Merge pull request #118 from chingu-voyages/prompt/ayer/speech

Prompt/ayer/speech

Author: aigul-ermak

backend/package.json

@@ -20,6 +20,7 @@
     "googleapis": "^148.0.0",
     "js-yaml": "^4.1.0",
     "jsonwebtoken": "^9.0.2",
+    "multer": "^1.4.5-lts.2",
     "octokit": "^4.1.2",
     "swagger-jsdoc": "^6.2.8",
     "swagger-ui-express": "^5.0.1",
@@ -42,6 +43,7 @@
     "@types/jest": "^29.5.14",
     "@types/js-yaml": "^4",
     "@types/jsonwebtoken": "^9.0.9",
+    "@types/multer": "^1",
     "@types/node": "^22.13.10",
     "@types/supertest": "^6.0.2",
     "@types/swagger-jsdoc": "^6",

backend/src/config/settings.ts

@@ -4,34 +4,35 @@ import cookieParser from "cookie-parser";
 import session from "express-session";
 import cors from "cors";
 
-import { corsOptions, securityHeaders, authMiddleware } from "../middleware/index.js";
-import { authRoute, promptRoute, userRoute } from "../routes/index.js";
+import {authMiddleware, corsOptions, securityHeaders} from "../middleware/index.js";
+import {audioRoute, authRoute, promptRoute, userRoute} from "../routes/index.js";
 
 export const configApp = async () => {
-  const app = express();
-  app.use(express.json());
+    const app = express();
+    app.use(express.json());
 
-  if (process.env.NODE_ENV !== "test") {
-    const { setupSwagger } = await import("../swagger.js");
-    setupSwagger(app);
-  }
+    if (process.env.NODE_ENV !== "test") {
+        const {setupSwagger} = await import("../swagger.js");
+        setupSwagger(app);
+    }
 
-  app.use(cookieParser());
-  app.use(securityHeaders); // Security headers for privacy-focused browsers
-  app.use(cors(corsOptions));
+    app.use(cookieParser());
+    app.use(securityHeaders); // Security headers for privacy-focused browsers
+    app.use(cors(corsOptions));
 
-  // Serve static files
-  const __dirname = path.resolve();
-  app.use(express.static(path.join(__dirname, "..", "static")));
+    // Serve static files
+    const __dirname = path.resolve();
+    app.use(express.static(path.join(__dirname, "..", "static")));
 
-  const sessionSecret = String(process.env.SESSION_SECRET);
-  app.use(
-    session({ secret: sessionSecret, resave: false, saveUninitialized: true })
-  );
+    const sessionSecret = String(process.env.SESSION_SECRET);
+    app.use(
+        session({secret: sessionSecret, resave: false, saveUninitialized: true})
+    );
 
-  app.use("/", authRoute);
-  app.use("/users", authMiddleware, userRoute);
-  app.use("/prompts", authMiddleware, promptRoute);
+    app.use("/", authRoute);
+    app.use("/users", authMiddleware, userRoute);
+    app.use("/prompts", authMiddleware, promptRoute);
+    app.use("/audio", audioRoute);
 
-  return app;
+    return app;
 };

backend/src/controllers/audioController.ts

@@ -0,0 +1,22 @@
+import {Request, Response} from "express";
+import {generateGeminiAudioResponse} from "../services/geminiService.js";
+
+
+export const transcribePrompt = async (req: Request, res: Response) => {
+    try {
+        if (!req.file) {
+           res.status(400).json({error: 'No audio file provided'});
+           return;
+        }
+
+        const result = await generateGeminiAudioResponse({
+            audioBuffer: req.file.buffer,
+            mimeType: req.file.mimetype
+        });
+
+        res.json(result);
+    } catch (error) {
+        console.error('Audio transcription error:', error);
+        res.status(500).json({error: 'Failed to transcribe audio'});
+    }
+};

backend/src/controllers/index.ts

@@ -1,3 +1,4 @@
 export * as authController from "./authController.js";
 export * as userController from "./userController.js";
 export * as promptController from "./promptController.js";
+export * as audioController from "./audioController.js";

backend/src/routes/audioRoutes.ts

@@ -0,0 +1,14 @@
+import {Router} from "express";
+import multer from 'multer';
+import {audioController} from "../controllers/index.js";
+
+const storage = multer.memoryStorage();
+const upload = multer({
+    storage,
+    limits: {fileSize: 20 * 1024 * 1024} // 20MB limit
+});
+
+export const audioRoute: Router = Router();
+
+audioRoute.post("/transcribe", upload.single('audio'), audioController.transcribePrompt);
+

backend/src/routes/index.ts

@@ -1,3 +1,4 @@
 export * from "./userRoutes.js";
 export * from "./promptRoutes.js";
 export * from "./authRoutes.js";
+export * from "./audioRoutes.js";

backend/src/routes/promptRoutes.ts

@@ -345,42 +345,6 @@ promptRoute.get("/", promptController.getAllPrompts);
  */
 promptRoute.put("/:promptId", promptController.updateScorePrompt);
 
-// /**
-//  * @swagger
-//  * /prompts:
-//  *   delete:
-//  *     summary: Delete all prompts for the authenticated user
-//  *     description: Removes all prompt records associated with the currently authenticated user.
-//  *     tags:
-//  *       - Prompts
-//  *     security:
-//  *       - cookieAuth: []
-//  *     responses:
-//  *       204:
-//  *         description: All prompts successfully deleted (no content returned)
-//  *       401:
-//  *         description: Unauthorized - user not authenticated
-//  *         content:
-//  *           application/json:
-//  *             schema:
-//  *               type: object
-//  *               properties:
-//  *                 error:
-//  *                   type: string
-//  *                   example: Unauthorized
-//  *       500:
-//  *         description: Internal server error
-//  *         content:
-//  *           application/json:
-//  *             schema:
-//  *               type: object
-//  *               properties:
-//  *                 error:
-//  *                   type: string
-//  *                   example: Something went wrong
-//  */
-// promptRoute.delete("/", authMiddleware, promptController.deleteAllPrompts);
-
 /**
  * @swagger
  * /prompts/{promptId}:
@@ -445,4 +409,3 @@ promptRoute.put("/:promptId", promptController.updateScorePrompt);
  */
 promptRoute.delete("/:promptId", promptController.deletePrompt);
 
-

backend/src/services/geminiService.ts

@@ -1,5 +1,6 @@
 import {GoogleGenerativeAI} from "@google/generative-ai";
 import {GeminiResponseType} from "../types/promptTypes.js";
+import {AudioRequest} from "../types/audioTypes.js";
 
 // Initialize Gemini AI with the API key from environment variables
 const genAI = new GoogleGenerativeAI(process.env.GEMINI_API_KEY!);
@@ -54,6 +55,39 @@ export const generateGeminiResponse = async (promptText: string): Promise<Gemini
     }
 };
 
+export const generateGeminiAudioResponse = async (
+    { audioBuffer, mimeType }: AudioRequest
+): Promise<string> => {
+    try {
+
+        const model = genAI.getGenerativeModel({ model: "gemini-2.0-flash" });
+        const base64Audio = audioBuffer.toString('base64');
+
+        const result = await model.generateContent({
+            contents: [{
+                role: "user",
+                parts: [
+                    { text: "Generate a transcript of this audio:" },
+                    {
+                        inlineData: {
+                            data: base64Audio,
+                            mimeType
+                        }
+                    }
+                ]
+            }]
+        });
+
+        return stripMarkdown(result.response.text());
+
+    } catch (error) {
+        console.error("Gemini Audio Error:", error);
+        throw error;
+    }
+};
+
+
+
 /**
  * Strips Markdown formatting from the given text.
  *
@@ -71,3 +105,6 @@ function stripMarkdown(text: string): string {
         .replace(/\n{2,}/g, '\n\n') // Normalize spacing
         .trim(); // Remove leading/trailing whitespace
 }
+
+
+

backend/src/types/audioTypes.ts

@@ -0,0 +1,4 @@
+export interface AudioRequest {
+    audioBuffer: Buffer;
+    mimeType: string;
+}