@coze/api
Version:
Official Coze Node.js SDK for seamless AI integration into your applications | 扣子官方 Node.js SDK，助您轻松集成 AI 能力到应用中
406 lines (405 loc) • 12.1 kB
TypeScript
import { type TranscriptionsUpdateEvent } from '../resources/websockets/types';
import { type AIDenoiserProcessorLevel, type AIDenoiserProcessorMode } from './recorder/pcm-recorder';
import { type WebsocketOptions } from '../core';
import { type ChatUpdateEvent, type CreateChatWsRes, type GetToken } from '..';
export interface WsToolsOptions {
/** Personal Access Token (PAT) or OAuth2.0 token, or a function to get token */
token: GetToken;
/** Whether to enable debug mode */
debug?: boolean;
/** Custom headers */
headers?: Headers | Record<string, unknown>;
/** Whether Personal Access Tokens (PAT) are allowed in browser environments */
allowPersonalAccessTokenInBrowser?: boolean;
/** base websocket URL, default is wss://ws.coze.cn */
baseWsURL?: string;
/** websocket options */
websocketOptions?: WebsocketOptions;
}
export interface WsSpeechClientOptions extends WsToolsOptions {
entityType?: 'bot' | 'workflow';
entityId?: string;
}
export interface SentenceItem {
/** 事件 ID */
id: string;
/** 句子文本 */
content: string;
/** 该句子的音频时长（毫秒） */
audioDuration: number;
/** 是否是最后一个句子 */
isLastSentence: boolean;
/** 句子音频时长是否已更新完成 */
isDurationFinish: boolean;
}
export interface AudioRecordEvent {
event_type: ClientEventType.AUDIO_INPUT_DUMP;
data: {
name: string;
wav: Blob;
};
}
export interface AudioSentencePlaybackStartEvent {
event_type: ClientEventType.AUDIO_SENTENCE_PLAYBACK_START;
data: {
/** 句子对应的事件 ID */
id: string;
/** 句子文本 */
content: string;
};
}
export interface AudioSentencePlaybackEndEvent {
event_type: ClientEventType.AUDIO_SENTENCE_PLAYBACK_ENDED;
}
export type WsChatEventData = CreateChatWsRes | AudioRecordEvent | undefined | ChatUpdateEvent | AudioSentencePlaybackStartEvent | AudioSentencePlaybackEndEvent;
export type WsChatCallbackHandler = (eventName: string, event: WsChatEventData) => void;
export interface AudioCaptureConfig {
/**
* en: Whether to enable noise suppression
* zh: 是否启用降噪
*/
noiseSuppression?: boolean;
/**
* en: Whether to enable echo cancellation
* zh: 是否启用回声消除
*/
echoCancellation?: boolean;
/**
* en: Whether to enable automatic gain control
* zh: 是否启用自动增益控制
*/
autoGainControl?: boolean;
}
export interface AIDenoisingConfig {
/**
* en: AI denoiser mode
* - `NSNG`: Enable non-stationary noise suppression.
* - `STATIONARY_NS`: Enable stationary noise suppression.
* zh: AI 降噪
* - `NSNG`: 启用非平稳噪声抑制
* - `STATIONARY_NS`: 启用平稳噪声抑制
*/
mode?: AIDenoiserProcessorMode;
/**
* en: AI denoiser level
* - `SOFT`: (Recommended) Soft mode
* - `AGGRESSIVE`: Aggressive mode
* zh: AI 降噪等级
* - `SOFT`: （推荐）舒缓降噪
* - `AGGRESSIVE`: 激进降噪。将降噪强度提高到激进降噪会增大损伤人声的概率。
*/
level?: AIDenoiserProcessorLevel;
/**
* en: The path to the AI denoiser assets, default is `/external`
* zh: AI 降噪 wasm 文件路径，默认是 `/external`
*/
assetsPath?: string;
}
export interface WavRecordConfig {
/**
* en: Whether to enable source record
* zh: 是否启用源音频录制
*/
enableSourceRecord: boolean;
/**
* en: Whether to enable denoise record
* zh: 是否启用降噪后音频录制
*/
enableDenoiseRecord: boolean;
}
export interface WsChatClientOptions extends WsToolsOptions {
/**
* en: Bot id
* zh: 智能体 ID
*/
botId: string;
/**
* en: Workflow id
* zh: 工作流 ID
*/
workflowId?: string;
/**
* en: Voice id
* zh: 音色 ID
*/
voiceId?: string;
/**
* en: Audio input device id
* zh: 音频输入设备 ID
*/
deviceId?: string;
/**
* en: Whether to mute by default
* zh: 是否默认静音
*/
audioMutedDefault?: boolean;
/**
* en: Audio capture config
* zh: 音频采集配置
*/
audioCaptureConfig?: AudioCaptureConfig;
/**
* en: AI denoising config
* zh: AI 降噪配置
*/
aiDenoisingConfig?: AIDenoisingConfig;
/**
* en: Custom media stream track
* zh: 自定义音频流
*/
mediaStreamTrack?: MediaStreamTrack;
/**
* en: Audio record config, only effective when debug = true
* zh: 音频录制配置，仅在 debug = true 时有效
*/
wavRecordConfig?: WavRecordConfig;
/**
* en: Whether to enable local loopback, default is true on mobile, false on PC
* zh: 是否启用本地回环播放音频，移动端(window.innerWidth <= 768)上默认设置为 true，PC 端默认设置为 false
*/
enableLocalLoopback?: boolean;
/**
* en: Audio playback volume default
* zh: 默认播放音量
*/
playbackVolumeDefault?: number;
}
export declare enum ClientEventType {
/**
* en: Audio input dump
* zh: 音频输入 dump
*/
AUDIO_INPUT_DUMP = "audio.input.dump",
/**
* en: Audio sentence playback start
* zh: 句子开始播放
*/
AUDIO_SENTENCE_PLAYBACK_START = "audio.sentence.playback_start",
/**
* en: Audio sentence playback ended
* zh: 句子播放结束
*/
AUDIO_SENTENCE_PLAYBACK_ENDED = "audio.sentence.playback_ended"
}
export declare enum WsChatEventNames {
/**
* en: All events
* zh: 所有事件
*/
ALL = "realtime.event",
/**
* en: Client connected
* zh: 客户端连接
*/
CONNECTED = "client.connected",
/**
* en: Client connecting
* zh: 客户端连接中
*/
CONNECTING = "client.connecting",
/**
* en: Client interrupted
* zh: 客户端中断
*/
INTERRUPTED = "client.interrupted",
/**
* en: Client disconnected
* zh: 客户端断开
*/
DISCONNECTED = "client.disconnected",
/**
* en: Client audio unmuted
* zh: 客户端音频未静音
*/
AUDIO_UNMUTED = "client.audio.unmuted",
/**
* en: Client audio muted
* zh: 客户端音频静音
*/
AUDIO_MUTED = "client.audio.muted",
/**
* en: Client error
* zh: 客户端错误
*/
ERROR = "client.error",
/**
* en: Denoiser enabled
* zh: 降噪开启
*/
DENOISER_ENABLED = "client.denoiser.enabled",
/**
* en: Denoiser disabled
* zh: 降噪关闭
*/
DENOISER_DISABLED = "client.denoiser.disabled",
/**
* en: Audio input device changed
* zh: 音频输入设备改变
*/
AUDIO_INPUT_DEVICE_CHANGED = "client.input.device.changed",
/**
* en: Audio output device changed
* zh: 音频输出设备改变
*/
AUDIO_OUTPUT_DEVICE_CHANGED = "client.output.device.changed",
/**
* en: Audio record dump
* zh: 音频 dump
*/
AUDIO_INPUT_DUMP = "client.audio.input.dump",
/**
* en: Audio sentence playback start
* zh: 句子开始播放
*/
AUDIO_SENTENCE_PLAYBACK_START = "client.audio.sentence.playback_start",
/**
* en: Audio sentence playback ended
* zh: 句子播放结束
*/
AUDIO_SENTENCE_PLAYBACK_ENDED = "client.audio.sentence.playback_ended",
/**
* en: Chat created
* zh: 对话创建成功
*/
CHAT_CREATED = "server.chat.created",
/**
* en: Chat updated
* zh: 对话更新
*/
CHAT_UPDATED = "server.chat.updated",
/**
* en: Conversation chat created
* zh: 会话对话创建
*/
CONVERSATION_CHAT_CREATED = "server.conversation.chat.created",
/**
* en: Conversation chat in progress
* zh: 对话正在处理中
*/
CONVERSATION_CHAT_IN_PROGRESS = "server.conversation.chat.in.progress",
/**
* en: Conversation message delta received
* zh: 文本消息增量返回
*/
CONVERSATION_MESSAGE_DELTA = "server.conversation.message.delta",
/**
* en: Conversation audio delta received
* zh: 语音消息增量返回
*/
CONVERSATION_AUDIO_DELTA = "server.conversation.audio.delta",
/**
* en: Conversation message completed
* zh: 文本消息完成
*/
CONVERSATION_MESSAGE_COMPLETED = "server.conversation.message.completed",
/**
* en: Conversation audio completed
* zh: 语音回复完成
*/
CONVERSATION_AUDIO_COMPLETED = "server.conversation.audio.completed",
/**
* en: Conversation chat completed
* zh: 对话完成
*/
CONVERSATION_CHAT_COMPLETED = "server.conversation.chat.completed",
/**
* en: Conversation chat failed
* zh: 对话失败
*/
CONVERSATION_CHAT_FAILED = "server.conversation.chat.failed",
/**
* en: Server error occurred
* zh: 服务端错误
*/
SERVER_ERROR = "server.error",
/**
* en: Input audio buffer completed
* zh: 语音输入缓冲区提交完成
*/
INPUT_AUDIO_BUFFER_COMPLETED = "server.input_audio_buffer.completed",
/**
* en: Input audio buffer cleared
* zh: 语音输入缓冲区已清除
*/
INPUT_AUDIO_BUFFER_CLEARED = "server.input_audio_buffer.cleared",
/**
* en: Conversation chat cancelled
* zh: 对话被取消
*/
CONVERSATION_CHAT_CANCELLED = "server.conversation.chat.cancelled",
/**
* en: Conversation context cleared
* zh: 对话上下文已清除
*/
CONVERSATION_CLEARED = "server.conversation.cleared",
/**
* en: Conversation audio transcript updated
* zh: 用户语音识别实时字幕更新
*/
CONVERSATION_AUDIO_TRANSCRIPT_UPDATE = "server.conversation.audio_transcript.update",
/**
* en: Conversation audio transcript completed
* zh: 用户语音识别完成
*/
CONVERSATION_AUDIO_TRANSCRIPT_COMPLETED = "server.conversation.audio_transcript.completed",
/**
* en: Conversation chat requires action
* zh: 对话需要端插件响应
*/
CONVERSATION_CHAT_REQUIRES_ACTION = "server.conversation.chat.requires_action",
/**
* en: User speech detected - started
* zh: 检测到用户开始说话
*/
INPUT_AUDIO_BUFFER_SPEECH_STARTED = "server.input_audio_buffer.speech_started",
/**
* en: User speech detected - stopped
* zh: 检测到用户停止说话
*/
INPUT_AUDIO_BUFFER_SPEECH_STOPPED = "server.input_audio_buffer.speech_stopped",
/**
* en: Audio dump
* zh: 音频 dump
*/
DUMP_AUDIO = "server.dump.audio",
/**
* en: Audio sentence start
* zh: 音频句子开始
*/
CONVERSATION_AUDIO_SENTENCE_START = "server.conversation.audio.sentence_start"
}
export interface WsTranscriptionClientOptions extends WsToolsOptions {
/**
* en: Audio input device id
* zh: 音频输入设备 ID
*/
deviceId?: string;
/**
* en: Audio capture config
* zh: 音频采集配置
*/
audioCaptureConfig?: AudioCaptureConfig;
/**
* en: AI denoising config
* zh: AI 降噪配置
*/
aiDenoisingConfig?: AIDenoisingConfig;
/**
* en: Custom media stream track
* zh: 自定义音频流
*/
mediaStreamTrack?: MediaStreamTrack | (() => Promise<MediaStreamTrack>);
/**
* en: Audio record config, only effective when debug = true
* zh: 音频录制配置，仅在 debug = true 时有效
*/
wavRecordConfig?: WavRecordConfig;
entityType?: 'bot' | 'workflow';
entityId?: string;
/**
* en: Custom data payload for initial transcription update
* zh: 初始语音识别更新的接口配置数据
*/
transcriptionUpdateData?: TranscriptionsUpdateEvent['data'];
}
export type WsSimultInterpretationClientOptions = WsTranscriptionClientOptions;