(`生成的语音文件不存在: ${outputPath}`);
throw new Error(`生成的语音文件不存在: ${outputPath}`)
}
const stats = fs.statSync(outputPath);
const stats = fs.statSync(outputPath)
if (stats.size === 0) {
throw new Error(`生成的语音文件大小为0: ${outputPath}`);
throw new Error(`生成的语音文件大小为0: ${outputPath}`)
}
log.info(`Microsoft Edge TTS合成成功: ${outputPath}, 文件大小: ${stats.size} 字节`);
return outputPath;
log.info(`Microsoft Edge TTS合成成功: ${outputPath}, 文件大小: ${stats.size} 字节`)
return outputPath
} catch (error: any) {
// 记录详细的错误信息
log.error(`Microsoft Edge TTS语音合成失败 (语音=${voice}):`, error);
log.error(`Microsoft Edge TTS语音合成失败 (语音=${voice}):`, error)
// 尝试提供更有用的错误信息
if (error.message && typeof error.message === 'string') {
if (error.message.includes('Timed out')) {
throw new Error(`语音合成超时，请检查网络连接或尝试其他语音`);
throw new Error(`语音合成超时，请检查网络连接或尝试其他语音`)
} else if (error.message.includes('ENOTFOUND')) {
throw new Error(`无法连接到Microsoft语音服务，请检查网络连接`);
throw new Error(`无法连接到Microsoft语音服务，请检查网络连接`)
} else if (error.message.includes('ECONNREFUSED')) {
throw new Error(`连接被拒绝，请检查网络设置或代理配置`);
throw new Error(`连接被拒绝，请检查网络设置或代理配置`)
}
}
throw error;
throw error
}
}
}
// 导出单例方法
export const getVoices = async () => {
return await MsEdgeTTSService.getInstance().getVoices();
};
return await MsEdgeTTSService.getInstance().getVoices()
}
export const synthesize = async (text: string, voice: string, outputFormat: string) => {
return await MsEdgeTTSService.getInstance().synthesize(text, voice, outputFormat);
};
return await MsEdgeTTSService.getInstance().synthesize(text, voice, outputFormat)
}

19
src/main/services/MsTTSIpcHandler.ts
View File

@ -1,18 +1,17 @@
import { IpcChannel } from '@shared/IpcChannel';
import { ipcMain } from 'electron';
import * as MsTTSService from './MsTTSService';
import { IpcChannel } from '@shared/IpcChannel'
import { ipcMain } from 'electron'
import * as MsTTSService from './MsTTSService'
/**
* 注册MsTTS相关的IPC处理程序
*/
export function registerMsTTSIpcHandlers(): void {
// 获取可用的语音列表
ipcMain.handle(IpcChannel.MsTTS_GetVoices, MsTTSService.getVoices);
ipcMain.handle(IpcChannel.MsTTS_GetVoices, MsTTSService.getVoices)
// 合成语音
ipcMain.handle(
IpcChannel.MsTTS_Synthesize,
(_, text: string, voice: string, outputFormat: string) =>
MsTTSService.synthesize(text, voice, outputFormat)
);
ipcMain.handle(IpcChannel.MsTTS_Synthesize, (_, text: string, voice: string, outputFormat: string) =>
MsTTSService.synthesize(text, voice, outputFormat)
)
}

561
src/main/services/MsTTSService.ts
View File

@ -1,115 +1,440 @@
import { EdgeTTS } from 'node-edge-tts'; // listVoices is no longer needed here
import fs from 'node:fs';
import path from 'node:path';
import { app } from 'electron';
import log from 'electron-log';
import fs from 'node:fs'
import path from 'node:path'
import { app } from 'electron'
import log from 'electron-log'
import { EdgeTTS } from 'node-edge-tts' // listVoices is no longer needed here
// --- START OF HARDCODED VOICE LIST ---
// WARNING: This list is static and may become outdated.
// It's generally recommended to use listVoices() for the most up-to-date list.
const hardcodedVoices = [
{ Name: 'Microsoft Server Speech Text to Speech Voice (af-ZA, AdriNeural)', ShortName: 'af-ZA-AdriNeural', Gender: 'Female', Locale: 'af-ZA' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (am-ET, MekdesNeural)', ShortName: 'am-ET-MekdesNeural', Gender: 'Female', Locale: 'am-ET' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (ar-AE, FatimaNeural)', ShortName: 'ar-AE-FatimaNeural', Gender: 'Female', Locale: 'ar-AE' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (ar-AE, HamdanNeural)', ShortName: 'ar-AE-HamdanNeural', Gender: 'Male', Locale: 'ar-AE' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (ar-BH, AliNeural)', ShortName: 'ar-BH-AliNeural', Gender: 'Male', Locale: 'ar-BH' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (ar-BH, LailaNeural)', ShortName: 'ar-BH-LailaNeural', Gender: 'Female', Locale: 'ar-BH' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (af-ZA, AdriNeural)',
ShortName: 'af-ZA-AdriNeural',
Gender: 'Female',
Locale: 'af-ZA'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (am-ET, MekdesNeural)',
ShortName: 'am-ET-MekdesNeural',
Gender: 'Female',
Locale: 'am-ET'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (ar-AE, FatimaNeural)',
ShortName: 'ar-AE-FatimaNeural',
Gender: 'Female',
Locale: 'ar-AE'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (ar-AE, HamdanNeural)',
ShortName: 'ar-AE-HamdanNeural',
Gender: 'Male',
Locale: 'ar-AE'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (ar-BH, AliNeural)',
ShortName: 'ar-BH-AliNeural',
Gender: 'Male',
Locale: 'ar-BH'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (ar-BH, LailaNeural)',
ShortName: 'ar-BH-LailaNeural',
Gender: 'Female',
Locale: 'ar-BH'
},
// ... (Many other Arabic locales/voices) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (ar-SA, ZariyahNeural)', ShortName: 'ar-SA-ZariyahNeural', Gender: 'Female', Locale: 'ar-SA' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (az-AZ, BabekNeural)', ShortName: 'az-AZ-BabekNeural', Gender: 'Male', Locale: 'az-AZ' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (az-AZ, BanuNeural)', ShortName: 'az-AZ-BanuNeural', Gender: 'Female', Locale: 'az-AZ' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (bg-BG, BorislavNeural)', ShortName: 'bg-BG-BorislavNeural', Gender: 'Male', Locale: 'bg-BG' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (bg-BG, KalinaNeural)', ShortName: 'bg-BG-KalinaNeural', Gender: 'Female', Locale: 'bg-BG' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (bn-BD, NabanitaNeural)', ShortName: 'bn-BD-NabanitaNeural', Gender: 'Female', Locale: 'bn-BD' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (bn-BD, PradeepNeural)', ShortName: 'bn-BD-PradeepNeural', Gender: 'Male', Locale: 'bn-BD' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (ar-SA, ZariyahNeural)',
ShortName: 'ar-SA-ZariyahNeural',
Gender: 'Female',
Locale: 'ar-SA'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (az-AZ, BabekNeural)',
ShortName: 'az-AZ-BabekNeural',
Gender: 'Male',
Locale: 'az-AZ'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (az-AZ, BanuNeural)',
ShortName: 'az-AZ-BanuNeural',
Gender: 'Female',
Locale: 'az-AZ'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (bg-BG, BorislavNeural)',
ShortName: 'bg-BG-BorislavNeural',
Gender: 'Male',
Locale: 'bg-BG'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (bg-BG, KalinaNeural)',
ShortName: 'bg-BG-KalinaNeural',
Gender: 'Female',
Locale: 'bg-BG'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (bn-BD, NabanitaNeural)',
ShortName: 'bn-BD-NabanitaNeural',
Gender: 'Female',
Locale: 'bn-BD'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (bn-BD, PradeepNeural)',
ShortName: 'bn-BD-PradeepNeural',
Gender: 'Male',
Locale: 'bn-BD'
},
// ... (Catalan, Czech, Welsh, Danish, German, Greek, English variants) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-AU, NatashaNeural)', ShortName: 'en-AU-NatashaNeural', Gender: 'Female', Locale: 'en-AU' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-AU, WilliamNeural)', ShortName: 'en-AU-WilliamNeural', Gender: 'Male', Locale: 'en-AU' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-CA, ClaraNeural)', ShortName: 'en-CA-ClaraNeural', Gender: 'Female', Locale: 'en-CA' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-CA, LiamNeural)', ShortName: 'en-CA-LiamNeural', Gender: 'Male', Locale: 'en-CA' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, LibbyNeural)', ShortName: 'en-GB-LibbyNeural', Gender: 'Female', Locale: 'en-GB' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, MaisieNeural)', ShortName: 'en-GB-MaisieNeural', Gender: 'Female', Locale: 'en-GB' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, RyanNeural)', ShortName: 'en-GB-RyanNeural', Gender: 'Male', Locale: 'en-GB' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, SoniaNeural)', ShortName: 'en-GB-SoniaNeural', Gender: 'Female', Locale: 'en-GB' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, ThomasNeural)', ShortName: 'en-GB-ThomasNeural', Gender: 'Male', Locale: 'en-GB' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-HK, SamNeural)', ShortName: 'en-HK-SamNeural', Gender: 'Male', Locale: 'en-HK' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-HK, YanNeural)', ShortName: 'en-HK-YanNeural', Gender: 'Female', Locale: 'en-HK' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-IE, ConnorNeural)', ShortName: 'en-IE-ConnorNeural', Gender: 'Male', Locale: 'en-IE' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-IE, EmilyNeural)', ShortName: 'en-IE-EmilyNeural', Gender: 'Female', Locale: 'en-IE' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-IN, NeerjaNeural)', ShortName: 'en-IN-NeerjaNeural', Gender: 'Female', Locale: 'en-IN' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-IN, PrabhatNeural)', ShortName: 'en-IN-PrabhatNeural', Gender: 'Male', Locale: 'en-IN' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-AU, NatashaNeural)',
ShortName: 'en-AU-NatashaNeural',
Gender: 'Female',
Locale: 'en-AU'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-AU, WilliamNeural)',
ShortName: 'en-AU-WilliamNeural',
Gender: 'Male',
Locale: 'en-AU'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-CA, ClaraNeural)',
ShortName: 'en-CA-ClaraNeural',
Gender: 'Female',
Locale: 'en-CA'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-CA, LiamNeural)',
ShortName: 'en-CA-LiamNeural',
Gender: 'Male',
Locale: 'en-CA'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, LibbyNeural)',
ShortName: 'en-GB-LibbyNeural',
Gender: 'Female',
Locale: 'en-GB'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, MaisieNeural)',
ShortName: 'en-GB-MaisieNeural',
Gender: 'Female',
Locale: 'en-GB'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, RyanNeural)',
ShortName: 'en-GB-RyanNeural',
Gender: 'Male',
Locale: 'en-GB'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, SoniaNeural)',
ShortName: 'en-GB-SoniaNeural',
Gender: 'Female',
Locale: 'en-GB'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-GB, ThomasNeural)',
ShortName: 'en-GB-ThomasNeural',
Gender: 'Male',
Locale: 'en-GB'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-HK, SamNeural)',
ShortName: 'en-HK-SamNeural',
Gender: 'Male',
Locale: 'en-HK'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-HK, YanNeural)',
ShortName: 'en-HK-YanNeural',
Gender: 'Female',
Locale: 'en-HK'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-IE, ConnorNeural)',
ShortName: 'en-IE-ConnorNeural',
Gender: 'Male',
Locale: 'en-IE'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-IE, EmilyNeural)',
ShortName: 'en-IE-EmilyNeural',
Gender: 'Female',
Locale: 'en-IE'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-IN, NeerjaNeural)',
ShortName: 'en-IN-NeerjaNeural',
Gender: 'Female',
Locale: 'en-IN'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-IN, PrabhatNeural)',
ShortName: 'en-IN-PrabhatNeural',
Gender: 'Male',
Locale: 'en-IN'
},
// ... (Many more English variants: KE, NG, NZ, PH, SG, TZ, US, ZA) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, AriaNeural)', ShortName: 'en-US-AriaNeural', Gender: 'Female', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, AnaNeural)', ShortName: 'en-US-AnaNeural', Gender: 'Female', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, ChristopherNeural)', ShortName: 'en-US-ChristopherNeural', Gender: 'Male', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, EricNeural)', ShortName: 'en-US-EricNeural', Gender: 'Male', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, GuyNeural)', ShortName: 'en-US-GuyNeural', Gender: 'Male', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, JennyNeural)', ShortName: 'en-US-JennyNeural', Gender: 'Female', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, MichelleNeural)', ShortName: 'en-US-MichelleNeural', Gender: 'Female', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, RogerNeural)', ShortName: 'en-US-RogerNeural', Gender: 'Male', Locale: 'en-US' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (en-US, SteffanNeural)', ShortName: 'en-US-SteffanNeural', Gender: 'Male', Locale: 'en-US' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, AriaNeural)',
ShortName: 'en-US-AriaNeural',
Gender: 'Female',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, AnaNeural)',
ShortName: 'en-US-AnaNeural',
Gender: 'Female',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, ChristopherNeural)',
ShortName: 'en-US-ChristopherNeural',
Gender: 'Male',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, EricNeural)',
ShortName: 'en-US-EricNeural',
Gender: 'Male',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, GuyNeural)',
ShortName: 'en-US-GuyNeural',
Gender: 'Male',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, JennyNeural)',
ShortName: 'en-US-JennyNeural',
Gender: 'Female',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, MichelleNeural)',
ShortName: 'en-US-MichelleNeural',
Gender: 'Female',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, RogerNeural)',
ShortName: 'en-US-RogerNeural',
Gender: 'Male',
Locale: 'en-US'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (en-US, SteffanNeural)',
ShortName: 'en-US-SteffanNeural',
Gender: 'Male',
Locale: 'en-US'
},
// ... (Spanish variants) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (es-MX, DaliaNeural)', ShortName: 'es-MX-DaliaNeural', Gender: 'Female', Locale: 'es-MX' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (es-MX, JorgeNeural)', ShortName: 'es-MX-JorgeNeural', Gender: 'Male', Locale: 'es-MX' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (es-MX, DaliaNeural)',
ShortName: 'es-MX-DaliaNeural',
Gender: 'Female',
Locale: 'es-MX'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (es-MX, JorgeNeural)',
ShortName: 'es-MX-JorgeNeural',
Gender: 'Male',
Locale: 'es-MX'
},
// ... (Estonian, Basque, Persian, Finnish, Filipino, French, Irish, Galician, Gujarati, Hebrew, Hindi, Croatian, Hungarian, Indonesian, Icelandic, Italian, Japanese) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (ja-JP, KeitaNeural)', ShortName: 'ja-JP-KeitaNeural', Gender: 'Male', Locale: 'ja-JP' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (ja-JP, NanamiNeural)', ShortName: 'ja-JP-NanamiNeural', Gender: 'Female', Locale: 'ja-JP' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (ja-JP, KeitaNeural)',
ShortName: 'ja-JP-KeitaNeural',
Gender: 'Male',
Locale: 'ja-JP'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (ja-JP, NanamiNeural)',
ShortName: 'ja-JP-NanamiNeural',
Gender: 'Female',
Locale: 'ja-JP'
},
// ... (Javanese, Georgian, Kazakh, Khmer, Kannada, Korean) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (ko-KR, InJoonNeural)', ShortName: 'ko-KR-InJoonNeural', Gender: 'Male', Locale: 'ko-KR' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (ko-KR, SunHiNeural)', ShortName: 'ko-KR-SunHiNeural', Gender: 'Female', Locale: 'ko-KR' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (ko-KR, InJoonNeural)',
ShortName: 'ko-KR-InJoonNeural',
Gender: 'Male',
Locale: 'ko-KR'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (ko-KR, SunHiNeural)',
ShortName: 'ko-KR-SunHiNeural',
Gender: 'Female',
Locale: 'ko-KR'
},
// ... (Lao, Lithuanian, Latvian, Macedonian, Malayalam, Mongolian, Marathi, Malay, Maltese, Burmese, Norwegian, Dutch, Polish, Pashto, Portuguese) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (pt-BR, AntonioNeural)', ShortName: 'pt-BR-AntonioNeural', Gender: 'Male', Locale: 'pt-BR' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (pt-BR, FranciscaNeural)', ShortName: 'pt-BR-FranciscaNeural', Gender: 'Female', Locale: 'pt-BR' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (pt-PT, DuarteNeural)', ShortName: 'pt-PT-DuarteNeural', Gender: 'Male', Locale: 'pt-PT' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (pt-PT, RaquelNeural)', ShortName: 'pt-PT-RaquelNeural', Gender: 'Female', Locale: 'pt-PT' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (pt-BR, AntonioNeural)',
ShortName: 'pt-BR-AntonioNeural',
Gender: 'Male',
Locale: 'pt-BR'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (pt-BR, FranciscaNeural)',
ShortName: 'pt-BR-FranciscaNeural',
Gender: 'Female',
Locale: 'pt-BR'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (pt-PT, DuarteNeural)',
ShortName: 'pt-PT-DuarteNeural',
Gender: 'Male',
Locale: 'pt-PT'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (pt-PT, RaquelNeural)',
ShortName: 'pt-PT-RaquelNeural',
Gender: 'Female',
Locale: 'pt-PT'
},
// ... (Romanian, Russian, Sinhala, Slovak, Slovenian, Somali, Albanian, Serbian, Sundanese, Swedish, Swahili, Tamil, Telugu, Thai) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (th-TH, NiwatNeural)', ShortName: 'th-TH-NiwatNeural', Gender: 'Male', Locale: 'th-TH' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (th-TH, PremwadeeNeural)', ShortName: 'th-TH-PremwadeeNeural', Gender: 'Female', Locale: 'th-TH' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (th-TH, NiwatNeural)',
ShortName: 'th-TH-NiwatNeural',
Gender: 'Male',
Locale: 'th-TH'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (th-TH, PremwadeeNeural)',
ShortName: 'th-TH-PremwadeeNeural',
Gender: 'Female',
Locale: 'th-TH'
},
// ... (Turkish, Ukrainian, Urdu, Uzbek, Vietnamese) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (vi-VN, HoaiMyNeural)', ShortName: 'vi-VN-HoaiMyNeural', Gender: 'Female', Locale: 'vi-VN' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (vi-VN, NamMinhNeural)', ShortName: 'vi-VN-NamMinhNeural', Gender: 'Male', Locale: 'vi-VN' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (vi-VN, HoaiMyNeural)',
ShortName: 'vi-VN-HoaiMyNeural',
Gender: 'Female',
Locale: 'vi-VN'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (vi-VN, NamMinhNeural)',
ShortName: 'vi-VN-NamMinhNeural',
Gender: 'Male',
Locale: 'vi-VN'
},
// ... (Chinese variants) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, XiaoxiaoNeural)', ShortName: 'zh-CN-XiaoxiaoNeural', Gender: 'Female', Locale: 'zh-CN' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunxiNeural)', ShortName: 'zh-CN-YunxiNeural', Gender: 'Male', Locale: 'zh-CN' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunjianNeural)', ShortName: 'zh-CN-YunjianNeural', Gender: 'Male', Locale: 'zh-CN' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunxiaNeural)', ShortName: 'zh-CN-YunxiaNeural', Gender: 'Male', Locale: 'zh-CN' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunyangNeural)', ShortName: 'zh-CN-YunyangNeural', Gender: 'Male', Locale: 'zh-CN' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN-liaoning, XiaobeiNeural)', ShortName: 'zh-CN-liaoning-XiaobeiNeural', Gender: 'Female', Locale: 'zh-CN-liaoning' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, XiaoxiaoNeural)',
ShortName: 'zh-CN-XiaoxiaoNeural',
Gender: 'Female',
Locale: 'zh-CN'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunxiNeural)',
ShortName: 'zh-CN-YunxiNeural',
Gender: 'Male',
Locale: 'zh-CN'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunjianNeural)',
ShortName: 'zh-CN-YunjianNeural',
Gender: 'Male',
Locale: 'zh-CN'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunxiaNeural)',
ShortName: 'zh-CN-YunxiaNeural',
Gender: 'Male',
Locale: 'zh-CN'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN, YunyangNeural)',
ShortName: 'zh-CN-YunyangNeural',
Gender: 'Male',
Locale: 'zh-CN'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN-liaoning, XiaobeiNeural)',
ShortName: 'zh-CN-liaoning-XiaobeiNeural',
Gender: 'Female',
Locale: 'zh-CN-liaoning'
},
// { Name: 'Microsoft Server Speech Text to Speech Voice (zh-CN-shaanxi, XiaoniNeural)', ShortName: 'zh-CN-shaanxi-XiaoniNeural', Gender: 'Female', Locale: 'zh-CN-shaanxi' }, // Example regional voice
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-HK, HiuGaaiNeural)', ShortName: 'zh-HK-HiuGaaiNeural', Gender: 'Female', Locale: 'zh-HK' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-HK, HiuMaanNeural)', ShortName: 'zh-HK-HiuMaanNeural', Gender: 'Female', Locale: 'zh-HK' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-HK, WanLungNeural)', ShortName: 'zh-HK-WanLungNeural', Gender: 'Male', Locale: 'zh-HK' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-TW, HsiaoChenNeural)', ShortName: 'zh-TW-HsiaoChenNeural', Gender: 'Female', Locale: 'zh-TW' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-TW, HsiaoYuNeural)', ShortName: 'zh-TW-HsiaoYuNeural', Gender: 'Female', Locale: 'zh-TW' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zh-TW, YunJheNeural)', ShortName: 'zh-TW-YunJheNeural', Gender: 'Male', Locale: 'zh-TW' },
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-HK, HiuGaaiNeural)',
ShortName: 'zh-HK-HiuGaaiNeural',
Gender: 'Female',
Locale: 'zh-HK'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-HK, HiuMaanNeural)',
ShortName: 'zh-HK-HiuMaanNeural',
Gender: 'Female',
Locale: 'zh-HK'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-HK, WanLungNeural)',
ShortName: 'zh-HK-WanLungNeural',
Gender: 'Male',
Locale: 'zh-HK'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-TW, HsiaoChenNeural)',
ShortName: 'zh-TW-HsiaoChenNeural',
Gender: 'Female',
Locale: 'zh-TW'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-TW, HsiaoYuNeural)',
ShortName: 'zh-TW-HsiaoYuNeural',
Gender: 'Female',
Locale: 'zh-TW'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zh-TW, YunJheNeural)',
ShortName: 'zh-TW-YunJheNeural',
Gender: 'Male',
Locale: 'zh-TW'
},
// ... (Zulu) ...
{ Name: 'Microsoft Server Speech Text to Speech Voice (zu-ZA, ThandoNeural)', ShortName: 'zu-ZA-ThandoNeural', Gender: 'Female', Locale: 'zu-ZA' },
{ Name: 'Microsoft Server Speech Text to Speech Voice (zu-ZA, ThembaNeural)', ShortName: 'zu-ZA-ThembaNeural', Gender: 'Male', Locale: 'zu-ZA' },
];
{
Name: 'Microsoft Server Speech Text to Speech Voice (zu-ZA, ThandoNeural)',
ShortName: 'zu-ZA-ThandoNeural',
Gender: 'Female',
Locale: 'zu-ZA'
},
{
Name: 'Microsoft Server Speech Text to Speech Voice (zu-ZA, ThembaNeural)',
ShortName: 'zu-ZA-ThembaNeural',
Gender: 'Male',
Locale: 'zu-ZA'
}
]
// --- END OF HARDCODED VOICE LIST ---
/**
* 免费在线TTS服务
* 使用免费的在线TTS服务，不需要API密钥
*/
class MsTTSService {
private static instance: MsTTSService;
private tempDir: string;
private static instance: MsTTSService
private tempDir: string
private constructor() {
this.tempDir = path.join(app.getPath('temp'), 'cherry-tts');
this.tempDir = path.join(app.getPath('temp'), 'cherry-tts')
if (!fs.existsSync(this.tempDir)) {
fs.mkdirSync(this.tempDir, { recursive: true });
fs.mkdirSync(this.tempDir, { recursive: true })
}
log.info('初始化免费在线TTS服务 (使用硬编码语音列表)');
log.info('初始化免费在线TTS服务 (使用硬编码语音列表)')
}
public static getInstance(): MsTTSService {
if (!MsTTSService.instance) {
MsTTSService.instance = new MsTTSService();
MsTTSService.instance = new MsTTSService()
}
return MsTTSService.instance;
return MsTTSService.instance
}
/**
@ -118,15 +443,15 @@ class MsTTSService {
*/
public async getVoices(): Promise<any[]> {
try {
log.info(`返回硬编码的 ${hardcodedVoices.length} 个语音列表`);
log.info(`返回硬编码的 ${hardcodedVoices.length} 个语音列表`)
// 直接返回硬编码的列表
// 注意：保持 async 是为了接口兼容性，虽然这里没有实际的异步操作
return hardcodedVoices;
return hardcodedVoices
} catch (error) {
// 这个 try/catch 在这里意义不大了，因为返回静态数据不会出错
// 但保留结构以防未来改动
log.error('获取硬编码语音列表时出错 (理论上不应发生):', error);
return []; // 返回空列表以防万一
// 这个 try/catch 在这里意义不大了，因为返回静态数据不会出错
// 但保留结构以防未来改动
log.error('获取硬编码语音列表时出错 (理论上不应发生):', error)
return [] // 返回空列表以防万一
}
}
@ -140,15 +465,15 @@ class MsTTSService {
public async synthesize(text: string, voice: string, outputFormat: string): Promise<string> {
try {
// 记录详细的请求信息
log.info(`微软在线TTS合成语音: 文本="${text.substring(0, 30)}...", 语音=${voice}, 格式=${outputFormat}`);
log.info(`微软在线TTS合成语音: 文本="${text.substring(0, 30)}...", 语音=${voice}, 格式=${outputFormat}`)
// 验证输入参数
if (!text || text.trim() === '') {
throw new Error('要合成的文本不能为空');
throw new Error('要合成的文本不能为空')
}
if (!voice || voice.trim() === '') {
throw new Error('语音名称不能为空');
throw new Error('语音名称不能为空')
}
// 创建一个新的EdgeTTS实例，并设置参数
@ -157,80 +482,80 @@ class MsTTSService {
voice: voice,
outputFormat: outputFormat,
timeout: 30000, // 30秒超时
rate: '+0%', // 正常语速
rate: '+0%', // 正常语速
pitch: '+0Hz', // 正常音调
volume: '+0%' // 正常音量
});
volume: '+0%' // 正常音量
})
// 生成临时文件路径
const timestamp = Date.now();
const fileExtension = outputFormat.includes('mp3') ? 'mp3' : outputFormat.split('-').pop() || 'audio';
const outputPath = path.join(this.tempDir, `tts_${timestamp}.${fileExtension}`);
const timestamp = Date.now()
const fileExtension = outputFormat.includes('mp3') ? 'mp3' : outputFormat.split('-').pop() || 'audio'
const outputPath = path.join(this.tempDir, `tts_${timestamp}.${fileExtension}`)
log.info(`开始生成语音文件: ${outputPath}`);
log.info(`开始生成语音文件: ${outputPath}`)
// 使用ttsPromise方法生成文件
await tts.ttsPromise(text, outputPath);
await tts.ttsPromise(text, outputPath)
// 验证生成的文件是否存在且大小大于0
if (!fs.existsSync(outputPath)) {
throw new Error(`生成的语音文件不存在: ${outputPath}`);
throw new Error(`生成的语音文件不存在: ${outputPath}`)
}
const stats = fs.statSync(outputPath);
const stats = fs.statSync(outputPath)
if (stats.size === 0) {
throw new Error(`生成的语音文件大小为0: ${outputPath}`);
throw new Error(`生成的语音文件大小为0: ${outputPath}`)
}
log.info(`微软在线TTS合成成功: ${outputPath}, 文件大小: ${stats.size} 字节`);
return outputPath;
log.info(`微软在线TTS合成成功: ${outputPath}, 文件大小: ${stats.size} 字节`)
return outputPath
} catch (error: any) {
// 记录详细的错误信息
log.error(`微软在线TTS语音合成失败 (语音=${voice}):`, error);
log.error(`微软在线TTS语音合成失败 (语音=${voice}):`, error)
// 尝试提供更有用的错误信息
if (error.message && typeof error.message === 'string') {
if (error.message.includes('Timed out')) {
throw new Error(`语音合成超时，请检查网络连接或尝试其他语音`);
throw new Error(`语音合成超时，请检查网络连接或尝试其他语音`)
} else if (error.message.includes('ENOTFOUND')) {
throw new Error(`无法连接到微软语音服务，请检查网络连接`);
throw new Error(`无法连接到微软语音服务，请检查网络连接`)
} else if (error.message.includes('ECONNREFUSED')) {
throw new Error(`连接被拒绝，请检查网络设置或代理配置`);
throw new Error(`连接被拒绝，请检查网络设置或代理配置`)
}
}
throw error;
throw error
}
}
/**
/**
* (可选) 清理临时文件目录
*/
public async cleanupTempDir(): Promise<void> {
// (Cleanup method remains the same)
try {
const files = await fs.promises.readdir(this.tempDir);
try {
const files = await fs.promises.readdir(this.tempDir)
for (const file of files) {
if (file.startsWith('tts_')) {
await fs.promises.unlink(path.join(this.tempDir, file));
await fs.promises.unlink(path.join(this.tempDir, file))
}
}
log.info('TTS 临时文件已清理');
log.info('TTS 临时文件已清理')
} catch (error) {
log.error('清理 TTS 临时文件失败:', error);
log.error('清理 TTS 临时文件失败:', error)
}
}
}
// 导出单例方法 (保持不变)
export const getVoices = async () => {
return await MsTTSService.getInstance().getVoices();
};
return await MsTTSService.getInstance().getVoices()
}
export const synthesize = async (text: string, voice: string, outputFormat: string) => {
return await MsTTSService.getInstance().synthesize(text, voice, outputFormat);
};
return await MsTTSService.getInstance().synthesize(text, voice, outputFormat)
}
export const cleanupTtsTempFiles = async () => {
await MsTTSService.getInstance().cleanupTempDir();
};
await MsTTSService.getInstance().cleanupTempDir()
}

54
src/renderer/src/components/VoiceCallButton.tsx
View File

@ -1,35 +1,36 @@
import React, { useState } from 'react';
import { Button, Tooltip } from 'antd';
import { PhoneOutlined, LoadingOutlined } from '@ant-design/icons';
import { useTranslation } from 'react-i18next';
import VoiceCallModal from './VoiceCallModal';
import { VoiceCallService } from '../services/VoiceCallService';
import { LoadingOutlined, PhoneOutlined } from '@ant-design/icons'
import { Button, Tooltip } from 'antd'
import React, { useState } from 'react'
import { useTranslation } from 'react-i18next'
import { VoiceCallService } from '../services/VoiceCallService'
import VoiceCallModal from './VoiceCallModal'
interface Props {
disabled?: boolean;
style?: React.CSSProperties;
disabled?: boolean
style?: React.CSSProperties
}
const VoiceCallButton: React.FC<Props> = ({ disabled = false, style }) => {
const { t } = useTranslation();
const [isModalVisible, setIsModalVisible] = useState(false);
const [isLoading, setIsLoading] = useState(false);
const { t } = useTranslation()
const [isModalVisible, setIsModalVisible] = useState(false)
const [isLoading, setIsLoading] = useState(false)
const handleClick = async () => {
if (disabled || isLoading) return;
setIsLoading(true);
if (disabled || isLoading) return
setIsLoading(true)
try {
// 初始化语音服务
await VoiceCallService.initialize();
setIsModalVisible(true);
await VoiceCallService.initialize()
setIsModalVisible(true)
} catch (error) {
console.error('Failed to initialize voice call:', error);
window.message.error(t('voice_call.initialization_failed'));
console.error('Failed to initialize voice call:', error)
window.message.error(t('voice_call.initialization_failed'))
} finally {
setIsLoading(false);
setIsLoading(false)
}
};
}
return (
<>
@ -42,14 +43,9 @@ const VoiceCallButton: React.FC<Props> = ({ disabled = false, style }) => {
style={style}
/>
</Tooltip>
{isModalVisible && (
<VoiceCallModal
visible={isModalVisible}
onClose={() => setIsModalVisible(false)}
/>
)}
{isModalVisible && <VoiceCallModal visible={isModalVisible} onClose={() => setIsModalVisible(false)} />}
</>
);
};
)
}
export default VoiceCallButton;
export default VoiceCallButton

167
src/renderer/src/components/VoiceCallModal.tsx
View File

@ -1,5 +1,3 @@
import React, { useEffect, useState } from 'react';
import { Modal, Button, Space, Tooltip } from 'antd';
import {
AudioMutedOutlined,
AudioOutlined,
@ -7,125 +5,128 @@ import {
PauseCircleOutlined,
PlayCircleOutlined,
SoundOutlined
} from '@ant-design/icons';
import styled from 'styled-components';
import { useTranslation } from 'react-i18next';
import VoiceVisualizer from './VoiceVisualizer';
import { VoiceCallService } from '../services/VoiceCallService';
} from '@ant-design/icons'
import { Button, Modal, Space, Tooltip } from 'antd'
import React, { useCallback, useEffect, useState } from 'react'
import { useTranslation } from 'react-i18next'
import styled from 'styled-components'
import { VoiceCallService } from '../services/VoiceCallService'
import VoiceVisualizer from './VoiceVisualizer'
interface Props {
visible: boolean;
onClose: () => void;
visible: boolean
onClose: () => void
}
const VoiceCallModal: React.FC<Props> = ({ visible, onClose }) => {
const { t } = useTranslation();
const [isMuted, setIsMuted] = useState(false);
const [isPaused, setIsPaused] = useState(false);
const [transcript, setTranscript] = useState('');
const [response, setResponse] = useState('');
const [isListening, setIsListening] = useState(false);
const [isSpeaking, setIsSpeaking] = useState(false);
const [isRecording, setIsRecording] = useState(false);
const [isProcessing, setIsProcessing] = useState(false);
const { t } = useTranslation()
const [isMuted, setIsMuted] = useState(false)
const [isPaused, setIsPaused] = useState(false)
const [transcript, setTranscript] = useState('')
const [response, setResponse] = useState('')
const [isListening, setIsListening] = useState(false)
const [isSpeaking, setIsSpeaking] = useState(false)
const [isRecording, setIsRecording] = useState(false)
const [isProcessing, setIsProcessing] = useState(false)
const handleClose = useCallback(() => {
VoiceCallService.endCall()
onClose()
}, [onClose])
useEffect(() => {
const startVoiceCall = async () => {
try {
await VoiceCallService.startCall({
onTranscript: (text) => setTranscript(text),
onResponse: (text) => setResponse(text),
onTranscript: (text: string) => setTranscript(text),
onResponse: (text: string) => setResponse(text),
onListeningStateChange: setIsListening,
onSpeakingStateChange: setIsSpeaking,
});
onSpeakingStateChange: setIsSpeaking
})
} catch (error) {
console.error('Voice call error:', error);
window.message.error(t('voice_call.error'));
handleClose();
console.error('Voice call error:', error)
window.message.error(t('voice_call.error'))
handleClose()
}
};
}
if (visible) {
startVoiceCall();
startVoiceCall()
}
return () => {
VoiceCallService.endCall();
};
}, [visible, t]);
const handleClose = () => {
VoiceCallService.endCall();
onClose();
};
VoiceCallService.endCall()
}
}, [visible, t, handleClose])
const toggleMute = () => {
const newMuteState = !isMuted;
setIsMuted(newMuteState);
VoiceCallService.setMuted(newMuteState);
};
const newMuteState = !isMuted
setIsMuted(newMuteState)
VoiceCallService.setMuted(newMuteState)
}
const togglePause = () => {
const newPauseState = !isPaused;
setIsPaused(newPauseState);
VoiceCallService.setPaused(newPauseState);
};
const newPauseState = !isPaused
setIsPaused(newPauseState)
VoiceCallService.setPaused(newPauseState)
}
// 长按说话相关处理
const handleRecordStart = async (e: React.MouseEvent | React.TouchEvent) => {
e.preventDefault(); // 防止触摸事件的默认行为
e.preventDefault() // 防止触摸事件的默认行为
if (isProcessing || isPaused) return;
if (isProcessing || isPaused) return
setIsRecording(true);
await VoiceCallService.startRecording();
};
setIsRecording(true)
await VoiceCallService.startRecording()
}
const handleRecordEnd = async (e: React.MouseEvent | React.TouchEvent) => {
e.preventDefault(); // 防止触摸事件的默认行为
e.preventDefault() // 防止触摸事件的默认行为
if (!isRecording) return;
if (!isRecording) return
// 立即更新UI状态
setIsRecording(false);
setIsProcessing(true);
setIsRecording(false)
setIsProcessing(true)
// 确保录音完全停止
try {
await VoiceCallService.stopRecording();
console.log('录音已停止');
await VoiceCallService.stopRecording()
console.log('录音已停止')
} catch (error) {
console.error('停止录音出错:', error);
console.error('停止录音出错:', error)
}
// 处理结果会通过回调函数返回，不需要在这里处理
setTimeout(() => {
setIsProcessing(false);
}, 500); // 添加短暂延迟，防止用户立即再次点击
};
setIsProcessing(false)
}, 500) // 添加短暂延迟，防止用户立即再次点击
}
// 处理鼠标/触摸离开按钮的情况
const handleRecordCancel = async (e: React.MouseEvent | React.TouchEvent) => {
e.preventDefault();
e.preventDefault()
if (isRecording) {
// 立即更新UI状态
setIsRecording(false);
setIsProcessing(true);
setIsRecording(false)
setIsProcessing(true)
// 取消录音，不发送给AI
try {
await VoiceCallService.cancelRecording();
console.log('录音已取消');
await VoiceCallService.cancelRecording()
console.log('录音已取消')
} catch (error) {
console.error('取消录音出错:', error);
console.error('取消录音出错:', error)
}
setTimeout(() => {
setIsProcessing(false);
}, 500);
setIsProcessing(false)
}, 500)
}
};
}
return (
<Modal
@ -135,8 +136,7 @@ const VoiceCallModal: React.FC<Props> = ({ visible, onClose }) => {
footer={null}
width={500}
centered
maskClosable={false}
>
maskClosable={false}>
<Container>
<VisualizerContainer>
<VoiceVisualizer isActive={isListening || isRecording} type="input" />
@ -174,7 +174,7 @@ const VoiceCallModal: React.FC<Props> = ({ visible, onClose }) => {
/>
<Tooltip title={t('voice_call.press_to_talk')}>
<RecordButton
type={isRecording ? "primary" : "default"}
type={isRecording ? 'primary' : 'default'}
icon={<SoundOutlined />}
onMouseDown={handleRecordStart}
onMouseUp={handleRecordEnd}
@ -183,8 +183,7 @@ const VoiceCallModal: React.FC<Props> = ({ visible, onClose }) => {
onTouchEnd={handleRecordEnd}
onTouchCancel={handleRecordCancel}
size="large"
disabled={isProcessing || isPaused}
>
disabled={isProcessing || isPaused}>
{isRecording ? t('voice_call.release_to_send') : t('voice_call.press_to_talk')}
</RecordButton>
</Tooltip>
@ -200,21 +199,21 @@ const VoiceCallModal: React.FC<Props> = ({ visible, onClose }) => {
</ControlsContainer>
</Container>
</Modal>
);
};
)
}
const Container = styled.div`
display: flex;
flex-direction: column;
gap: 20px;
height: 400px;
`;
`
const VisualizerContainer = styled.div`
display: flex;
justify-content: space-between;
height: 100px;
`;
`
const TranscriptContainer = styled.div`
flex: 1;
@ -223,33 +222,33 @@ const TranscriptContainer = styled.div`
border-radius: 8px;
padding: 16px;
background-color: var(--color-background-2);
`;
`
const TranscriptText = styled.p`
margin-bottom: 8px;
color: var(--color-text-1);
`;
`
const ResponseText = styled.p`
margin-bottom: 8px;
color: var(--color-primary);
`;
`
const UserLabel = styled.span`
font-weight: bold;
color: var(--color-text-1);
`;
`
const AILabel = styled.span`
font-weight: bold;
color: var(--color-primary);
`;
`
const ControlsContainer = styled.div`
display: flex;
justify-content: center;
padding: 10px 0;
`;
`
const RecordButton = styled(Button)`
min-width: 150px;
@ -258,6 +257,6 @@ const RecordButton = styled(Button)`
&:active {
transform: scale(0.95);
}
`;
`
export default VoiceCallModal;
export default VoiceCallModal

90
src/renderer/src/components/VoiceVisualizer.tsx
View File

@ -1,74 +1,74 @@
import React, { useEffect, useRef } from 'react';
import styled from 'styled-components';
import { useTranslation } from 'react-i18next';
import React, { useEffect, useRef } from 'react'
import { useTranslation } from 'react-i18next'
import styled from 'styled-components'
interface Props {
isActive: boolean;
type: 'input' | 'output';
isActive: boolean
type: 'input' | 'output'
}
const VoiceVisualizer: React.FC<Props> = ({ isActive, type }) => {
const { t } = useTranslation();
const canvasRef = useRef<HTMLCanvasElement>(null);
const animationRef = useRef<number | undefined>(undefined);
const { t } = useTranslation()
const canvasRef = useRef<HTMLCanvasElement>(null)
const animationRef = useRef<number | undefined>(undefined)
useEffect(() => {
const canvas = canvasRef.current;
if (!canvas) return;
const canvas = canvasRef.current
if (!canvas) return
const ctx = canvas.getContext('2d');
if (!ctx) return;
const ctx = canvas.getContext('2d')
if (!ctx) return
const width = canvas.width;
const height = canvas.height;
const width = canvas.width
const height = canvas.height
const drawVisualizer = () => {
ctx.clearRect(0, 0, width, height);
ctx.clearRect(0, 0, width, height)
if (!isActive) {
// 绘制静态波形
ctx.beginPath();
ctx.moveTo(0, height / 2);
ctx.lineTo(width, height / 2);
ctx.strokeStyle = type === 'input' ? 'var(--color-text-2)' : 'var(--color-primary)';
ctx.lineWidth = 2;
ctx.stroke();
return;
ctx.beginPath()
ctx.moveTo(0, height / 2)
ctx.lineTo(width, height / 2)
ctx.strokeStyle = type === 'input' ? 'var(--color-text-2)' : 'var(--color-primary)'
ctx.lineWidth = 2
ctx.stroke()
return
}
// 绘制动态波形
const barCount = 30;
const barWidth = width / barCount;
const color = type === 'input' ? 'var(--color-text-1)' : 'var(--color-primary)';
const barCount = 30
const barWidth = width / barCount
const color = type === 'input' ? 'var(--color-text-1)' : 'var(--color-primary)'
for (let i = 0; i < barCount; i++) {
const barHeight = Math.random() * (height / 2) + 10;
const x = i * barWidth;
const y = height / 2 - barHeight / 2;
const barHeight = Math.random() * (height / 2) + 10
const x = i * barWidth
const y = height / 2 - barHeight / 2
ctx.fillStyle = color;
ctx.fillRect(x, y, barWidth - 2, barHeight);
ctx.fillStyle = color
ctx.fillRect(x, y, barWidth - 2, barHeight)
}
animationRef.current = requestAnimationFrame(drawVisualizer);
};
animationRef.current = requestAnimationFrame(drawVisualizer)
}
drawVisualizer();
drawVisualizer()
return () => {
if (animationRef.current) {
cancelAnimationFrame(animationRef.current);
cancelAnimationFrame(animationRef.current)
}
};
}, [isActive, type]);
}
}, [isActive, type])
return (
<Container $type={type}>
<Label>{type === 'input' ? t('voice_call.you') : t('voice_call.ai')}</Label>
<Canvas ref={canvasRef} width={200} height={50} />
</Container>
);
};
)
}
const Container = styled.div<{ $type: 'input' | 'output' }>`
display: flex;
@ -77,21 +77,17 @@ const Container = styled.div<{ $type: 'input' | 'output' }>`
width: 45%;
border-radius: 8px;
padding: 10px;
background-color: ${props =>
props.$type === 'input'
? 'var(--color-background-3)'
: 'var(--color-primary-bg)'
};
`;
background-color: ${(props) => (props.$type === 'input' ? 'var(--color-background-3)' : 'var(--color-primary-bg)')};
`
const Label = styled.div`
margin-bottom: 8px;
font-weight: bold;
`;
`
const Canvas = styled.canvas`
width: 100%;
height: 50px;
`;
`
export default VoiceVisualizer;
export default VoiceVisualizer

15
src/renderer/src/i18n/locales/ja-jp.json
View File

@ -1507,6 +1507,21 @@
"quit": "終了",
"show_window": "ウィンドウを表示",
"visualization": "可視化"
},
"voice_call": {
"title": "[to be translated]:语音通话",
"start": "[to be translated]:开始语音通话",
"end": "[to be translated]:结束通话",
"mute": "[to be translated]:静音",
"unmute": "[to be translated]:取消静音",
"pause": "[to be translated]:暂停",
"resume": "[to be translated]:继续",
"you": "[to be translated]:您",
"ai": "[to be translated]:AI",
"press_to_talk": "[to be translated]:长按说话",
"release_to_send": "[to be translated]:松开发送",
"initialization_failed": "[to be translated]:初始化语音通话失败",
"error": "[to be translated]:语音通话出错"
}
}
}

15
src/renderer/src/i18n/locales/ru-ru.json
View File

@ -1507,6 +1507,21 @@
"quit": "Выйти",
"show_window": "Показать окно",
"visualization": "Визуализация"
},
"voice_call": {
"title": "[to be translated]:语音通话",
"start": "[to be translated]:开始语音通话",
"end": "[to be translated]:结束通话",
"mute": "[to be translated]:静音",
"unmute": "[to be translated]:取消静音",
"pause": "[to be translated]:暂停",
"resume": "[to be translated]:继续",
"you": "[to be translated]:您",
"ai": "[to be translated]:AI",
"press_to_talk": "[to be translated]:长按说话",
"release_to_send": "[to be translated]:松开发送",
"initialization_failed": "[to be translated]:初始化语音通话失败",
"error": "[to be translated]:语音通话出错"
}
}
}

15
src/renderer/src/i18n/locales/zh-tw.json
View File

@ -1507,6 +1507,21 @@
"quit": "結束",
"show_window": "顯示視窗",
"visualization": "視覺化"
},
"voice_call": {
"title": "[to be translated]:语音通话",
"start": "[to be translated]:开始语音通话",
"end": "[to be translated]:结束通话",
"mute": "[to be translated]:静音",
"unmute": "[to be translated]:取消静音",
"pause": "[to be translated]:暂停",
"resume": "[to be translated]:继续",
"you": "[to be translated]:您",
"ai": "[to be translated]:AI",
"press_to_talk": "[to be translated]:长按说话",
"release_to_send": "[to be translated]:松开发送",
"initialization_failed": "[to be translated]:初始化语音通话失败",
"error": "[to be translated]:语音通话出错"
}
}
}

10
src/renderer/src/pages/home/Messages/MessageMenubar.tsx
View File

@ -407,10 +407,12 @@ const MessageMenubar: FC<Props> = (props) => {
)}
{isAssistantMessage && ttsEnabled && (
<Tooltip title={t('chat.tts.play')} mouseEnterDelay={0.8}>
<ActionButton className="message-action-button" onClick={() => {
console.log('点击MessageMenubar中的TTS按钮，开始播放消息')
TTSService.speakFromMessage(message)
}}>
<ActionButton
className="message-action-button"
onClick={() => {
console.log('点击MessageMenubar中的TTS按钮，开始播放消息')
TTSService.speakFromMessage(message)
}}>
<SoundOutlined />
</ActionButton>
</Tooltip>

106
src/renderer/src/pages/settings/TTSSettings/TTSSettings.tsx
View File

@ -14,19 +14,19 @@ import {
setTtsEnabled,
setTtsFilterOptions,
setTtsModel,
setTtsMsOutputFormat,
setTtsMsVoice,
setTtsServiceType,
setTtsVoice,
setTtsSiliconflowApiKey,
setTtsSiliconflowApiUrl,
setTtsSiliconflowVoice,
setTtsSiliconflowModel,
setTtsSiliconflowResponseFormat,
setTtsSiliconflowSpeed,
setTtsMsVoice,
setTtsMsOutputFormat
setTtsSiliconflowVoice,
setTtsVoice
} from '@renderer/store/settings'
import { Button, Form, Input, InputNumber, message, Select, Space, Switch, Tabs, Tag } from 'antd'
import { FC, useEffect, useState, useCallback } from 'react'
import { FC, useCallback, useEffect, useState } from 'react'
import { useTranslation } from 'react-i18next'
import { useSelector } from 'react-redux'
import styled from 'styled-components'
@ -176,32 +176,30 @@ const TTSSettings: FC = () => {
// 免费在线TTS可用的语音列表
const [msTtsVoices, setMsTtsVoices] = useState<{ label: string; value: string }[]>([])
// 获取免费在线TTS可用的语音列表
const getMsTtsVoices = useCallback(async () => {
try {
// 调用API获取免费在线TTS语音列表
const response = await window.api.msTTS.getVoices();
console.log('获取到的免费在线TTS语音列表:', response);
const response = await window.api.msTTS.getVoices()
console.log('获取到的免费在线TTS语音列表:', response)
// 转换为选项格式
const voices = response.map((voice: any) => ({
label: `${voice.ShortName} (${voice.Gender === 'Female' ? '女声' : '男声'})`,
value: voice.ShortName
}));
}))
// 按语言和性别排序
voices.sort((a: any, b: any) => {
const localeA = a.value.split('-')[0] + a.value.split('-')[1];
const localeB = b.value.split('-')[0] + b.value.split('-')[1];
if (localeA !== localeB) return localeA.localeCompare(localeB);
return a.label.localeCompare(b.label);
});
const localeA = a.value.split('-')[0] + a.value.split('-')[1]
const localeB = b.value.split('-')[0] + b.value.split('-')[1]
if (localeA !== localeB) return localeA.localeCompare(localeB)
return a.label.localeCompare(b.label)
})
setMsTtsVoices(voices);
setMsTtsVoices(voices)
} catch (error) {
console.error('获取免费在线TTS语音列表失败:', error);
console.error('获取免费在线TTS语音列表失败:', error)
// 如果获取失败，设置一些默认的中文语音
setMsTtsVoices([
{ label: 'zh-CN-XiaoxiaoNeural (女声)', value: 'zh-CN-XiaoxiaoNeural' },
@ -211,10 +209,10 @@ const TTSSettings: FC = () => {
{ label: 'zh-CN-XiaomoNeural (女声)', value: 'zh-CN-XiaomoNeural' },
{ label: 'zh-CN-XiaoxuanNeural (女声)', value: 'zh-CN-XiaoxuanNeural' },
{ label: 'zh-CN-XiaoruiNeural (女声)', value: 'zh-CN-XiaoruiNeural' },
{ label: 'zh-CN-YunfengNeural (男声)', value: 'zh-CN-YunfengNeural' },
]);
{ label: 'zh-CN-YunfengNeural (男声)', value: 'zh-CN-YunfengNeural' }
])
}
}, []);
}, [])
// 获取浏览器可用的语音列表
const getVoices = useCallback(() => {
@ -323,8 +321,8 @@ const TTSSettings: FC = () => {
// 获取免费在线TTS语音列表
useEffect(() => {
// 获取免费在线TTS语音列表
getMsTtsVoices();
}, [getMsTtsVoices]);
getMsTtsVoices()
}, [getMsTtsVoices])
useEffect(() => {
// 初始化语音合成引擎
@ -634,9 +632,7 @@ const TTSSettings: FC = () => {
<Select
value={ttsSiliconflowModel}
onChange={(value) => dispatch(setTtsSiliconflowModel(value))}
options={[
{ label: 'FunAudioLLM/CosyVoice2-0.5B', value: 'FunAudioLLM/CosyVoice2-0.5B' }
]}
options={[{ label: 'FunAudioLLM/CosyVoice2-0.5B', value: 'FunAudioLLM/CosyVoice2-0.5B' }]}
disabled={!ttsEnabled}
style={{ width: '100%' }}
placeholder={t('settings.tts.siliconflow_model.placeholder')}
@ -726,24 +722,28 @@ const TTSSettings: FC = () => {
<Form.Item label={t('settings.tts.mstts.voice')} style={{ marginBottom: 16 }}>
<VoiceSelectContainer>
<Select
value={ttsMsVoice}
onChange={(value) => dispatch(setTtsMsVoice(value))}
disabled={!ttsEnabled}
style={{ width: '100%' }}
options={msTtsVoices.length > 0 ? msTtsVoices : [
{ label: 'zh-CN-XiaoxiaoNeural (女声)', value: 'zh-CN-XiaoxiaoNeural' },
{ label: 'zh-CN-YunxiNeural (男声)', value: 'zh-CN-YunxiNeural' },
{ label: 'zh-CN-YunyangNeural (男声)', value: 'zh-CN-YunyangNeural' },
{ label: 'zh-CN-XiaohanNeural (女声)', value: 'zh-CN-XiaohanNeural' },
{ label: 'zh-CN-XiaomoNeural (女声)', value: 'zh-CN-XiaomoNeural' },
{ label: 'zh-CN-XiaoxuanNeural (女声)', value: 'zh-CN-XiaoxuanNeural' },
{ label: 'zh-CN-XiaoruiNeural (女声)', value: 'zh-CN-XiaoruiNeural' },
{ label: 'zh-CN-YunfengNeural (男声)', value: 'zh-CN-YunfengNeural' },
]}
showSearch
optionFilterProp="label"
placeholder={t('settings.tts.voice.placeholder', { defaultValue: '请选择音色' })}
notFoundContent={t('settings.tts.voice.not_found', { defaultValue: '未找到音色' })}
value={ttsMsVoice}
onChange={(value) => dispatch(setTtsMsVoice(value))}
disabled={!ttsEnabled}
style={{ width: '100%' }}
options={
msTtsVoices.length > 0
? msTtsVoices
: [
{ label: 'zh-CN-XiaoxiaoNeural (女声)', value: 'zh-CN-XiaoxiaoNeural' },
{ label: 'zh-CN-YunxiNeural (男声)', value: 'zh-CN-YunxiNeural' },
{ label: 'zh-CN-YunyangNeural (男声)', value: 'zh-CN-YunyangNeural' },
{ label: 'zh-CN-XiaohanNeural (女声)', value: 'zh-CN-XiaohanNeural' },
{ label: 'zh-CN-XiaomoNeural (女声)', value: 'zh-CN-XiaomoNeural' },
{ label: 'zh-CN-XiaoxuanNeural (女声)', value: 'zh-CN-XiaoxuanNeural' },
{ label: 'zh-CN-XiaoruiNeural (女声)', value: 'zh-CN-XiaoruiNeural' },
{ label: 'zh-CN-YunfengNeural (男声)', value: 'zh-CN-YunfengNeural' }
]
}
showSearch
optionFilterProp="label"
placeholder={t('settings.tts.voice.placeholder', { defaultValue: '请选择音色' })}
notFoundContent={t('settings.tts.voice.not_found', { defaultValue: '未找到音色' })}
/>
<Button
icon={<ReloadOutlined />}
@ -754,7 +754,10 @@ const TTSSettings: FC = () => {
</VoiceSelectContainer>
{msTtsVoices.length > 0 && (
<InfoText>
{t('settings.tts.mstts.available_count', { count: msTtsVoices.length, defaultValue: '可用语音: {{count}}个' })}
{t('settings.tts.mstts.available_count', {
count: msTtsVoices.length,
defaultValue: '可用语音: {{count}}个'
})}
</InfoText>
)}
</Form.Item>
@ -767,11 +770,15 @@ const TTSSettings: FC = () => {
options={[
{ label: 'MP3 (24kHz, 48kbps)', value: 'audio-24khz-48kbitrate-mono-mp3' },
{ label: 'MP3 (24kHz, 96kbps)', value: 'audio-24khz-96kbitrate-mono-mp3' },
{ label: 'Webm (24kHz)', value: 'webm-24khz-16bit-mono-opus' },
{ label: 'Webm (24kHz)', value: 'webm-24khz-16bit-mono-opus' }
]}
/>
</Form.Item>
<InfoText>{t('settings.tts.mstts.info', { defaultValue: '免费在线TTS服务不需要API密钥，完全免费使用。' })}</InfoText>
<InfoText>
{t('settings.tts.mstts.info', {
defaultValue: '免费在线TTS服务不需要API密钥，完全免费使用。'
})}
</InfoText>
</>
)}
@ -960,11 +967,8 @@ const TTSSettings: FC = () => {
!ttsEnabled ||
(ttsServiceType === 'openai' && (!ttsApiKey || !ttsVoice || !ttsModel)) ||
(ttsServiceType === 'edge' && !ttsEdgeVoice) ||
(ttsServiceType === 'siliconflow' && (
!ttsSiliconflowApiKey ||
!ttsSiliconflowVoice ||
!ttsSiliconflowModel
))
(ttsServiceType === 'siliconflow' &&
(!ttsSiliconflowApiKey || !ttsSiliconflowVoice || !ttsSiliconflowModel))
}>
{t('settings.tts.test')}
</Button>

6
src/renderer/src/services/ASRService.ts
View File

@ -138,9 +138,9 @@ class ASRService {
// 如果有回调函数，调用一次空字符串，触发按钮状态重置
if (this.resultCallback && typeof this.resultCallback === 'function') {
// 使用空字符串调用回调，不会影响输入框，但可以触发按钮状态重置
const callback = this.resultCallback as (text: string, isFinal?: boolean) => void; // 明确指定类型
const callback = this.resultCallback as (text: string, isFinal?: boolean) => void // 明确指定类型
setTimeout(() => {
callback('', false);
callback('', false)
}, 100)
}
}
@ -334,7 +334,7 @@ class ASRService {
// 使用空字符串调用回调，不会影响输入框，但可以触发按钮状态重置
// 传递false表示这不是最终结果，只是状态更新
setTimeout(() => {
onTranscribed('', false);
onTranscribed('', false)
}, 100)
}
} else {

3
src/renderer/src/services/TTSService.ts
View File

@ -3,9 +3,10 @@
* 这个文件仅作兼容性保留，将在后续版本中移除
*/
import { TTSService as NewTTSService } from './tts/index'
import { Message } from '@renderer/types'
import { TTSService as NewTTSService } from './tts/index'
/**
* TTS服务，用于将文本转换为语音
* @deprecated 请使用 src/renderer/src/services/tts/TTSService.ts

458
src/renderer/src/services/VoiceCallService.ts
View File

@ -1,180 +1,180 @@
import store from '@renderer/store';
import { fetchChatCompletion } from '@renderer/services/ApiService';
import { getAssistantMessage, getUserMessage } from '@renderer/services/MessagesService';
import { getDefaultAssistant } from '@renderer/services/AssistantService';
import TTSService from '@renderer/services/TTSService';
import ASRService from '@renderer/services/ASRService';
import { fetchChatCompletion } from '@renderer/services/ApiService'
import ASRService from '@renderer/services/ASRService'
import { getDefaultAssistant } from '@renderer/services/AssistantService'
import { getAssistantMessage, getUserMessage } from '@renderer/services/MessagesService'
import TTSService from '@renderer/services/TTSService'
import store from '@renderer/store'
// 导入类型
import type { Message } from '@renderer/types';
import type { Message } from '@renderer/types'
interface VoiceCallCallbacks {
onTranscript: (text: string) => void;
onResponse: (text: string) => void;
onListeningStateChange: (isListening: boolean) => void;
onSpeakingStateChange: (isSpeaking: boolean) => void;
onTranscript: (text: string) => void
onResponse: (text: string) => void
onListeningStateChange: (isListening: boolean) => void
onSpeakingStateChange: (isSpeaking: boolean) => void
}
// 为TypeScript添加SpeechRecognition类型
declare global {
interface Window {
SpeechRecognition: any;
webkitSpeechRecognition: any;
SpeechRecognition: any
webkitSpeechRecognition: any
}
}
class VoiceCallServiceClass {
private recognition: any = null;
private isCallActive = false;
private isRecording = false; // 新增录音状态
private isMuted = false;
private isPaused = false;
private callbacks: VoiceCallCallbacks | null = null;
private _currentTranscript = ''; // 使用下划线前缀避免未使用警告
private _accumulatedTranscript = ''; // 累积的语音识别结果
private conversationHistory: { role: string; content: string }[] = [];
private isProcessingResponse = false;
private ttsService = TTSService;
private recordingTimeout: NodeJS.Timeout | null = null; // 录音超时定时器
private recognition: any = null
private isCallActive = false
private isRecording = false // 新增录音状态
private isMuted = false
private isPaused = false
private callbacks: VoiceCallCallbacks | null = null
private _currentTranscript = '' // 使用下划线前缀避免未使用警告
private _accumulatedTranscript = '' // 累积的语音识别结果
private conversationHistory: { role: string; content: string }[] = []
private isProcessingResponse = false
private ttsService = TTSService
private recordingTimeout: NodeJS.Timeout | null = null // 录音超时定时器
async initialize() {
// 检查麦克风权限
try {
const stream = await navigator.mediaDevices.getUserMedia({ audio: true });
stream.getTracks().forEach(track => track.stop());
const stream = await navigator.mediaDevices.getUserMedia({ audio: true })
stream.getTracks().forEach((track) => track.stop())
} catch (error) {
console.error('Microphone permission denied:', error);
throw new Error('Microphone permission denied');
console.error('Microphone permission denied:', error)
throw new Error('Microphone permission denied')
}
// 获取当前ASR服务类型
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
// 如果使用浏览器ASR，检查浏览器支持
if (asrServiceType === 'browser') {
if (!('webkitSpeechRecognition' in window) && !('SpeechRecognition' in window)) {
throw new Error('Speech recognition not supported in this browser');
throw new Error('Speech recognition not supported in this browser')
}
// 初始化浏览器语音识别
const SpeechRecognition = window.SpeechRecognition || window.webkitSpeechRecognition;
this.recognition = new SpeechRecognition();
this.recognition.continuous = true;
this.recognition.interimResults = true;
this.recognition.lang = navigator.language || 'zh-CN';
const SpeechRecognition = window.SpeechRecognition || window.webkitSpeechRecognition
this.recognition = new SpeechRecognition()
this.recognition.continuous = true
this.recognition.interimResults = true
this.recognition.lang = navigator.language || 'zh-CN'
} else if (asrServiceType === 'local') {
// 如果使用本地服务器ASR，检查连接
try {
// 尝试连接本地ASR服务器
const connected = await ASRService.connectToWebSocketServer();
const connected = await ASRService.connectToWebSocketServer()
if (!connected) {
throw new Error('无法连接到语音识别服务');
throw new Error('无法连接到语音识别服务')
}
} catch (error) {
console.error('Failed to connect to ASR server:', error);
throw new Error('Failed to connect to ASR server');
console.error('Failed to connect to ASR server:', error)
throw new Error('Failed to connect to ASR server')
}
}
return true;
return true
}
async startCall(callbacks: VoiceCallCallbacks) {
this.callbacks = callbacks;
this.isCallActive = true;
this.conversationHistory = [];
this.callbacks = callbacks
this.isCallActive = true
this.conversationHistory = []
// 获取当前ASR服务类型
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
// 根据不同的ASR服务类型进行初始化
if (asrServiceType === 'browser') {
if (!this.recognition) {
throw new Error('Browser speech recognition not initialized');
throw new Error('Browser speech recognition not initialized')
}
// 设置浏览器语音识别事件处理
this.recognition.onresult = (event: any) => {
let interimTranscript = '';
let finalTranscript = '';
let interimTranscript = ''
let finalTranscript = ''
for (let i = event.resultIndex; i < event.results.length; ++i) {
if (event.results[i].isFinal) {
finalTranscript += event.results[i][0].transcript;
finalTranscript += event.results[i][0].transcript
} else {
interimTranscript += event.results[i][0].transcript;
interimTranscript += event.results[i][0].transcript
}
}
if (interimTranscript) {
// 更新当前的临时识别结果
this._currentTranscript = interimTranscript;
this._currentTranscript = interimTranscript
// 显示累积结果 + 当前临时结果
this.callbacks?.onTranscript(this._accumulatedTranscript + ' ' + interimTranscript);
this.callbacks?.onTranscript(this._accumulatedTranscript + ' ' + interimTranscript)
}
if (finalTranscript) {
// 将最终结果累积到总结果中
if (this._accumulatedTranscript) {
// 如果已经有累积的文本，添加空格再追加
this._accumulatedTranscript += ' ' + finalTranscript;
this._accumulatedTranscript += ' ' + finalTranscript
} else {
// 如果是第一段文本，直接设置
this._accumulatedTranscript = finalTranscript;
this._accumulatedTranscript = finalTranscript
}
// 更新当前的识别结果
this._currentTranscript = '';
this._currentTranscript = ''
// 显示累积的完整结果
this.callbacks?.onTranscript(this._accumulatedTranscript);
this.callbacks?.onTranscript(this._accumulatedTranscript)
// 在录音过程中只更新transcript，不触发handleUserSpeech
// 松开按钮后才会处理完整的录音内容
}
};
}
this.recognition.onstart = () => {
this.isRecording = true;
this.callbacks?.onListeningStateChange(true);
};
this.isRecording = true
this.callbacks?.onListeningStateChange(true)
}
this.recognition.onend = () => {
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
};
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
}
this.recognition.onerror = (event: any) => {
console.error('Speech recognition error', event.error);
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
};
console.error('Speech recognition error', event.error)
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
}
}
// 播放欢迎语音
const welcomeMessage = '您好，我是您的AI助手，请长按说话按钮进行对话。';
this.callbacks?.onResponse(welcomeMessage);
const welcomeMessage = '您好，我是您的AI助手，请长按说话按钮进行对话。'
this.callbacks?.onResponse(welcomeMessage)
// 监听TTS状态
const ttsStateHandler = (isPlaying: boolean) => {
this.callbacks?.onSpeakingStateChange(isPlaying);
};
this.callbacks?.onSpeakingStateChange(isPlaying)
}
// 监听TTS播放状态
window.addEventListener('tts-state-change', (event: any) => {
ttsStateHandler(event.detail.isPlaying);
});
ttsStateHandler(event.detail.isPlaying)
})
// 播放欢迎语音，并手动设置初始状态
this.callbacks?.onSpeakingStateChange(true);
this.ttsService.speak(welcomeMessage);
this.callbacks?.onSpeakingStateChange(true)
this.ttsService.speak(welcomeMessage)
// 确保欢迎语音结束后状态正确
setTimeout(() => {
if (this.ttsService && !this.ttsService.isCurrentlyPlaying()) {
this.callbacks?.onSpeakingStateChange(false);
this.callbacks?.onSpeakingStateChange(false)
}
}, 5000); // 5秒后检查TTS状态
}, 5000) // 5秒后检查TTS状态
return true;
return true
}
/**
@ -183,25 +183,24 @@ class VoiceCallServiceClass {
*/
async startRecording(): Promise<boolean> {
if (!this.isCallActive || this.isPaused || this.isProcessingResponse || this.isRecording) {
return false;
return false
}
// 重置累积的文本
this._accumulatedTranscript = '';
this._accumulatedTranscript = ''
// 获取当前ASR服务类型
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
try {
if (asrServiceType === 'browser') {
// 浏览器ASR
if (!this.recognition) {
throw new Error('Browser speech recognition not initialized');
throw new Error('Browser speech recognition not initialized')
}
this.recognition.start();
this.isRecording = true;
this.recognition.start()
this.isRecording = true
} else if (asrServiceType === 'local') {
// 本地服务器ASR
await ASRService.startRecording((text, isFinal) => {
@ -210,51 +209,50 @@ class VoiceCallServiceClass {
// 如果是最终结果，累积到总结果中
if (this._accumulatedTranscript) {
// 如果已经有累积的文本，添加空格再追加
this._accumulatedTranscript += ' ' + text;
this._accumulatedTranscript += ' ' + text
} else {
// 如果是第一段文本，直接设置
this._accumulatedTranscript = text;
this._accumulatedTranscript = text
}
// 更新当前的识别结果
this._currentTranscript = '';
this._currentTranscript = ''
// 显示累积的完整结果
this.callbacks?.onTranscript(this._accumulatedTranscript);
this.callbacks?.onTranscript(this._accumulatedTranscript)
} else {
// 如果是临时结果，更新当前的识别结果
this._currentTranscript = text;
this._currentTranscript = text
// 显示累积结果 + 当前临时结果
this.callbacks?.onTranscript(this._accumulatedTranscript + ' ' + text);
this.callbacks?.onTranscript(this._accumulatedTranscript + ' ' + text)
}
// 在录音过程中只更新transcript，不触发handleUserSpeech
// 松开按钮后才会处理完整的录音内容
}
});
this.isRecording = true;
this.callbacks?.onListeningStateChange(true);
})
this.isRecording = true
this.callbacks?.onListeningStateChange(true)
} else if (asrServiceType === 'openai') {
// OpenAI ASR
await ASRService.startRecording();
this.isRecording = true;
this.callbacks?.onListeningStateChange(true);
await ASRService.startRecording()
this.isRecording = true
this.callbacks?.onListeningStateChange(true)
}
// 设置最长录音时间，防止用户忘记松开
this.recordingTimeout = setTimeout(() => {
if (this.isRecording) {
this.stopRecording();
this.stopRecording()
}
}, 60000); // 60秒最长录音时间
}, 60000) // 60秒最长录音时间
return true;
return true
} catch (error) {
console.error('Failed to start recording:', error);
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
return false;
console.error('Failed to start recording:', error)
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
return false
}
}
@ -264,143 +262,141 @@ class VoiceCallServiceClass {
*/
async stopRecording(): Promise<boolean> {
if (!this.isCallActive || !this.isRecording) {
return false;
return false
}
// 清除录音超时定时器
if (this.recordingTimeout) {
clearTimeout(this.recordingTimeout);
this.recordingTimeout = null;
clearTimeout(this.recordingTimeout)
this.recordingTimeout = null
}
// 获取当前ASR服务类型
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
try {
// 存储当前的语音识别结果，用于松开按钮后发送给AI
const currentTranscript = this._currentTranscript;
const currentTranscript = this._currentTranscript
// 存储累积的语音识别结果
const accumulatedTranscript = this._accumulatedTranscript;
const accumulatedTranscript = this._accumulatedTranscript
if (asrServiceType === 'browser') {
// 浏览器ASR
if (!this.recognition) {
throw new Error('Browser speech recognition not initialized');
throw new Error('Browser speech recognition not initialized')
}
this.recognition.stop();
this.recognition.stop()
// onend事件将设置isRecording = false
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
// 优先使用累积的文本，如果有的话
if (accumulatedTranscript) {
console.log('发送累积的语音识别结果给AI:', accumulatedTranscript);
this.handleUserSpeech(accumulatedTranscript);
console.log('发送累积的语音识别结果给AI:', accumulatedTranscript)
this.handleUserSpeech(accumulatedTranscript)
} else if (currentTranscript) {
// 如果没有累积结果，使用当前结果
console.log('没有累积结果，使用当前结果:', currentTranscript);
this.handleUserSpeech(currentTranscript);
console.log('没有累积结果，使用当前结果:', currentTranscript)
this.handleUserSpeech(currentTranscript)
}
} else if (asrServiceType === 'local') {
// 本地服务器ASR
// 创建一个承诺，等待最终结果
const finalResultPromise = new Promise<string>((resolve) => {
// 设置一个超时器，确保不会无限等待
const timeoutId = setTimeout(() => {
console.log('等待最终结果超时，使用当前结果');
resolve(this._currentTranscript);
}, 1500); // 1.5秒超时
console.log('等待最终结果超时，使用当前结果')
resolve(this._currentTranscript)
}, 1500) // 1.5秒超时
// 设置回调函数来接收最终结果
const resultCallback = (text: string) => {
// 如果是空字符串，表示只是重置状态，不处理
if (text === '') return;
if (text === '') return
if (text) {
clearTimeout(timeoutId);
console.log('收到最终语音识别结果:', text);
this._currentTranscript = text;
this.callbacks?.onTranscript(text);
resolve(text);
clearTimeout(timeoutId)
console.log('收到最终语音识别结果:', text)
this._currentTranscript = text
this.callbacks?.onTranscript(text)
resolve(text)
}
};
}
// 停止录音，但不取消，以获取最终结果
ASRService.stopRecording(resultCallback);
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
ASRService.stopRecording(resultCallback)
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
// 添加额外的安全措施，在停止后立即发送重置命令
setTimeout(() => {
// 发送重置命令，确保浏览器不会继续发送结果
ASRService.cancelRecording();
}, 2000); // 2秒后强制取消，作为安全措施
});
ASRService.cancelRecording()
}, 2000) // 2秒后强制取消，作为安全措施
})
// 等待最终结果
const finalText = await finalResultPromise;
const finalText = await finalResultPromise
// 优先使用累积的文本，如果有的话
if (accumulatedTranscript) {
console.log('发送累积的语音识别结果给AI:', accumulatedTranscript);
this.handleUserSpeech(accumulatedTranscript);
console.log('发送累积的语音识别结果给AI:', accumulatedTranscript)
this.handleUserSpeech(accumulatedTranscript)
} else if (finalText) {
// 如果没有累积结果，使用最终结果
console.log('发送最终语音识别结果给AI:', finalText);
this.handleUserSpeech(finalText);
console.log('发送最终语音识别结果给AI:', finalText)
this.handleUserSpeech(finalText)
} else if (currentTranscript) {
// 如果没有最终结果，使用当前结果
console.log('没有最终结果，使用当前结果:', currentTranscript);
this.handleUserSpeech(currentTranscript);
console.log('没有最终结果，使用当前结果:', currentTranscript)
this.handleUserSpeech(currentTranscript)
}
} else if (asrServiceType === 'openai') {
// OpenAI ASR
await ASRService.stopRecording((text) => {
// 更新最终的语音识别结果
if (text) {
this._currentTranscript = text;
this.callbacks?.onTranscript(text);
this._currentTranscript = text
this.callbacks?.onTranscript(text)
}
});
})
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
// 使用最新的语音识别结果
const finalTranscript = this._currentTranscript;
const finalTranscript = this._currentTranscript
if (finalTranscript) {
this.handleUserSpeech(finalTranscript);
this.handleUserSpeech(finalTranscript)
}
}
return true;
return true
} catch (error) {
console.error('Failed to stop recording:', error);
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
return false;
console.error('Failed to stop recording:', error)
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
return false
}
}
async handleUserSpeech(text: string) {
if (!this.isCallActive || this.isProcessingResponse || this.isPaused) return;
if (!this.isCallActive || this.isProcessingResponse || this.isPaused) return
// 暂停语音识别，避免在AI回复时继续识别
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
if (asrServiceType === 'browser') {
this.recognition?.stop();
this.recognition?.stop()
} else if (asrServiceType === 'local' || asrServiceType === 'openai') {
ASRService.cancelRecording();
ASRService.cancelRecording()
}
this.isProcessingResponse = true;
this.isProcessingResponse = true
try {
// 获取当前助手
const assistant = getDefaultAssistant();
const assistant = getDefaultAssistant()
// 创建一个简单的Topic对象
const topic = {
@ -410,7 +406,7 @@ class VoiceCallServiceClass {
createdAt: new Date().toISOString(),
updatedAt: new Date().toISOString(),
messages: []
};
}
// 创建用户消息
const userMessage = getUserMessage({
@ -418,35 +414,35 @@ class VoiceCallServiceClass {
topic,
type: 'text',
content: text
});
})
// 创建助手消息
const assistantMessage = getAssistantMessage({
assistant,
topic
});
})
// 更新对话历史
this.conversationHistory.push({ role: 'user', content: text });
this.conversationHistory.push({ role: 'user', content: text })
// 构建消息列表
// 将历史消息转换为正确的Message对象
const historyMessages = this.conversationHistory.map(msg => {
const historyMessages = this.conversationHistory.map((msg) => {
if (msg.role === 'user') {
return getUserMessage({
assistant,
topic,
type: 'text',
content: msg.content
});
})
} else {
const assistantMsg = getAssistantMessage({
assistant,
topic
});
return { ...assistantMsg, content: msg.content, status: 'success' };
})
return { ...assistantMsg, content: msg.content, status: 'success' }
}
});
})
// 修改用户消息，添加语音通话提示
const voiceCallPrompt = `当前是语音通话模式。请注意：
@ -457,7 +453,7 @@ class VoiceCallServiceClass {
5. 回答应该简短有力，便于用户通过语音理解。
6. 避免使用特殊符号、表情符号、标点符号等，因为这些在语音播放时会影响理解。
7. 使用完整的句子而非简单的关键词列表。
8. 尽量使用常见词汇，避免生僻或专业术语，除非用户特别询问。`;
8. 尽量使用常见词汇，避免生僻或专业术语，除非用户特别询问。`
// 创建系统指令消息
const systemMessage = getUserMessage({
@ -465,17 +461,17 @@ class VoiceCallServiceClass {
topic,
type: 'text',
content: voiceCallPrompt
});
})
// 修改用户消息的内容
userMessage.content = text;
userMessage.content = text
// 构建最终消息列表
// 使用类型断言解决类型问题
const messages = [systemMessage, ...historyMessages, userMessage] as Message[];
const messages = [systemMessage, ...historyMessages, userMessage] as Message[]
// 流式响应处理
let fullResponse = '';
let fullResponse = ''
try {
// 调用真实的LLM API
@ -485,60 +481,58 @@ class VoiceCallServiceClass {
assistant,
onResponse: async (msg) => {
if (msg.content && msg.content !== fullResponse) {
fullResponse = msg.content;
fullResponse = msg.content
// 更新UI
this.callbacks?.onResponse(fullResponse);
this.callbacks?.onResponse(fullResponse)
// 如果TTS正在播放，停止它
if (this.ttsService.isCurrentlyPlaying()) {
this.ttsService.stop();
this.ttsService.stop()
}
}
}
});
})
// 播放完整响应
if (!this.isMuted && this.isCallActive) {
// 手动设置语音状态
this.callbacks?.onSpeakingStateChange(true);
this.ttsService.speak(fullResponse);
this.callbacks?.onSpeakingStateChange(true)
this.ttsService.speak(fullResponse)
// 确保语音结束后状态正确
setTimeout(() => {
if (this.ttsService && !this.ttsService.isCurrentlyPlaying()) {
this.callbacks?.onSpeakingStateChange(false);
this.callbacks?.onSpeakingStateChange(false)
}
}, 1000); // 1秒后检查TTS状态
}, 1000) // 1秒后检查TTS状态
}
// 更新对话历史
this.conversationHistory.push({ role: 'assistant', content: fullResponse });
this.conversationHistory.push({ role: 'assistant', content: fullResponse })
} catch (innerError) {
console.error('Error generating response:', innerError);
console.error('Error generating response:', innerError)
// 如果出错，使用一个简单的回复
fullResponse = `抱歉，处理您的请求时出错了。`;
this.callbacks?.onResponse(fullResponse);
fullResponse = `抱歉，处理您的请求时出错了。`
this.callbacks?.onResponse(fullResponse)
if (!this.isMuted && this.isCallActive) {
// 手动设置语音状态
this.callbacks?.onSpeakingStateChange(true);
this.ttsService.speak(fullResponse);
this.callbacks?.onSpeakingStateChange(true)
this.ttsService.speak(fullResponse)
// 确保语音结束后状态正确
setTimeout(() => {
if (this.ttsService && !this.ttsService.isCurrentlyPlaying()) {
this.callbacks?.onSpeakingStateChange(false);
this.callbacks?.onSpeakingStateChange(false)
}
}, 1000); // 1秒后检查TTS状态
}, 1000) // 1秒后检查TTS状态
}
}
} catch (error) {
console.error('Error processing voice response:', error);
console.error('Error processing voice response:', error)
} finally {
this.isProcessingResponse = false;
this.isProcessingResponse = false
// 不自动恢复语音识别，等待用户长按按钮
// 长按说话模式下，我们不需要自动恢复语音识别
@ -551,106 +545,104 @@ class VoiceCallServiceClass {
*/
async cancelRecording(): Promise<boolean> {
if (!this.isCallActive || !this.isRecording) {
return false;
return false
}
// 清除录音超时定时器
if (this.recordingTimeout) {
clearTimeout(this.recordingTimeout);
this.recordingTimeout = null;
clearTimeout(this.recordingTimeout)
this.recordingTimeout = null
}
// 获取当前ASR服务类型
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
try {
if (asrServiceType === 'browser') {
// 浏览器ASR
if (!this.recognition) {
throw new Error('Browser speech recognition not initialized');
throw new Error('Browser speech recognition not initialized')
}
this.recognition.stop();
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
this.recognition.stop()
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
} else if (asrServiceType === 'local') {
// 本地服务器ASR
ASRService.cancelRecording();
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
ASRService.cancelRecording()
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
} else if (asrServiceType === 'openai') {
// OpenAI ASR
ASRService.cancelRecording();
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
ASRService.cancelRecording()
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
}
// 清除当前识别结果
this._currentTranscript = '';
this.callbacks?.onTranscript('');
this._currentTranscript = ''
this.callbacks?.onTranscript('')
return true;
return true
} catch (error) {
console.error('Failed to cancel recording:', error);
this.isRecording = false;
this.callbacks?.onListeningStateChange(false);
return false;
console.error('Failed to cancel recording:', error)
this.isRecording = false
this.callbacks?.onListeningStateChange(false)
return false
}
}
setMuted(muted: boolean) {
this.isMuted = muted;
this.isMuted = muted
// 如果设置为静音，停止当前TTS播放
if (muted && this.ttsService.isCurrentlyPlaying()) {
this.ttsService.stop();
this.ttsService.stop()
}
}
setPaused(paused: boolean) {
this.isPaused = paused;
this.isPaused = paused
// 获取当前ASR服务类型
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
if (paused) {
// 暂停语音识别
if (asrServiceType === 'browser') {
this.recognition?.stop();
this.recognition?.stop()
} else if (asrServiceType === 'local' || asrServiceType === 'openai') {
ASRService.cancelRecording();
ASRService.cancelRecording()
}
// 暂停TTS
if (this.ttsService.isCurrentlyPlaying()) {
this.ttsService.stop();
this.ttsService.stop()
}
}
// 不自动恢复语音识别，等待用户长按按钮
}
endCall() {
this.isCallActive = false;
this.isCallActive = false
// 获取当前ASR服务类型
const { asrServiceType } = store.getState().settings;
const { asrServiceType } = store.getState().settings
// 停止语音识别
if (asrServiceType === 'browser') {
this.recognition?.stop();
this.recognition?.stop()
} else if (asrServiceType === 'local' || asrServiceType === 'openai') {
ASRService.cancelRecording();
ASRService.cancelRecording()
}
// 停止TTS
if (this.ttsService.isCurrentlyPlaying()) {
this.ttsService.stop();
this.ttsService.stop()
}
this.callbacks = null;
this.callbacks = null
}
}
export const VoiceCallService = new VoiceCallServiceClass();
export const VoiceCallService = new VoiceCallServiceClass()

191
src/renderer/src/services/tts/EdgeTTSService.ts
View File

@ -1,22 +1,23 @@
import { TTSServiceInterface } from './TTSServiceInterface';
import i18n from '@renderer/i18n';
import i18n from '@renderer/i18n'
import { TTSServiceInterface } from './TTSServiceInterface'
// 全局变量来跟踪当前正在播放的语音
let currentUtterance: SpeechSynthesisUtterance | null = null;
let currentUtterance: SpeechSynthesisUtterance | null = null
/**
* Edge TTS服务实现类
*/
export class EdgeTTSService implements TTSServiceInterface {
private edgeVoice: string;
private edgeVoice: string
/**
* 构造函数
* @param edgeVoice Edge语音
*/
constructor(edgeVoice: string) {
this.edgeVoice = edgeVoice;
console.log('初始化EdgeTTSService，语音:', edgeVoice);
this.edgeVoice = edgeVoice
console.log('初始化EdgeTTSService，语音:', edgeVoice)
}
/**
@ -25,7 +26,7 @@ export class EdgeTTSService implements TTSServiceInterface {
*/
private validateParams(): void {
if (!this.edgeVoice) {
throw new Error(i18n.t('settings.tts.error.no_edge_voice'));
throw new Error(i18n.t('settings.tts.error.no_edge_voice'))
}
}
@ -37,79 +38,79 @@ export class EdgeTTSService implements TTSServiceInterface {
private playDirectly(text: string): boolean {
try {
// 验证参数
this.validateParams();
this.validateParams()
// 使用Web Speech API
if (!('speechSynthesis' in window)) {
throw new Error(i18n.t('settings.tts.error.browser_not_support'));
throw new Error(i18n.t('settings.tts.error.browser_not_support'))
}
// 停止当前正在播放的语音
window.speechSynthesis.cancel();
window.speechSynthesis.cancel()
if (currentUtterance) {
currentUtterance = null;
currentUtterance = null
}
// 创建语音合成器实例
const utterance = new SpeechSynthesisUtterance(text);
currentUtterance = utterance;
const utterance = new SpeechSynthesisUtterance(text)
currentUtterance = utterance
// 获取可用的语音合成声音
const voices = window.speechSynthesis.getVoices();
console.log('可用的语音合成声音:', voices);
const voices = window.speechSynthesis.getVoices()
console.log('可用的语音合成声音:', voices)
// 查找指定的语音
let selectedVoice = voices.find((v) => v.name === this.edgeVoice);
let selectedVoice = voices.find((v) => v.name === this.edgeVoice)
// 如果没有找到指定的语音，尝试使用中文语音
if (!selectedVoice) {
console.warn('未找到指定的语音:', this.edgeVoice);
console.warn('未找到指定的语音:', this.edgeVoice)
// 尝试找中文语音
selectedVoice = voices.find((v) => v.lang === 'zh-CN');
selectedVoice = voices.find((v) => v.lang === 'zh-CN')
if (selectedVoice) {
console.log('使用替代中文语音:', selectedVoice.name);
console.log('使用替代中文语音:', selectedVoice.name)
} else {
// 如果没有中文语音，使用第一个可用的语音
if (voices.length > 0) {
selectedVoice = voices[0];
console.log('使用第一个可用的语音:', selectedVoice.name);
selectedVoice = voices[0]
console.log('使用第一个可用的语音:', selectedVoice.name)
} else {
console.warn('没有可用的语音');
return false;
console.warn('没有可用的语音')
return false
}
}
} else {
console.log('已选择语音:', selectedVoice.name);
console.log('已选择语音:', selectedVoice.name)
}
// 设置语音
if (selectedVoice) {
utterance.voice = selectedVoice;
utterance.voice = selectedVoice
}
// 设置事件处理程序
utterance.onend = () => {
console.log('语音合成已结束');
currentUtterance = null;
console.log('语音合成已结束')
currentUtterance = null
// 分发一个自定义事件，通知语音合成已结束
// 这样TTSService可以监听这个事件并重置播放状态
const event = new CustomEvent('edgeTTSComplete', { detail: { text } });
document.dispatchEvent(event);
};
const event = new CustomEvent('edgeTTSComplete', { detail: { text } })
document.dispatchEvent(event)
}
utterance.onerror = (event) => {
console.error('语音合成错误:', event);
currentUtterance = null;
};
console.error('语音合成错误:', event)
currentUtterance = null
}
// 开始语音合成
window.speechSynthesis.speak(utterance);
return true;
window.speechSynthesis.speak(utterance)
return true
} catch (error) {
console.error('直接播放语音失败:', error);
return false;
console.error('直接播放语音失败:', error)
return false
}
}
@ -120,151 +121,151 @@ export class EdgeTTSService implements TTSServiceInterface {
*/
async synthesize(text: string): Promise<Blob> {
// 验证参数
this.validateParams();
this.validateParams()
// 先尝试直接播放
const playResult = this.playDirectly(text);
const playResult = this.playDirectly(text)
if (playResult) {
// 如果直接播放成功，返回一个有效的音频Blob
// 创建一个简单的音频文件，包含一个短暂停
// 这个文件可以被浏览器正常播放，但实际上不会发出声音
// 因为我们已经使用Web Speech API直接播放了语音
const silentAudioBase64 = 'UklGRiQAAABXQVZFZm10IBAAAAABAAEARKwAAIhYAQACABAAZGF0YQAAAAA=';
const silentAudioBuffer = Uint8Array.from(atob(silentAudioBase64), c => c.charCodeAt(0));
return new Blob([silentAudioBuffer], { type: 'audio/wav' });
const silentAudioBase64 = 'UklGRiQAAABXQVZFZm10IBAAAAABAAEARKwAAIhYAQACABAAZGF0YQAAAAA='
const silentAudioBuffer = Uint8Array.from(atob(silentAudioBase64), (c) => c.charCodeAt(0))
return new Blob([silentAudioBuffer], { type: 'audio/wav' })
}
// 如果直接播放失败，尝试录制方法
console.log('直接播放失败，尝试录制方法');
console.log('直接播放失败，尝试录制方法')
try {
console.log('使用浏览器TTS生成语音，音色:', this.edgeVoice);
console.log('使用浏览器TTS生成语音，音色:', this.edgeVoice)
// 使用Web Speech API
if (!('speechSynthesis' in window)) {
throw new Error(i18n.t('settings.tts.error.browser_not_support'));
throw new Error(i18n.t('settings.tts.error.browser_not_support'))
}
// 停止当前正在播放的语音
window.speechSynthesis.cancel();
window.speechSynthesis.cancel()
// 创建语音合成器实例
const utterance = new SpeechSynthesisUtterance(text);
const utterance = new SpeechSynthesisUtterance(text)
// 获取可用的语音合成声音
const voices = window.speechSynthesis.getVoices();
console.log('初始可用的语音合成声音:', voices);
const voices = window.speechSynthesis.getVoices()
console.log('初始可用的语音合成声音:', voices)
// 如果没有可用的声音，等待声音加载
if (voices.length === 0) {
try {
await new Promise<void>((resolve) => {
const voicesChangedHandler = () => {
window.speechSynthesis.onvoiceschanged = null;
resolve();
};
window.speechSynthesis.onvoiceschanged = voicesChangedHandler;
window.speechSynthesis.onvoiceschanged = null
resolve()
}
window.speechSynthesis.onvoiceschanged = voicesChangedHandler
// 设置超时，防止无限等待
setTimeout(() => {
window.speechSynthesis.onvoiceschanged = null;
resolve();
}, 5000);
});
window.speechSynthesis.onvoiceschanged = null
resolve()
}, 5000)
})
} catch (error) {
console.error('等待语音加载超时:', error);
console.error('等待语音加载超时:', error)
}
}
// 重新获取可用的语音合成声音
const updatedVoices = window.speechSynthesis.getVoices();
console.log('更新后可用的语音合成声音:', updatedVoices);
const updatedVoices = window.speechSynthesis.getVoices()
console.log('更新后可用的语音合成声音:', updatedVoices)
// 查找指定的语音
let selectedVoice = updatedVoices.find((v) => v.name === this.edgeVoice);
let selectedVoice = updatedVoices.find((v) => v.name === this.edgeVoice)
// 如果没有找到指定的语音，尝试使用中文语音
if (!selectedVoice) {
console.warn('未找到指定的语音:', this.edgeVoice);
console.warn('未找到指定的语音:', this.edgeVoice)
// 尝试找中文语音
selectedVoice = updatedVoices.find((v) => v.lang === 'zh-CN');
selectedVoice = updatedVoices.find((v) => v.lang === 'zh-CN')
if (selectedVoice) {
console.log('使用替代中文语音:', selectedVoice.name);
console.log('使用替代中文语音:', selectedVoice.name)
} else {
// 如果没有中文语音，使用第一个可用的语音
if (updatedVoices.length > 0) {
selectedVoice = updatedVoices[0];
console.log('使用第一个可用的语音:', selectedVoice.name);
selectedVoice = updatedVoices[0]
console.log('使用第一个可用的语音:', selectedVoice.name)
} else {
console.warn('没有可用的语音');
console.warn('没有可用的语音')
}
}
} else {
console.log('已选择语音:', selectedVoice.name);
console.log('已选择语音:', selectedVoice.name)
}
// 设置语音
if (selectedVoice) {
utterance.voice = selectedVoice;
utterance.voice = selectedVoice
}
// 创建一个Promise来等待语音合成完成
return await new Promise<Blob>((resolve, reject) => {
try {
// 使用AudioContext捕获语音合成的音频
const audioContext = new (window.AudioContext || (window as any).webkitAudioContext)();
const audioDestination = audioContext.createMediaStreamDestination();
const mediaRecorder = new MediaRecorder(audioDestination.stream);
const audioChunks: BlobPart[] = [];
const audioContext = new (window.AudioContext || (window as any).webkitAudioContext)()
const audioDestination = audioContext.createMediaStreamDestination()
const mediaRecorder = new MediaRecorder(audioDestination.stream)
const audioChunks: BlobPart[] = []
mediaRecorder.ondataavailable = (event) => {
if (event.data.size > 0) {
audioChunks.push(event.data);
audioChunks.push(event.data)
}
};
}
mediaRecorder.onstop = () => {
const audioBlob = new Blob(audioChunks, { type: 'audio/wav' });
resolve(audioBlob);
};
const audioBlob = new Blob(audioChunks, { type: 'audio/wav' })
resolve(audioBlob)
}
// 开始录制
mediaRecorder.start();
mediaRecorder.start()
// 设置语音合成事件
utterance.onend = () => {
// 语音合成结束后停止录制
setTimeout(() => {
mediaRecorder.stop();
}, 500); // 等待一下，确保所有音频都被捕获
};
mediaRecorder.stop()
}, 500) // 等待一下，确保所有音频都被捕获
}
utterance.onerror = (event) => {
console.error('语音合成错误:', event);
mediaRecorder.stop();
reject(new Error('语音合成错误'));
};
console.error('语音合成错误:', event)
mediaRecorder.stop()
reject(new Error('语音合成错误'))
}
// 开始语音合成
window.speechSynthesis.speak(utterance);
window.speechSynthesis.speak(utterance)
// 设置超时，防止无限等待
setTimeout(() => {
if (mediaRecorder.state === 'recording') {
console.warn('语音合成超时，强制停止');
mediaRecorder.stop();
console.warn('语音合成超时，强制停止')
mediaRecorder.stop()
}
}, 10000); // 10秒超时
}, 10000) // 10秒超时
} catch (error: any) {
console.error('浏览器TTS语音合成失败:', error);
reject(new Error(`浏览器TTS语音合成失败: ${error?.message || '未知错误'}`));
console.error('浏览器TTS语音合成失败:', error)
reject(new Error(`浏览器TTS语音合成失败: ${error?.message || '未知错误'}`))
}
});
})
} catch (error: any) {
console.error('浏览器TTS语音合成失败:', error);
console.error('浏览器TTS语音合成失败:', error)
// 即使失败也返回一个空的Blob，而不是抛出异常
// 这样可以避免在UI上显示错误消息
return new Blob([], { type: 'audio/wav' });
return new Blob([], { type: 'audio/wav' })
}
}
}

31
src/renderer/src/services/tts/MsTTSService.ts
View File

@ -1,13 +1,14 @@
import { TTSServiceInterface } from './TTSServiceInterface';
import i18n from '@renderer/i18n';
import i18n from '@renderer/i18n'
import { TTSServiceInterface } from './TTSServiceInterface'
/**
* 免费在线TTS服务实现类
* 使用免费的在线TTS服务，不需要API密钥
*/
export class MsTTSService implements TTSServiceInterface {
private voice: string;
private outputFormat: string;
private voice: string
private outputFormat: string
/**
* 构造函数
@ -15,9 +16,9 @@ export class MsTTSService implements TTSServiceInterface {
* @param outputFormat 输出格式
*/
constructor(voice: string, outputFormat: string) {
this.voice = voice;
this.outputFormat = outputFormat;
console.log('初始化MsTTSService，语音:', voice, '输出格式:', outputFormat);
this.voice = voice
this.outputFormat = outputFormat
console.log('初始化MsTTSService，语音:', voice, '输出格式:', outputFormat)
}
/**
@ -26,7 +27,7 @@ export class MsTTSService implements TTSServiceInterface {
*/
private validateParams(): void {
if (!this.voice) {
throw new Error(i18n.t('settings.tts.error.no_mstts_voice'));
throw new Error(i18n.t('settings.tts.error.no_mstts_voice'))
}
}
@ -37,22 +38,22 @@ export class MsTTSService implements TTSServiceInterface {
*/
async synthesize(text: string): Promise<Blob> {
// 验证参数
this.validateParams();
this.validateParams()
try {
console.log('使用免费在线TTS生成语音，音色:', this.voice);
console.log('使用免费在线TTS生成语音，音色:', this.voice)
// 通过IPC调用主进程的MsTTSService
const outputPath = await window.api.msTTS.synthesize(text, this.voice, this.outputFormat);
const outputPath = await window.api.msTTS.synthesize(text, this.voice, this.outputFormat)
// 读取生成的音频文件
const audioData = await window.api.fs.read(outputPath);
const audioData = await window.api.fs.read(outputPath)
// 将Buffer转换为Blob
return new Blob([audioData], { type: 'audio/mp3' });
return new Blob([audioData], { type: 'audio/mp3' })
} catch (error: any) {
console.error('免费在线TTS语音合成失败:', error);
throw new Error(`免费在线TTS语音合成失败: ${error?.message || '未知错误'}`);
console.error('免费在线TTS语音合成失败:', error)
throw new Error(`免费在线TTS语音合成失败: ${error?.message || '未知错误'}`)
}
}
}

49
src/renderer/src/services/tts/OpenAITTSService.ts
View File

@ -1,14 +1,15 @@
import { TTSServiceInterface } from './TTSServiceInterface';
import i18n from '@renderer/i18n';
import i18n from '@renderer/i18n'
import { TTSServiceInterface } from './TTSServiceInterface'
/**
* OpenAI TTS服务实现类
*/
export class OpenAITTSService implements TTSServiceInterface {
private apiKey: string;
private apiUrl: string;
private voice: string;
private model: string;
private apiKey: string
private apiUrl: string
private voice: string
private model: string
/**
* 构造函数
@ -18,10 +19,10 @@ export class OpenAITTSService implements TTSServiceInterface {
* @param model 模型
*/
constructor(apiKey: string, apiUrl: string, voice: string, model: string) {
this.apiKey = apiKey;
this.apiUrl = apiUrl;
this.voice = voice;
this.model = model;
this.apiKey = apiKey
this.apiUrl = apiUrl
this.voice = voice
this.model = model
}
/**
@ -30,19 +31,19 @@ export class OpenAITTSService implements TTSServiceInterface {
*/
private validateParams(): void {
if (!this.apiKey) {
throw new Error(i18n.t('settings.tts.error.no_api_key'));
throw new Error(i18n.t('settings.tts.error.no_api_key'))
}
if (!this.apiUrl) {
throw new Error(i18n.t('settings.tts.error.no_api_url'));
throw new Error(i18n.t('settings.tts.error.no_api_url'))
}
if (!this.voice) {
throw new Error(i18n.t('settings.tts.error.no_voice'));
throw new Error(i18n.t('settings.tts.error.no_voice'))
}
if (!this.model) {
throw new Error(i18n.t('settings.tts.error.no_model'));
throw new Error(i18n.t('settings.tts.error.no_model'))
}
}
@ -53,24 +54,24 @@ export class OpenAITTSService implements TTSServiceInterface {
*/
async synthesize(text: string): Promise<Blob> {
// 验证参数
this.validateParams();
this.validateParams()
// 准备OpenAI TTS请求体
const requestBody: any = {
input: text
};
}
// 只有当模型和音色不为空时才添加到请求体中
if (this.model) {
requestBody.model = this.model;
requestBody.model = this.model
}
if (this.voice) {
requestBody.voice = this.voice;
requestBody.voice = this.voice
}
// 调用OpenAI TTS API
console.log('调用OpenAI TTS API，开始合成语音');
console.log('调用OpenAI TTS API，开始合成语音')
const response = await fetch(this.apiUrl, {
method: 'POST',
headers: {
@ -78,15 +79,15 @@ export class OpenAITTSService implements TTSServiceInterface {
Authorization: `Bearer ${this.apiKey}`
},
body: JSON.stringify(requestBody)
});
})
if (!response.ok) {
const errorData = await response.json();
throw new Error(errorData.error?.message || 'OpenAI语音合成失败');
const errorData = await response.json()
throw new Error(errorData.error?.message || 'OpenAI语音合成失败')
}
// 获取音频数据
console.log('获取到OpenAI TTS响应，开始处理音频数据');
return await response.blob();
console.log('获取到OpenAI TTS响应，开始处理音频数据')
return await response.blob()
}
}

61
src/renderer/src/services/tts/SiliconflowTTSService.ts
View File

@ -1,16 +1,17 @@
import { TTSServiceInterface } from './TTSServiceInterface';
import i18n from '@renderer/i18n';
import i18n from '@renderer/i18n'
import { TTSServiceInterface } from './TTSServiceInterface'
/**
* 硅基流动TTS服务实现类
*/
export class SiliconflowTTSService implements TTSServiceInterface {
private apiKey: string;
private apiUrl: string;
private voice: string;
private model: string;
private responseFormat: string;
private speed: number;
private apiKey: string
private apiUrl: string
private voice: string
private model: string
private responseFormat: string
private speed: number
/**
* 构造函数
@ -29,12 +30,12 @@ export class SiliconflowTTSService implements TTSServiceInterface {
responseFormat: string = 'mp3',
speed: number = 1.0
) {
this.apiKey = apiKey;
this.apiUrl = apiUrl || 'https://api.siliconflow.cn/v1/audio/speech';
this.voice = voice;
this.model = model;
this.responseFormat = responseFormat;
this.speed = speed;
this.apiKey = apiKey
this.apiUrl = apiUrl || 'https://api.siliconflow.cn/v1/audio/speech'
this.voice = voice
this.model = model
this.responseFormat = responseFormat
this.speed = speed
}
/**
@ -43,15 +44,15 @@ export class SiliconflowTTSService implements TTSServiceInterface {
*/
private validateParams(): void {
if (!this.apiKey) {
throw new Error(i18n.t('settings.tts.error.no_api_key'));
throw new Error(i18n.t('settings.tts.error.no_api_key'))
}
if (!this.voice) {
throw new Error(i18n.t('settings.tts.error.no_voice'));
throw new Error(i18n.t('settings.tts.error.no_voice'))
}
if (!this.model) {
throw new Error(i18n.t('settings.tts.error.no_model'));
throw new Error(i18n.t('settings.tts.error.no_model'))
}
}
@ -62,7 +63,7 @@ export class SiliconflowTTSService implements TTSServiceInterface {
*/
async synthesize(text: string): Promise<Blob> {
// 验证参数
this.validateParams();
this.validateParams()
// 准备硅基流动TTS请求体
const requestBody: any = {
@ -73,44 +74,44 @@ export class SiliconflowTTSService implements TTSServiceInterface {
response_format: this.responseFormat,
stream: false,
speed: this.speed
};
}
console.log('硅基流动TTS请求参数:', {
model: this.model,
voice: this.voice,
response_format: 'mp3',
speed: this.speed
});
})
// 调用硅基流动TTS API
console.log('调用硅基流动TTS API，开始合成语音');
console.log('调用硅基流动TTS API，开始合成语音')
const response = await fetch(this.apiUrl, {
method: 'POST',
headers: {
'Content-Type': 'application/json',
'Authorization': `Bearer ${this.apiKey}`
Authorization: `Bearer ${this.apiKey}`
},
body: JSON.stringify(requestBody)
});
})
if (!response.ok) {
let errorMessage = '硅基流动语音合成失败';
let errorMessage = '硅基流动语音合成失败'
try {
const errorData = await response.json();
errorMessage = errorData.error?.message || errorMessage;
const errorData = await response.json()
errorMessage = errorData.error?.message || errorMessage
} catch (e) {
// 如果无法解析JSON，使用默认错误消息
}
throw new Error(errorMessage);
throw new Error(errorMessage)
}
// 获取音频数据
console.log('获取到硅基流动TTS响应，开始处理音频数据');
console.log('获取到硅基流动TTS响应，开始处理音频数据')
// 获取原始Blob
const originalBlob = await response.blob();
const originalBlob = await response.blob()
// 创建一个新的Blob，并指定正确的MIME类型
return new Blob([originalBlob], { type: 'audio/mpeg' });
return new Blob([originalBlob], { type: 'audio/mpeg' })
}
}

161
src/renderer/src/services/tts/TTSService.ts
View File

@ -1,21 +1,22 @@
import store from '@renderer/store';
import i18n from '@renderer/i18n';
import { TTSServiceFactory } from './TTSServiceFactory';
import { TTSTextFilter } from './TTSTextFilter';
import { Message } from '@renderer/types';
import i18n from '@renderer/i18n'
import store from '@renderer/store'
import { Message } from '@renderer/types'
import { TTSServiceFactory } from './TTSServiceFactory'
import { TTSTextFilter } from './TTSTextFilter'
/**
* TTS服务类
* 用于处理文本到语音的转换
*/
export class TTSService {
private static instance: TTSService;
private audioElement: HTMLAudioElement | null = null;
private isPlaying = false;
private static instance: TTSService
private audioElement: HTMLAudioElement | null = null
private isPlaying = false
// 错误消息节流控制
private lastErrorTime = 0;
private errorThrottleTime = 2000; // 2秒内不重复显示相同错误
private lastErrorTime = 0
private errorThrottleTime = 2000 // 2秒内不重复显示相同错误
/**
* 获取单例实例
@ -25,8 +26,8 @@ export class TTSService {
// 每次调用时强制重新创建实例，确保使用最新的设置
// 注意：这会导致每次调用时都创建新的音频元素，可能会有内存泄漏风险
// 但在当前情况下，这是解决TTS服务类型切换问题的最简单方法
TTSService.instance = new TTSService();
return TTSService.instance;
TTSService.instance = new TTSService()
return TTSService.instance
}
/**
@ -34,21 +35,21 @@ export class TTSService {
*/
private constructor() {
// 创建音频元素
this.audioElement = document.createElement('audio');
this.audioElement.style.display = 'none';
document.body.appendChild(this.audioElement);
this.audioElement = document.createElement('audio')
this.audioElement.style.display = 'none'
document.body.appendChild(this.audioElement)
// 监听音频播放结束事件
this.audioElement.addEventListener('ended', () => {
this.isPlaying = false;
console.log('TTS播放结束');
});
this.isPlaying = false
console.log('TTS播放结束')
})
// 监听浏览器TTS直接播放结束的自定义事件
document.addEventListener('edgeTTSComplete', () => {
console.log('收到浏览器TTS直接播放结束事件');
this.isPlaying = false;
});
console.log('收到浏览器TTS直接播放结束事件')
this.isPlaying = false
})
}
/**
@ -58,21 +59,21 @@ export class TTSService {
*/
public async speakFromMessage(message: Message): Promise<boolean> {
// 获取最新的TTS过滤选项
const settings = store.getState().settings;
const settings = store.getState().settings
const ttsFilterOptions = settings.ttsFilterOptions || {
filterThinkingProcess: true,
filterMarkdown: true,
filterCodeBlocks: true,
filterHtmlTags: true,
maxTextLength: 4000
};
}
// 应用过滤
const filteredText = TTSTextFilter.filterText(message.content, ttsFilterOptions);
console.log('TTS过滤前文本长度:', message.content.length, '过滤后:', filteredText.length);
const filteredText = TTSTextFilter.filterText(message.content, ttsFilterOptions)
console.log('TTS过滤前文本长度:', message.content.length, '过滤后:', filteredText.length)
// 播放过滤后的文本
return this.speak(filteredText);
return this.speak(filteredText)
}
/**
@ -83,30 +84,30 @@ export class TTSService {
public async speak(text: string): Promise<boolean> {
try {
// 检查TTS是否启用
const settings = store.getState().settings;
const ttsEnabled = settings.ttsEnabled;
const settings = store.getState().settings
const ttsEnabled = settings.ttsEnabled
if (!ttsEnabled) {
this.showErrorMessage(i18n.t('settings.tts.error.not_enabled'));
return false;
this.showErrorMessage(i18n.t('settings.tts.error.not_enabled'))
return false
}
// 如果正在播放，先停止
if (this.isPlaying) {
this.stop();
this.stop()
}
// 确保文本不为空
if (!text || text.trim() === '') {
this.showErrorMessage(i18n.t('settings.tts.error.empty_text'));
return false;
this.showErrorMessage(i18n.t('settings.tts.error.empty_text'))
return false
}
// 获取最新的设置
// 强制刷新状态对象，确保获取最新的设置
const latestSettings = store.getState().settings;
const serviceType = latestSettings.ttsServiceType || 'openai';
console.log('使用的TTS服务类型:', serviceType);
const latestSettings = store.getState().settings
const serviceType = latestSettings.ttsServiceType || 'openai'
console.log('使用的TTS服务类型:', serviceType)
console.log('当前TTS设置详情:', {
ttsServiceType: serviceType,
ttsEdgeVoice: latestSettings.ttsEdgeVoice,
@ -115,18 +116,18 @@ export class TTSService {
ttsSiliconflowModel: latestSettings.ttsSiliconflowModel,
ttsSiliconflowResponseFormat: latestSettings.ttsSiliconflowResponseFormat,
ttsSiliconflowSpeed: latestSettings.ttsSiliconflowSpeed
});
})
try {
// 使用工厂创建TTS服务
const ttsService = TTSServiceFactory.createService(serviceType, latestSettings);
const ttsService = TTSServiceFactory.createService(serviceType, latestSettings)
// 合成语音
const audioBlob = await ttsService.synthesize(text);
const audioBlob = await ttsService.synthesize(text)
// 播放音频
if (audioBlob) {
const audioUrl = URL.createObjectURL(audioBlob);
const audioUrl = URL.createObjectURL(audioBlob)
if (this.audioElement) {
// 打印音频Blob信息，帮助调试
@ -134,36 +135,36 @@ export class TTSService {
size: audioBlob.size,
type: audioBlob.type,
serviceType: serviceType
});
})
this.audioElement.src = audioUrl;
this.audioElement.src = audioUrl
this.audioElement.play().catch((error) => {
// 检查是否是浏览器TTS直接播放的情况
// 如果是浏览器TTS且音频大小很小，则不显示错误消息
const isEdgeTTS = serviceType === 'edge';
const isSmallBlob = audioBlob.size < 100; // 小于100字节的音频文件可能是我们的静音文件
const isEdgeTTS = serviceType === 'edge'
const isSmallBlob = audioBlob.size < 100 // 小于100字节的音频文件可能是我们的静音文件
if (isEdgeTTS && isSmallBlob) {
console.log('浏览器TTS直接播放中，忽略音频元素错误');
console.log('浏览器TTS直接播放中，忽略音频元素错误')
} else {
console.error('播放TTS音频失败:', error);
console.error('音频URL:', audioUrl);
console.error('音频Blob类型:', audioBlob.type);
console.error('音频Blob大小:', audioBlob.size);
this.showErrorMessage(i18n.t('settings.tts.error.play_failed'));
console.error('播放TTS音频失败:', error)
console.error('音频URL:', audioUrl)
console.error('音频Blob类型:', audioBlob.type)
console.error('音频Blob大小:', audioBlob.size)
this.showErrorMessage(i18n.t('settings.tts.error.play_failed'))
}
});
})
this.isPlaying = true;
console.log('开始播放TTS音频');
this.isPlaying = true
console.log('开始播放TTS音频')
// 释放URL对象
this.audioElement.onended = () => {
URL.revokeObjectURL(audioUrl);
URL.revokeObjectURL(audioUrl)
// 检查是否是浏览器TTS直接播放的情况
const isEdgeTTS = serviceType === 'edge';
const isSmallBlob = audioBlob.size < 100;
const isEdgeTTS = serviceType === 'edge'
const isSmallBlob = audioBlob.size < 100
// 如果是浏览器TTS直接播放，则等待当前语音合成结束
if (isEdgeTTS && isSmallBlob) {
@ -171,33 +172,33 @@ export class TTSService {
// 如果还在播放，则不重置播放状态
// 注意：这里我们无法直接访问 EdgeTTSService 中的 currentUtterance
// 所以我们使用定时器来检查语音合成是否完成
console.log('浏览器TTS直接播放中，等待语音合成结束');
console.log('浏览器TTS直接播放中，等待语音合成结束')
// 保持播放状态，直到语音合成结束
// 使用定时器来检查语音合成是否完成
// 大多数语音合成应该在几秒内完成
setTimeout(() => {
this.isPlaying = false;
console.log('浏览器TTS直接播放完成');
}, 10000); // 10秒后自动重置状态
this.isPlaying = false
console.log('浏览器TTS直接播放完成')
}, 10000) // 10秒后自动重置状态
} else {
this.isPlaying = false;
this.isPlaying = false
}
};
}
return true;
return true
}
}
return false;
return false
} catch (error: any) {
console.error('TTS合成失败:', error);
this.showErrorMessage(error?.message || i18n.t('settings.tts.error.synthesis_failed'));
return false;
console.error('TTS合成失败:', error)
this.showErrorMessage(error?.message || i18n.t('settings.tts.error.synthesis_failed'))
return false
}
} catch (error) {
console.error('TTS播放失败:', error);
this.showErrorMessage(i18n.t('settings.tts.error.general'));
return false;
console.error('TTS播放失败:', error)
this.showErrorMessage(i18n.t('settings.tts.error.general'))
return false
}
}
@ -206,10 +207,10 @@ export class TTSService {
*/
public stop(): void {
if (this.audioElement && this.isPlaying) {
this.audioElement.pause();
this.audioElement.currentTime = 0;
this.isPlaying = false;
console.log('停止TTS播放');
this.audioElement.pause()
this.audioElement.currentTime = 0
this.isPlaying = false
console.log('停止TTS播放')
}
}
@ -218,7 +219,7 @@ export class TTSService {
* @returns 是否正在播放
*/
public isCurrentlyPlaying(): boolean {
return this.isPlaying;
return this.isPlaying
}
/**
@ -226,15 +227,15 @@ export class TTSService {
* @param message 错误消息
*/
private showErrorMessage(message: string): void {
const now = Date.now();
const now = Date.now()
// 如果距离上次错误消息的时间小于节流时间，则不显示
if (now - this.lastErrorTime < this.errorThrottleTime) {
console.log('错误消息被节流：', message);
return;
console.log('错误消息被节流：', message)
return
}
// 更新上次错误消息时间
this.lastErrorTime = now;
window.message.error({ content: message, key: 'tts-error' });
this.lastErrorTime = now
window.message.error({ content: message, key: 'tts-error' })
}
}

45
src/renderer/src/services/tts/TTSServiceFactory.ts
View File

@ -1,9 +1,10 @@
import { TTSServiceInterface } from './TTSServiceInterface';
import { OpenAITTSService } from './OpenAITTSService';
import { EdgeTTSService } from './EdgeTTSService';
import { SiliconflowTTSService } from './SiliconflowTTSService';
import { MsTTSService } from './MsTTSService';
import i18n from '@renderer/i18n';
import i18n from '@renderer/i18n'
import { EdgeTTSService } from './EdgeTTSService'
import { MsTTSService } from './MsTTSService'
import { OpenAITTSService } from './OpenAITTSService'
import { SiliconflowTTSService } from './SiliconflowTTSService'
import { TTSServiceInterface } from './TTSServiceInterface'
/**
* TTS服务工厂类
@ -17,24 +18,19 @@ export class TTSServiceFactory {
* @returns TTS服务实例
*/
static createService(serviceType: string, settings: any): TTSServiceInterface {
console.log('创建TTS服务实例，类型:', serviceType);
console.log('创建TTS服务实例，类型:', serviceType)
switch (serviceType) {
case 'openai':
console.log('创建OpenAI TTS服务实例');
return new OpenAITTSService(
settings.ttsApiKey,
settings.ttsApiUrl,
settings.ttsVoice,
settings.ttsModel
);
console.log('创建OpenAI TTS服务实例')
return new OpenAITTSService(settings.ttsApiKey, settings.ttsApiUrl, settings.ttsVoice, settings.ttsModel)
case 'edge':
console.log('创建Edge TTS服务实例');
return new EdgeTTSService(settings.ttsEdgeVoice);
console.log('创建Edge TTS服务实例')
return new EdgeTTSService(settings.ttsEdgeVoice)
case 'siliconflow':
console.log('创建硅基流动 TTS服务实例');
console.log('创建硅基流动 TTS服务实例')
console.log('硅基流动TTS设置:', {
apiKey: settings.ttsSiliconflowApiKey ? '已设置' : '未设置',
apiUrl: settings.ttsSiliconflowApiUrl,
@ -42,7 +38,7 @@ export class TTSServiceFactory {
model: settings.ttsSiliconflowModel,
responseFormat: settings.ttsSiliconflowResponseFormat,
speed: settings.ttsSiliconflowSpeed
});
})
return new SiliconflowTTSService(
settings.ttsSiliconflowApiKey,
settings.ttsSiliconflowApiUrl,
@ -50,21 +46,18 @@ export class TTSServiceFactory {
settings.ttsSiliconflowModel,
settings.ttsSiliconflowResponseFormat,
settings.ttsSiliconflowSpeed
);
)
case 'mstts':
console.log('创建免费在线TTS服务实例');
console.log('创建免费在线TTS服务实例')
console.log('免费在线TTS设置:', {
voice: settings.ttsMsVoice,
outputFormat: settings.ttsMsOutputFormat
});
return new MsTTSService(
settings.ttsMsVoice,
settings.ttsMsOutputFormat
);
})
return new MsTTSService(settings.ttsMsVoice, settings.ttsMsOutputFormat)
default:
throw new Error(i18n.t('settings.tts.error.unsupported_service_type', { serviceType }));
throw new Error(i18n.t('settings.tts.error.unsupported_service_type', { serviceType }))
}
}
}

2
src/renderer/src/services/tts/TTSServiceInterface.ts
View File

@ -8,5 +8,5 @@ export interface TTSServiceInterface {
* @param text 要合成的文本
* @returns 返回音频Blob对象的Promise
*/
synthesize(text: string): Promise<Blob>;
synthesize(text: string): Promise<Blob>
}

108
src/renderer/src/services/tts/TTSTextFilter.ts
View File

@ -12,43 +12,43 @@ export class TTSTextFilter {
public static filterText(
text: string,
options: {
filterThinkingProcess: boolean;
filterMarkdown: boolean;
filterCodeBlocks: boolean;
filterHtmlTags: boolean;
maxTextLength: number;
filterThinkingProcess: boolean
filterMarkdown: boolean
filterCodeBlocks: boolean
filterHtmlTags: boolean
maxTextLength: number
}
): string {
if (!text) return '';
if (!text) return ''
let filteredText = text;
let filteredText = text
// 过滤思考过程
if (options.filterThinkingProcess) {
filteredText = this.filterThinkingProcess(filteredText);
filteredText = this.filterThinkingProcess(filteredText)
}
// 过滤Markdown标记
if (options.filterMarkdown) {
filteredText = this.filterMarkdown(filteredText);
filteredText = this.filterMarkdown(filteredText)
}
// 过滤代码块
if (options.filterCodeBlocks) {
filteredText = this.filterCodeBlocks(filteredText);
filteredText = this.filterCodeBlocks(filteredText)
}
// 过滤HTML标签
if (options.filterHtmlTags) {
filteredText = this.filterHtmlTags(filteredText);
filteredText = this.filterHtmlTags(filteredText)
}
// 限制文本长度
if (options.maxTextLength > 0 && filteredText.length > options.maxTextLength) {
filteredText = filteredText.substring(0, options.maxTextLength);
filteredText = filteredText.substring(0, options.maxTextLength)
}
return filteredText.trim();
return filteredText.trim()
}
/**
@ -58,27 +58,27 @@ export class TTSTextFilter {
*/
private static filterThinkingProcess(text: string): string {
// 过滤<think>标签内容
text = text.replace(/<think>[\s\S]*?<\/think>/g, '');
text = text.replace(/<think>[\s\S]*?<\/think>/g, '')
// 过滤未闭合的<think>标签
if (text.includes('<think>')) {
const parts = text.split('<think>');
text = parts[0];
const parts = text.split('<think>')
text = parts[0]
}
// 过滤思考过程部分（###Thinking和###Response格式）
const thinkingMatch = text.match(/###\s*Thinking[\s\S]*?(?=###\s*Response|$)/);
const thinkingMatch = text.match(/###\s*Thinking[\s\S]*?(?=###\s*Response|$)/)
if (thinkingMatch) {
text = text.replace(thinkingMatch[0], '');
text = text.replace(thinkingMatch[0], '')
}
// 如果有Response部分，只保留Response部分
const responseMatch = text.match(/###\s*Response\s*([\s\S]*?)(?=###|$)/);
const responseMatch = text.match(/###\s*Response\s*([\s\S]*?)(?=###|$)/)
if (responseMatch) {
text = responseMatch[1];
text = responseMatch[1]
}
return text;
return text
}
/**
@ -88,29 +88,29 @@ export class TTSTextFilter {
*/
private static filterMarkdown(text: string): string {
// 过滤标题标记
text = text.replace(/#{1,6}\s+/g, '');
text = text.replace(/#{1,6}\s+/g, '')
// 过滤粗体和斜体标记
text = text.replace(/(\*\*|__)(.*?)\1/g, '$2');
text = text.replace(/(\*|_)(.*?)\1/g, '$2');
text = text.replace(/(\*\*|__)(.*?)\1/g, '$2')
text = text.replace(/(\*|_)(.*?)\1/g, '$2')
// 过滤链接
text = text.replace(/\[([^\]]+)\]\(([^)]+)\)/g, '$1');
text = text.replace(/\[([^\]]+)\]\(([^)]+)\)/g, '$1')
// 过滤图片
text = text.replace(/!\[([^\]]*)\]\(([^)]+)\)/g, '');
text = text.replace(/!\[([^\]]*)\]\(([^)]+)\)/g, '')
// 过滤引用
text = text.replace(/^\s*>\s+/gm, '');
text = text.replace(/^\s*>\s+/gm, '')
// 过滤水平线
text = text.replace(/^\s*[-*_]{3,}\s*$/gm, '');
text = text.replace(/^\s*[-*_]{3,}\s*$/gm, '')
// 过滤列表标记
text = text.replace(/^\s*[-*+]\s+/gm, '');
text = text.replace(/^\s*\d+\.\s+/gm, '');
return text;
text = text.replace(/^\s*[-*+]\s+/gm, '')
text = text.replace(/^\s*\d+\.\s+/gm, '')
return text
}
/**
@ -120,15 +120,15 @@ export class TTSTextFilter {
*/
private static filterCodeBlocks(text: string): string {
// 过滤围栏式代码块
text = text.replace(/```[\s\S]*?```/g, '');
text = text.replace(/```[\s\S]*?```/g, '')
// 过滤缩进式代码块
text = text.replace(/(?:^|\n)( {4}|\t).*(?:\n|$)/g, '\n');
text = text.replace(/(?:^|\n)( {4}|\t).*(?:\n|$)/g, '\n')
// 过滤行内代码
text = text.replace(/`([^`]+)`/g, '$1');
return text;
text = text.replace(/`([^`]+)`/g, '$1')
return text
}
/**
@ -138,11 +138,11 @@ export class TTSTextFilter {
*/
private static filterHtmlTags(text: string): string {
// 过滤HTML标签
text = text.replace(/<[^>]*>/g, '');
text = text.replace(/<[^>]*>/g, '')
// 过滤HTML实体
text = text.replace(/&[a-zA-Z0-9#]+;/g, ' ');
return text;
text = text.replace(/&[a-zA-Z0-9#]+;/g, ' ')
return text
}
}

14
src/renderer/src/services/tts/index.ts
View File

@ -1,7 +1,7 @@
export * from './TTSService';
export * from './TTSServiceInterface';
export * from './TTSServiceFactory';
export * from './OpenAITTSService';
export * from './EdgeTTSService';
export * from './SiliconflowTTSService';
export * from './MsTTSService';
export * from './EdgeTTSService'
export * from './MsTTSService'
export * from './OpenAITTSService'
export * from './SiliconflowTTSService'
export * from './TTSService'
export * from './TTSServiceFactory'
export * from './TTSServiceInterface'