whio-api-sdk 1.0.198-beta-staging → 1.0.199-beta-staging

package/WEBSOCKET_AUDIO_STREAMING.md

@@ -5,7 +5,7 @@ The SDK now supports real-time audio streaming via WebSocket connections, provid
 ## Features
 
 - 🔄 **Auto-reconnection** with exponential backoff
-- 📡 **Real-time audio streaming** with float array support
+- 📡 **Real-time audio streaming** with base64 audio support
 - 🎯 **Event-driven architecture** for connection and audio events
 - 🛡️ **Automatic authentication** using existing SDK tokens
 - 📊 **Connection statistics** and monitoring
@@ -85,7 +85,7 @@ sdk.onWebSocketEvent('connection-error', (error) => {
 
 ```typescript
 // Stream individual audio chunks
-const audioChunk = [0.1, 0.2, -0.3, 0.4]; // Float array
+const audioChunk = "SGVsbG8gV29ybGQ="; // Base64 encoded audio data
 sdk.streamAudioChunk(session.id, audioChunk, false);
 
 // Stream final chunk to trigger transcription
@@ -97,9 +97,9 @@ sdk.streamAudioChunk(session.id, finalChunk, true);
 ```typescript
 // Stream multiple chunks with automatic pacing
 const audioChunks = [
-  [0.1, 0.2, 0.3],
-  [0.4, 0.5, 0.6],
-  [0.7, 0.8, 0.9]
+  "SGVsbG8gV29ybGQ=", // Base64 chunk 1
+  "QmFzZTY0IGNodW5r", // Base64 chunk 2
+  "QXVkaW8gZGF0YQ==" // Base64 chunk 3
 ];
 
 await sdk.streamAudioChunks(session.id, audioChunks, {
@@ -120,9 +120,9 @@ processor.onaudioprocess = (event) => {
   const inputBuffer = event.inputBuffer;
   const inputData = inputBuffer.getChannelData(0);
   
-  // Convert to regular array and stream
-  const audioChunk = Array.from(inputData);
-  sdk.streamAudioChunk(session.id, audioChunk, false);
+  // Convert float data to base64 (you'll need your own conversion logic)
+  const base64Chunk = convertFloatArrayToBase64(inputData);
+  sdk.streamAudioChunk(session.id, base64Chunk, false);
 };
 
 source.connect(processor);
@@ -130,7 +130,7 @@ processor.connect(audioContext.destination);
 
 // Stop recording and trigger transcription
 setTimeout(() => {
-  sdk.streamAudioChunk(session.id, [], true);
+  sdk.streamAudioChunk(session.id, "", true); // Empty string for end
   // Cleanup...
 }, 10000);
 ```
@@ -187,11 +187,11 @@ The server automatically buffers audio chunks and flushes them when:
 
 ```typescript
 // The server handles buffering automatically
-sdk.streamAudioChunk(sessionId, chunk1, false); // Buffered
-sdk.streamAudioChunk(sessionId, chunk2, false); // Buffered
+sdk.streamAudioChunk(sessionId, base64Chunk1, false); // Buffered
+sdk.streamAudioChunk(sessionId, base64Chunk2, false); // Buffered
 // ... (8 more chunks buffered)
-sdk.streamAudioChunk(sessionId, chunk10, false); // Auto-flush triggered
-sdk.streamAudioChunk(sessionId, chunk11, true);  // End flag triggers transcription
+sdk.streamAudioChunk(sessionId, base64Chunk10, false); // Auto-flush triggered
+sdk.streamAudioChunk(sessionId, base64Chunk11, true);  // End flag triggers transcription
 ```
 
 ## Error Handling
@@ -253,6 +253,6 @@ for (const chunk of audioChunks) {
 ## Troubleshooting
 
 - **Connection Issues**: Check network connectivity and authentication
-- **Audio Not Processing**: Ensure chunks are valid float arrays
+- **Audio Not Processing**: Ensure chunks are valid base64 strings
 - **Transcription Not Starting**: Verify end flag is sent
 - **Reconnection Problems**: Check server availability and token validity

package/dist/src/sdk/modules/websocket.module.d.ts

@@ -33,11 +33,11 @@ export declare class WebSocketModule extends BaseClient {
     /**
      * Stream audio chunk to server
      */
-    streamAudioChunk(sessionId: string, audioChunk: number[] | number[][], isEnd?: boolean): void;
+    streamAudioChunk(sessionId: string, audioChunk: string, isEnd?: boolean): void;
     /**
      * Stream multiple audio chunks with automatic end detection
      */
-    streamAudioChunks(sessionId: string, audioChunks: number[], options?: AudioStreamingOptions): Promise<void>;
+    streamAudioChunks(sessionId: string, audioChunks: string[], options?: AudioStreamingOptions): Promise<void>;
     /**
      * Add event listener
      */

package/dist/src/sdk/modules/websocket.module.js

@@ -107,11 +107,10 @@ export class WebSocketModule extends BaseClient {
             if (!((_a = this.socket) === null || _a === void 0 ? void 0 : _a.connected)) {
                 throw new Error('WebSocket not connected. Cannot stream audio chunks.');
             }
-            const chunks = audioChunks;
             const delay = options.bufferSize || 100; // ms between chunks
-            for (let i = 0; i < chunks.length; i++) {
-                const isLastChunk = i === chunks.length - 1;
-                this.streamAudioChunk(sessionId, chunks, isLastChunk);
+            for (let i = 0; i < audioChunks.length; i++) {
+                const isLastChunk = i === audioChunks.length - 1;
+                this.streamAudioChunk(sessionId, audioChunks[i], isLastChunk);
                 // Small delay between chunks to prevent overwhelming the server
                 if (!isLastChunk) {
                     yield new Promise(resolve => setTimeout(resolve, delay));

package/dist/src/sdk/types/websocket.types.d.ts

@@ -1,6 +1,6 @@
 export interface AudioChunkPayload {
     sessionId: string;
-    audioChunk: number[] | number[][];
+    audioChunk: string;
     flag: 'end' | null;
 }
 export interface WebSocketConfig {

package/examples/websocket-audio-streaming.ts

@@ -72,21 +72,19 @@ async function exampleWebSocketAudioStreaming() {
 
   console.log('WebSocket connected, starting audio streaming...');
 
-  // Simulate audio data (in a real app, this would come from microphone)
-  const simulateAudioChunks = (): number[][] => {
-    const chunks: number[][] = [];
+  // Simulate base64 audio data (in a real app, this would come from your audio processing)
+  const simulateBase64AudioChunks = (): string[] => {
+    const chunks: string[] = [];
     for (let i = 0; i < 15; i++) {
-      const chunk: number[] = [];
-      for (let j = 0; j < 1024; j++) {
-        // Generate random audio samples between -1 and 1
-        chunk.push((Math.random() - 0.5) * 2);
-      }
-      chunks.push(chunk);
+      // Simulate different sized base64 chunks (this would be your actual audio data)
+      const mockData = `mock-audio-chunk-${i}-${Date.now()}`;
+      const base64Chunk = btoa(mockData);
+      chunks.push(base64Chunk);
     }
     return chunks;
   };
 
-  const audioChunks = simulateAudioChunks();
+  const audioChunks = simulateBase64AudioChunks();
 
   // Stream audio chunks one by one
   for (let i = 0; i < audioChunks.length; i++) {
@@ -159,11 +157,12 @@ async function exampleMicrophoneStreaming() {
       const inputBuffer = event.inputBuffer;
       const inputData = inputBuffer.getChannelData(0);
       
-      // Convert Float32Array to regular array
-      const audioChunk = Array.from(inputData);
+      // Convert Float32Array to base64 (you'll need your own conversion logic)
+      // This is just a placeholder - replace with your actual conversion
+      const mockBase64 = btoa(String.fromCharCode(...Array.from(inputData.slice(0, 100))));
       
-      // Stream the audio chunk
-      sdk.streamAudioChunk(session.id, audioChunk, false);
+      // Stream the base64 audio chunk
+      sdk.streamAudioChunk(session.id, mockBase64, false);
     };
 
     source.connect(processor);
@@ -173,8 +172,8 @@ async function exampleMicrophoneStreaming() {
 
     // Stop after 10 seconds and trigger transcription
     setTimeout(() => {
-      // Send empty chunk with end flag to trigger transcription
-      sdk.streamAudioChunk(session.id, [], true);
+      // Send empty base64 string with end flag to trigger transcription
+      sdk.streamAudioChunk(session.id, "", true);
       
       // Clean up
       processor.disconnect();

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "whio-api-sdk",
-  "version": "1.0.198-beta-staging",
+  "version": "1.0.199-beta-staging",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
   "type": "module",

package/src/sdk/modules/websocket.module.ts

@@ -113,7 +113,7 @@ export class WebSocketModule extends BaseClient {
    */
   public streamAudioChunk(
     sessionId: string, 
-    audioChunk: number[] | number[][], 
+    audioChunk: string, 
     isEnd: boolean = false
   ): void {
     if (!this.socket?.connected) {
@@ -134,20 +134,19 @@ export class WebSocketModule extends BaseClient {
    */
   public async streamAudioChunks(
     sessionId: string,
-    audioChunks: number[], 
+    audioChunks: string[], 
     options: AudioStreamingOptions = {}
   ): Promise<void> {
     if (!this.socket?.connected) {
       throw new Error('WebSocket not connected. Cannot stream audio chunks.');
     }
 
-    const chunks = audioChunks;
     const delay = options.bufferSize || 100; // ms between chunks
 
-    for (let i = 0; i < chunks.length; i++) {
-      const isLastChunk = i === chunks.length - 1;
+    for (let i = 0; i < audioChunks.length; i++) {
+      const isLastChunk = i === audioChunks.length - 1;
       
-      this.streamAudioChunk(sessionId, chunks, isLastChunk);
+      this.streamAudioChunk(sessionId, audioChunks[i], isLastChunk);
       
       // Small delay between chunks to prevent overwhelming the server
       if (!isLastChunk) {

package/src/sdk/types/websocket.types.ts

@@ -2,7 +2,7 @@
 
 export interface AudioChunkPayload {
   sessionId: string;
-  audioChunk: number[] | number[][];
+  audioChunk: string; // Base64 encoded audio data
   flag: 'end' | null;
 }