deepvoice_client/local_plugins/volcano_speech/lib/volcano_speech.dart

import 'dart:async';
import 'package:flutter/services.dart';

/// ASR 事件类型
enum AsrEventType {
  /// 会话开始
  sessionStarted,
  
  /// 会话结束
  sessionStopped,
  
  /// 正在识别（中间结果）
  recognizing,
  
  /// 识别结果（最终结果）
  result,
  
  /// 音量变化
  volumeChanged,
  
  /// 错误
  error
}

/// TTS 工作模式
enum TtsWorkMode {
  /// 在线合成
  online,
  
  /// 离线合成
  offline,
  
  /// 同时在线离线
  both,
  
  /// 先在线再离线（网络不好时自动切换）
  alternate,
  
  /// 文件模式
  file
}

/// TTS 文本类型
enum TtsTextType {
  /// 纯文本
  plain,
  
  /// SSML格式
  ssml
}

/// 协议类型
enum ProtocolType {
  /// 默认协议
  defaultProtocol,
  
  /// Seed协议（用于大模型）
  seed
}

/// TTS 播放进度事件
class TtsProgressEvent {
  /// 播放进度 0.0-1.0
  final double progress;
  
  /// 请求ID
  final String reqId;
  
  const TtsProgressEvent({
    required this.progress,
    required this.reqId,
  });
  
  @override
  String toString() {
    return 'TtsProgressEvent{progress: $progress, reqId: $reqId}';
  }
}

/// ASR 事件
class AsrEvent {
  /// 事件类型
  final AsrEventType type;
  
  /// 识别文本（仅在recognizing和result类型时有效）
  final String? text;
  
  /// 识别语言（仅在recognizing和result类型时有效）
  final String? language;
  
  /// 音量值（仅在volumeChanged类型时有效）
  final int? volume;
  
  /// 错误信息（仅在error类型时有效）
  final String? errorMessage;

  const AsrEvent({
    required this.type,
    this.text,
    this.language,
    this.volume,
    this.errorMessage,
  });

  factory AsrEvent.fromMap(Map<dynamic, dynamic> map) {
    final typeStr = map['type'] as String;
    
    AsrEventType type;
    switch (typeStr) {
      case 'sessionStarted':
        type = AsrEventType.sessionStarted;
        break;
      case 'sessionStopped':
        type = AsrEventType.sessionStopped;
        break;
      case 'recognizing':
        type = AsrEventType.recognizing;
        break;
      case 'result':
        type = AsrEventType.result;
        break;
      case 'volumeChanged':
        type = AsrEventType.volumeChanged;
        break;
      case 'error':
        type = AsrEventType.error;
        break;
      default:
        throw ArgumentError('未知的事件类型: $typeStr');
    }
    
    return AsrEvent(
      type: type,
      text: map['text'] as String?,
      language: map['language'] as String?,
      volume: map['volume'] as int?,
      errorMessage: map['message'] as String?,
    );
  }

  @override
  String toString() {
    return 'AsrEvent{type: $type, text: $text, language: $language, volume: $volume, errorMessage: $errorMessage}';
  }
}

/// 火山引擎语音服务插件
class VolcanoSpeech {
  static final VolcanoSpeechAsr asr = VolcanoSpeechAsr._();
  static final VolcanoSpeechTts tts = VolcanoSpeechTts._();
  
  /// 释放资源
  static Future<void> dispose() async {
    await asr.dispose();
    await tts.dispose();
  }
}

/// 火山引擎语音识别服务
class VolcanoSpeechAsr {
  static const MethodChannel _channel = MethodChannel('volcano_speech/asr');
  static const EventChannel _eventChannel = EventChannel('volcano_speech/asr_events');
  
  /// ASR事件流控制器
  static final StreamController<AsrEvent> _eventStreamController = StreamController<AsrEvent>.broadcast();
  
  /// ASR事件流
  Stream<AsrEvent> get events => _eventStreamController.stream;
  
  /// 是否已初始化事件监听
  bool _eventListenerInitialized = false;
  
  VolcanoSpeechAsr._() {
    _initEventListener();
  }
  
  /// 初始化ASR事件监听
  void _initEventListener() {
    if (_eventListenerInitialized) return;
    
    _eventChannel.receiveBroadcastStream().listen((dynamic event) {
      if (event is Map) {
        _eventStreamController.add(AsrEvent.fromMap(event));
      }
    });
    
    _eventListenerInitialized = true;
  }
  
  /// 获取平台版本信息
  Future<String?> getPlatformVersion() async {
    return await _channel.invokeMethod('getPlatformVersion');
  }
  
  /// 初始化语音识别引擎
  /// 
  /// [appId] 火山引擎AppID
  /// [apiKey] 火山引擎ApiKey
  /// [supportedLanguages] 支持的语言列表
  /// [useBigModel] 是否使用大模型识别
  /// [resourceId] 大模型资源ID（仅当useBigModel为true时有效）
  Future<bool> initialize({
    required String appId,
    required String apiKey,
    List<String> supportedLanguages = const ['zh-CN'],
    bool useBigModel = false,
    String resourceId = '',
  }) async {
    return await _channel.invokeMethod('initialize', {
      'appId': appId,
      'apiKey': apiKey,
      'supportedLanguages': supportedLanguages,
      'useBigModel': useBigModel,
      'resourceId': resourceId,
    }) ?? false;
  }
  
  /// 设置识别语言
  /// 
  /// [language] 语言代码，例如 zh-CN、en-US
  Future<bool> setLanguage(String language) async {
    return await _channel.invokeMethod('setLanguage', {
      'language': language,
    }) ?? false;
  }
  
  /// 设置热词
  /// 
  /// [hotWords] 热词JSON字符串，例如 {"hotwords":[{"word":"快速入门","scale":2.0}]}
  Future<bool> setHotWords(String hotWords) async {
    return await _channel.invokeMethod('setHotWords', {
      'hotWords': hotWords,
    }) ?? false;
  }
  
  /// 设置ASR请求参数
  /// 
  /// [params] 请求参数JSON字符串
  Future<bool> setRequestParams(String params) async {
    return await _channel.invokeMethod('setRequestParams', {
      'params': params,
    }) ?? false;
  }
  
  /// 启用语音停顿、分句、分词信息输出
  /// 
  /// [enable] 是否启用
  Future<bool> setShowUtterances(bool enable) async {
    return await _channel.invokeMethod('setShowUtterances', {
      'enable': enable,
    }) ?? false;
  }
  
  /// 一次性识别（直到说话结束）
  /// 
  /// 返回识别结果文本和语言
  Future<Map<String, String>> recognizeOnce() async {
    final result = await _channel.invokeMethod('recognizeOnce');
    return {
      'text': result['text'] ?? '',
      'language': result['language'] ?? 'zh-CN',
    };
  }
  
  /// 开始连续识别
  /// 
  /// 通过[events]流监听识别结果
  Future<bool> startContinuousRecognition() async {
    return await _channel.invokeMethod('startContinuousRecognition') ?? false;
  }
  
  /// 停止连续识别
  Future<bool> stopContinuousRecognition() async {
    return await _channel.invokeMethod('stopContinuousRecognition') ?? false;
  }
  
  /// 检查是否正在连续识别
  Future<bool> isContinuousRecognitionActive() async {
    return await _channel.invokeMethod('isContinuousRecognitionActive') ?? false;
  }
  
  /// 开始长按识别（按下开始，抬起结束）
  /// 
  /// 通过[events]流监听识别结果
  Future<bool> startListening() async {
    return await _channel.invokeMethod('startListening') ?? false;
  }
  
  /// 停止长按识别
  Future<bool> stopListening() async {
    return await _channel.invokeMethod('stopListening') ?? false;
  }
  
  /// 释放ASR资源
  Future<bool> dispose() async {
    return await _channel.invokeMethod('dispose') ?? false;
  }
}

/// 火山引擎语音合成服务
class VolcanoSpeechTts {
  static const MethodChannel _channel = MethodChannel('volcano_speech/tts');
  static const EventChannel _eventChannel = EventChannel('volcano_speech/tts_events');
  
  /// TTS进度事件流控制器
  static final StreamController<TtsProgressEvent> _progressStreamController = StreamController<TtsProgressEvent>.broadcast();
  
  /// TTS进度事件流
  Stream<TtsProgressEvent> get progressEvents => _progressStreamController.stream;
  
  /// 是否已初始化事件监听
  bool _eventListenerInitialized = false;
  
  VolcanoSpeechTts._() {
    _initEventListener();
  }
  
  /// 初始化TTS事件监听
  void _initEventListener() {
    if (_eventListenerInitialized) return;
    
    _eventChannel.receiveBroadcastStream().listen((dynamic event) {
      if (event is Map) {
        final progress = event['progress'] as double?;
        final reqId = event['reqId'] as String?;
        
        if (progress != null && reqId != null) {
          _progressStreamController.add(TtsProgressEvent(
            progress: progress,
            reqId: reqId,
          ));
        }
      }
    });
    
    _eventListenerInitialized = true;
  }
  
  /// 获取平台版本信息
  Future<String?> getPlatformVersion() async {
    return await _channel.invokeMethod('getPlatformVersion');
  }
  
  /// 启用大模型TTS
  /// 
  /// [enable] 是否启用大模型TTS
  /// [resourceId] 资源ID
  Future<bool> enableBigModelTts({
    required bool enable,
    String resourceId = '',
  }) async {
    return await _channel.invokeMethod('enableBigModelTts', {
      'enable': enable,
      'resourceId': resourceId,
    }) ?? false;
  }
  
  /// 初始化语音合成引擎
  /// 
  /// [appId] 火山引擎AppID
  /// [apiKey] 火山引擎ApiKey
  Future<bool> initialize({
    required String appId,
    required String apiKey,
  }) async {
    return await _channel.invokeMethod('initialize', {
      'appId': appId,
      'apiKey': apiKey,
    }) ?? false;
  }
  
  /// 设置音色
  /// 
  /// [voiceName] 音色名称
  /// [voiceType] 音色类型，默认 qingxin
  Future<bool> setVoice({
    required String voiceName,
    String voiceType = 'qingxin',
  }) async {
    return await _channel.invokeMethod('setVoice', {
      'voiceName': voiceName,
      'voiceType': voiceType,
    }) ?? false;
  }
  
  /// 设置离线音色
  /// 
  /// [voiceName] 音色名称
  /// [voiceType] 音色类型，默认 qingxin
  Future<bool> setOfflineVoice({
    required String voiceName,
    String voiceType = 'qingxin',
  }) async {
    return await _channel.invokeMethod('setOfflineVoice', {
      'voiceName': voiceName,
      'voiceType': voiceType,
    }) ?? false;
  }
  
  /// 设置大模型声音ID
  /// 
  /// [voiceId] 声音ID
  Future<bool> setBigModelVoiceId(String voiceId) async {
    return await _channel.invokeMethod('setBigModelVoiceId', {
      'voiceId': voiceId,
    }) ?? false;
  }
  
  /// 设置大模型请求参数
  /// 
  /// [params] 请求参数，JSON字符串
  Future<bool> setBigModelRequestParams(String params) async {
    return await _channel.invokeMethod('setBigModelRequestParams', {
      'params': params,
    }) ?? false;
  }
  
  /// 设置工作模式
  /// 
  /// [mode] 工作模式
  Future<bool> setWorkMode(TtsWorkMode mode) async {
    String modeStr;
    switch (mode) {
      case TtsWorkMode.online:
        modeStr = 'online';
        break;
      case TtsWorkMode.offline:
        modeStr = 'offline';
        break;
      case TtsWorkMode.both:
        modeStr = 'both';
        break;
      case TtsWorkMode.alternate:
        modeStr = 'alternate';
        break;
      case TtsWorkMode.file:
        modeStr = 'file';
        break;
    }
    
    return await _channel.invokeMethod('setWorkMode', {
      'mode': modeStr,
    }) ?? false;
  }
  
  /// 设置文本类型
  /// 
  /// [type] 文本类型，plain或ssml
  Future<bool> setTextType(TtsTextType type) async {
    String typeStr;
    switch (type) {
      case TtsTextType.plain:
        typeStr = 'plain';
        break;
      case TtsTextType.ssml:
        typeStr = 'ssml';
        break;
    }
    
    return await _channel.invokeMethod('setTextType', {
      'type': typeStr,
    }) ?? false;
  }
  
  /// 设置是否启用缓存
  /// 
  /// [enable] 是否启用
  Future<bool> setEnableCache(bool enable) async {
    return await _channel.invokeMethod('setEnableCache', {
      'enable': enable,
    }) ?? false;
  }
  
  /// 设置情感
  /// 
  /// [emotion] 情感，例如 neutral、happy、angry、sad等
  Future<bool> setEmotion(String emotion) async {
    return await _channel.invokeMethod('setEmotion', {
      'emotion': emotion,
    }) ?? false;
  }
  
  /// 设置是否启用情感预测
  /// 
  /// [enable] 是否启用
  Future<bool> setEnableEmotionPredict(bool enable) async {
    return await _channel.invokeMethod('setEnableEmotionPredict', {
      'enable': enable,
    }) ?? false;
  }
  
  /// 设置是否启用声音克隆
  /// 
  /// [enable] 是否启用
  /// [backendCluster] 后端集群
  Future<bool> setEnableVoiceClone({
    required bool enable,
    String backendCluster = '',
  }) async {
    return await _channel.invokeMethod('setEnableVoiceClone', {
      'enable': enable,
      'backendCluster': backendCluster,
    }) ?? false;
  }
  
  /// 设置是否启用回声消除
  /// 
  /// [enable] 是否启用
  Future<bool> setEnableAEC(bool enable) async {
    return await _channel.invokeMethod('setEnableAEC', {
      'enable': enable,
    }) ?? false;
  }
  
  /// 下载离线资源
  /// 
  /// [voiceTypes] 音色类型列表
  /// [languages] 语言列表
  Future<bool> downloadOfflineResource({
    List<String> voiceTypes = const ['qingxin'],
    List<String> languages = const ['zh-CN'],
  }) async {
    return await _channel.invokeMethod('downloadOfflineResource', {
      'voiceTypes': voiceTypes,
      'languages': languages,
    }) ?? false;
  }
  
  /// 设置语音参数
  /// 
  /// [rate] 语速 -500~500
  /// [volume] 音量 0~100
  /// [pitch] 音调 -500~500
  /// [silenceDuration] 静音时长，毫秒
  Future<bool> setSpeechParams({
    int rate = 0,
    int volume = 100,
    int pitch = 0,
    int silenceDuration = 0,
  }) async {
    return await _channel.invokeMethod('setSpeechParams', {
      'rate': rate,
      'volume': volume,
      'pitch': pitch,
      'silenceDuration': silenceDuration,
    }) ?? false;
  }
  
  /// 设置音频输出类型
  /// 
  /// [outputType] 输出类型，speaker(扬声器)，earpiece(听筒)，auto(自动)
  Future<bool> setAudioOutputType(String outputType) async {
    return await _channel.invokeMethod('setAudioOutputType', {
      'outputType': outputType,
    }) ?? false;
  }
  
  /// 合成并播放文本
  /// 
  /// [text] 待合成的文本
  Future<bool> speakText(String text) async {
    return await _channel.invokeMethod('speakText', {
      'text': text,
    }) ?? false;
  }
  
  /// 使用大模型合成并播放文本
  /// 
  /// [text] 待合成的文本
  /// [voiceId] 声音ID
  /// [params] 额外参数，JSON字符串
  Future<bool> speakWithBigModel({
    required String text,
    required String voiceId,
    String params = '',
  }) async {
    return await _channel.invokeMethod('speakWithBigModel', {
      'text': text,
      'voiceId': voiceId,
      'params': params,
    }) ?? false;
  }
  
  /// 暂停播放
  Future<bool> pausePlayback() async {
    return await _channel.invokeMethod('pausePlayback') ?? false;
  }
  
  /// 恢复播放
  Future<bool> resumePlayback() async {
    return await _channel.invokeMethod('resumePlayback') ?? false;
  }
  
  /// 停止播放
  Future<bool> stopSpeaking() async {
    return await _channel.invokeMethod('stopSpeaking') ?? false;
  }
  
  /// 释放TTS资源
  Future<bool> dispose() async {
    return await _channel.invokeMethod('dispose') ?? false;
  }
}