index.mjs

// Import necessary libraries
import express from "express";
import 'dotenv/config.js'
import { createServer } from "http";
import { Server } from "socket.io";
import { client } from "@gradio/client";
import { createRequire } from 'node:module';
import fs from 'fs';
import path from 'path';

const require = createRequire(import.meta.url);
global.EventSource = require('eventsource');

// Initialize the Express app and HTTP server
const app = express();
const httpServer = createServer(app);

// Serve static files from 'public' directory
app.use(express.static('public'));

// Initialize Socket.IO for real-time web socket communication
const io = new Server(httpServer, {});

// Directory for saving inference files
const outputDir = 'inference_outputs';
if (!fs.existsSync(outputDir)){
    fs.mkdirSync(outputDir);
}

// Handle new socket connection
io.on("connection", (socket) => {
    console.log("🔌 New socket connection");
    
    // Listen for 'ask_api' events from the client
    socket.on("ask_api", (client_data) => {
        console.log("📸 Received data from client");
        asyncAPICall(client_data, socket);
    });
});

// Example function to test server communication with a Gradio API
async function test_servers(){
    try{
        const grapi_test = await client("https://gradio-hello-world.hf.space");
        const apitest_result = await grapi_test.predict("/predict", ["John"]);
        console.log(apitest_result);
    }
    catch(e){
        console.log("❌ Error testing servers:", e);
    }
}

// Function to call the Gradio API asynchronously
async function asyncAPICall(data, socket) {
    const grapi = await client("fffiloni/mndrm-call");
    try{
        // Perform the API call with the provided image blob and question
        const api_result = await grapi.predict("/infer", [data[0], data[1]]);
        console.log("✅ API Result:", api_result);
        
        // Emit the API response back to the client
        socket.emit("api_response", api_result.data);
        
        // Save the inference input and output
        saveInference(data, api_result.data);
    }
    catch(e){
        console.log("❌ API Call Error:", e);
        socket.emit("api_error", "ERROR ON API SIDE, SORRY...");
    }    
}

// Save the inference input (image and question) and output (API response) to files
function saveInference(data, apiResponse) {
    const timestamp = new Date().toISOString().replace(/[:.]/g, '-');
    const basePath = path.join(outputDir, timestamp);
    
    // Save image
    const imageData = data[0].split(',')[1]; // Assuming data[0] is a base64 image string
    fs.writeFileSync(`${basePath}_image.png`, imageData, 'base64');
    
    // Save question and API response in a text file
    const textContent = `Question: ${data[1]}\nAPI Response: ${apiResponse}`;
    fs.writeFileSync(`${basePath}_info.txt`, textContent);
    
    console.log("📁 Inference data saved:", basePath);
}

// Start the HTTP server
httpServer.listen(7860, () => console.log("🚀 App running on localhost:7860"));