breakpilot-lehrer/studio-v2/lib/voice/voice-api.ts

/**
 * Voice API Client
 * WebSocket-based voice streaming to voice-service
 */

import { VoiceEncryption } from './voice-encryption'

const VOICE_SERVICE_URL =
  process.env.NEXT_PUBLIC_VOICE_SERVICE_URL || 'http://localhost:8091'
const WS_URL = VOICE_SERVICE_URL.replace('http', 'ws')

export interface VoiceSession {
  id: string
  namespace_id: string
  status: string
  created_at: string
  websocket_url: string
}

export interface VoiceTask {
  id: string
  session_id: string
  type: string
  state: string
  created_at: string
  updated_at: string
  result_available: boolean
  error_message?: string
}

export interface TranscriptMessage {
  type: 'transcript'
  text: string
  final: boolean
  confidence: number
}

export interface IntentMessage {
  type: 'intent'
  intent: string
  confidence: number
  parameters: Record<string, unknown>
}

export interface ResponseMessage {
  type: 'response'
  text: string
}

export interface StatusMessage {
  type: 'status'
  status: string
}

export interface TaskCreatedMessage {
  type: 'task_created'
  task_id: string
  task_type: string
  state: string
}

export interface ErrorMessage {
  type: 'error'
  message: string
  code: string
}

export type VoiceMessage =
  | TranscriptMessage
  | IntentMessage
  | ResponseMessage
  | StatusMessage
  | TaskCreatedMessage
  | ErrorMessage

export type VoiceEventHandler = (message: VoiceMessage) => void
export type AudioHandler = (audioData: ArrayBuffer) => void
export type ErrorHandler = (error: Error) => void

/**
 * Voice API Client
 * Handles session management and WebSocket streaming
 */
export class VoiceAPI {
  private encryption: VoiceEncryption
  private session: VoiceSession | null = null
  private ws: WebSocket | null = null
  private audioContext: AudioContext | null = null
  private mediaStream: MediaStream | null = null
  private processor: ScriptProcessorNode | null = null

  private onMessage: VoiceEventHandler | null = null
  private onAudio: AudioHandler | null = null
  private onError: ErrorHandler | null = null
  private onStatusChange: ((status: string) => void) | null = null

  constructor() {
    this.encryption = new VoiceEncryption()
  }

  /**
   * Initialize the voice API
   */
  async initialize(): Promise<void> {
    await this.encryption.initialize()
  }

  /**
   * Check if API is ready
   */
  isReady(): boolean {
    return this.encryption.isInitialized()
  }

  /**
   * Create a new voice session
   */
  async createSession(): Promise<VoiceSession> {
    const namespaceId = this.encryption.getNamespaceId()
    const keyHash = this.encryption.getKeyHash()

    if (!namespaceId || !keyHash) {
      throw new Error('Encryption not initialized')
    }

    const response = await fetch(`${VOICE_SERVICE_URL}/api/v1/sessions`, {
      method: 'POST',
      headers: { 'Content-Type': 'application/json' },
      body: JSON.stringify({
        namespace_id: namespaceId,
        key_hash: keyHash,
        device_type: 'pwa',
        client_version: '1.0.0',
      }),
    })

    if (!response.ok) {
      throw new Error(`Failed to create session: ${response.statusText}`)
    }

    this.session = await response.json()
    return this.session!
  }

  /**
   * Connect to WebSocket for voice streaming
   */
  async connect(): Promise<void> {
    if (!this.session) {
      await this.createSession()
    }

    return new Promise((resolve, reject) => {
      const wsUrl = this.session!.websocket_url

      this.ws = new WebSocket(wsUrl)

      this.ws.onopen = () => {
        console.log('WebSocket connected')
        this.onStatusChange?.('connected')
        resolve()
      }

      this.ws.onerror = (event) => {
        console.error('WebSocket error:', event)
        this.onError?.(new Error('WebSocket connection failed'))
        reject(new Error('WebSocket connection failed'))
      }

      this.ws.onclose = () => {
        console.log('WebSocket closed')
        this.onStatusChange?.('disconnected')
      }

      this.ws.onmessage = (event) => {
        if (event.data instanceof Blob) {
          // Binary audio data
          event.data.arrayBuffer().then((buffer) => {
            this.onAudio?.(buffer)
          })
        } else {
          // JSON message
          try {
            const message = JSON.parse(event.data) as VoiceMessage
            this.onMessage?.(message)

            if (message.type === 'status') {
              this.onStatusChange?.(message.status)
            }
          } catch (e) {
            console.warn('Failed to parse message:', event.data)
          }
        }
      }
    })
  }

  /**
   * Start capturing audio from microphone
   */
  async startCapture(): Promise<void> {
    try {
      // Request microphone access
      this.mediaStream = await navigator.mediaDevices.getUserMedia({
        audio: {
          sampleRate: 24000,
          channelCount: 1,
          echoCancellation: true,
          noiseSuppression: true,
        },
      })

      // Create audio context
      this.audioContext = new AudioContext({ sampleRate: 24000 })
      const source = this.audioContext.createMediaStreamSource(this.mediaStream)

      // Create processor for capturing audio
      // Note: ScriptProcessorNode is deprecated but still widely supported
      // In production, use AudioWorklet
      this.processor = this.audioContext.createScriptProcessor(2048, 1, 1)

      this.processor.onaudioprocess = (event) => {
        if (!this.ws || this.ws.readyState !== WebSocket.OPEN) return

        const inputData = event.inputBuffer.getChannelData(0)

        // Convert Float32 to Int16
        const int16Data = new Int16Array(inputData.length)
        for (let i = 0; i < inputData.length; i++) {
          const s = Math.max(-1, Math.min(1, inputData[i]))
          int16Data[i] = s < 0 ? s * 0x8000 : s * 0x7fff
        }

        // Send audio chunk
        this.ws.send(int16Data.buffer)
      }

      source.connect(this.processor)
      this.processor.connect(this.audioContext.destination)

      this.onStatusChange?.('listening')
    } catch (error) {
      console.error('Failed to start capture:', error)
      this.onError?.(error as Error)
      throw error
    }
  }

  /**
   * Stop capturing audio
   */
  stopCapture(): void {
    if (this.processor) {
      this.processor.disconnect()
      this.processor = null
    }

    if (this.mediaStream) {
      this.mediaStream.getTracks().forEach((track) => track.stop())
      this.mediaStream = null
    }

    if (this.audioContext) {
      this.audioContext.close()
      this.audioContext = null
    }

    // Signal end of turn
    if (this.ws && this.ws.readyState === WebSocket.OPEN) {
      this.ws.send(JSON.stringify({ type: 'end_turn' }))
    }

    this.onStatusChange?.('processing')
  }

  /**
   * Send interrupt signal
   */
  interrupt(): void {
    if (this.ws && this.ws.readyState === WebSocket.OPEN) {
      this.ws.send(JSON.stringify({ type: 'interrupt' }))
    }
  }

  /**
   * Disconnect from voice service
   */
  async disconnect(): Promise<void> {
    this.stopCapture()

    if (this.ws) {
      this.ws.close()
      this.ws = null
    }

    if (this.session) {
      try {
        await fetch(
          `${VOICE_SERVICE_URL}/api/v1/sessions/${this.session.id}`,
          { method: 'DELETE' }
        )
      } catch (e) {
        console.warn('Failed to close session:', e)
      }
      this.session = null
    }
  }

  /**
   * Get pending tasks for current session
   */
  async getTasks(): Promise<VoiceTask[]> {
    if (!this.session) return []

    const response = await fetch(
      `${VOICE_SERVICE_URL}/api/v1/sessions/${this.session.id}/tasks`
    )

    if (!response.ok) {
      throw new Error('Failed to get tasks')
    }

    return response.json()
  }

  /**
   * Approve a task
   */
  async approveTask(taskId: string): Promise<void> {
    const response = await fetch(
      `${VOICE_SERVICE_URL}/api/v1/tasks/${taskId}/transition`,
      {
        method: 'PUT',
        headers: { 'Content-Type': 'application/json' },
        body: JSON.stringify({
          new_state: 'approved',
          reason: 'user_approved',
        }),
      }
    )

    if (!response.ok) {
      throw new Error('Failed to approve task')
    }
  }

  /**
   * Reject a task
   */
  async rejectTask(taskId: string): Promise<void> {
    const response = await fetch(
      `${VOICE_SERVICE_URL}/api/v1/tasks/${taskId}/transition`,
      {
        method: 'PUT',
        headers: { 'Content-Type': 'application/json' },
        body: JSON.stringify({
          new_state: 'rejected',
          reason: 'user_rejected',
        }),
      }
    )

    if (!response.ok) {
      throw new Error('Failed to reject task')
    }
  }

  /**
   * Set event handlers
   */
  setOnMessage(handler: VoiceEventHandler): void {
    this.onMessage = handler
  }

  setOnAudio(handler: AudioHandler): void {
    this.onAudio = handler
  }

  setOnError(handler: ErrorHandler): void {
    this.onError = handler
  }

  setOnStatusChange(handler: (status: string) => void): void {
    this.onStatusChange = handler
  }

  /**
   * Get current session
   */
  getSession(): VoiceSession | null {
    return this.session
  }

  /**
   * Get connection status
   */
  isConnected(): boolean {
    return this.ws !== null && this.ws.readyState === WebSocket.OPEN
  }
}