VideoTube Backend — Complete Documentation

Project: VideoTube Backend
Stack: Node.js · Express 5 · MongoDB (Mongoose) · Socket.io · Cloudinary · Groq AI (LLaMA 3.3) · OpenAI (Whisper + Embeddings) · PDFKit
Entry Point: src/index.js
Port: 4000 (from .env)

1. Product Requirements Document (PRD)

1.1 Vision

VideoTube is a YouTube-like video-sharing platform enhanced with an AI-powered learning assistant. Users can upload, watch, like, comment on, and subscribe to video channels. A unique differentiator is the AI doubt-resolution system that lets students ask questions — either general or based on a specific video's transcript — and receive AI-generated answers.

1.2 Core Features

#	Feature	Status
1	User Registration with Avatar & Cover Image	✅ Implemented
2	Login / Logout with JWT (Access + Refresh Tokens)	✅ Implemented
3	Token Refresh Flow	✅ Implemented
4	Password Change	✅ Implemented
5	Profile Update (name, email, avatar, cover image)	✅ Implemented
6	Channel Profile with Subscriber Count	✅ Implemented
7	Watch History	✅ Implemented
8	Video Upload (Video + Thumbnail to Cloudinary)	✅ Implemented
9	Video Feed (paginated, sorted by newest)	✅ Implemented
10	Watch Video (auto-increment views)	✅ Implemented
11	Like / Unlike Toggle	✅ Implemented
12	Comment System (add, list, delete)	✅ Implemented
13	Subscribe / Unsubscribe Toggle	✅ Implemented
14	AI Doubt Resolution (general questions)	✅ Implemented
15	AI Doubt History	✅ Implemented
16	PDF Notes Generation from Doubt History	🔧 Has bugs
17	Video-specific AI Q&A (transcript-based RAG)	🔧 Has bugs
18	Real-time Live Chat per Video (Socket.io)	🔧 Partial
19	Audio Extraction + Transcript Generation	🔧 Services exist, not wired
20	Embedding-based Semantic Search (RAG)	🔧 Services exist, not wired

1.3 Target Users

Content Creators — upload educational/general videos
Students / Viewers — watch videos, ask AI questions, get PDF study notes
Community — interact via comments and live chat

2. System Architecture

2.1 High-Level Architecture

graph TB
    Client["Client (Browser / Mobile)"]
    
    subgraph Server["Node.js Server"]
        Express["Express 5 App"]
        SocketIO["Socket.io Server"]
        
        subgraph MW["Middleware Layer"]
            Auth["JWT Auth"]
            Multer["Multer (File Upload)"]
            RateLimiter["Rate Limiter"]
            Helmet["Helmet (Security)"]
            CORS["CORS"]
        end
        
        subgraph Routes["Route Layer"]
            UserRoutes["/api/v1/users"]
            VideoRoutes["/api/v1/video"]
            LikeRoutes["/api/v1/likes"]
            CommentRoutes["/api/v1/commnets"]
            SubRoutes["/api/v1/subscriptions"]
            ChatRoutes["/api/v1/chat"]
            AIRoutes["/api/v1/ai"]
        end
        
        subgraph Controllers["Controller Layer"]
            UserCtrl["user.controller"]
            VideoCtrl["video.controller"]
            LikeCtrl["like.controller"]
            CommentCtrl["comment.controller"]
            SubCtrl["subscription.controller"]
            ChatCtrl["chat.controller"]
            AICtrl["ai.controller"]
        end
        
        subgraph Services["Service Layer"]
            AIService["ai.service (Groq LLaMA)"]
            AudioService["audio.service (FFmpeg)"]
            TranscriptService["transcript.service (Whisper)"]
            EmbeddingService["embedding.service (OpenAI)"]
            PDFService["pdf.service (PDFKit)"]
        end
    end
    
    MongoDB["MongoDB Atlas"]
    Cloudinary["Cloudinary CDN"]
    GroqAPI["Groq API"]
    OpenAIAPI["OpenAI API"]
    
    Client --> Express
    Client --> SocketIO
    Express --> MW --> Routes --> Controllers
    Controllers --> Services
    Controllers --> MongoDB
    Controllers --> Cloudinary
    AIService --> GroqAPI
    TranscriptService --> OpenAIAPI
    EmbeddingService --> OpenAIAPI

2.2 Request Lifecycle

sequenceDiagram
    participant C as Client
    participant RL as Rate Limiter
    participant H as Helmet + CORS
    participant R as Router
    participant JWT as JWT Auth Middleware
    participant Ctrl as Controller
    participant SVC as Service
    participant DB as MongoDB
    participant CDN as Cloudinary

    C->>RL: HTTP Request
    RL->>H: Pass (if under limit)
    H->>R: Route matching
    R->>JWT: Protected route check
    JWT->>JWT: Verify access token
    JWT->>DB: Fetch user by token._id
    JWT->>Ctrl: req.user attached
    Ctrl->>SVC: Business logic
    Ctrl->>DB: CRUD operations
    Ctrl->>CDN: Upload/delete files
    Ctrl->>C: ApiResponse JSON

2.3 Directory Structure

Backend/
├── .env                          # Environment variables
├── constant.js                   # DB_NAME = "videotube"
├── package.json                  # Dependencies & scripts
├── public/                       # Temp file storage (multer destination)
├── src/
│   ├── index.js                  # Entry: HTTP server + Socket.io + DB connect
│   ├── app.js                    # Express app config + route mounting
│   ├── db/index.js               # MongoDB connection via Mongoose
│   ├── Socket.io/socket.js       # Real-time chat socket events
│   ├── controller/               # 7 controller files
│   │   ├── ai.controller.js      # AI doubt resolution & PDF notes
│   │   ├── chat.controller.js    # Live chat message retrieval
│   │   ├── comment.controller.js # Add, get, delete comments
│   │   ├── like.controller.js    # Toggle like/unlike
│   │   ├── subscription.controller.js  # Toggle subscribe/unsubscribe
│   │   ├── user.controller.js    # Auth, profile, channel, watch history
│   │   └── video.controller.js   # Upload, feed, watch video
│   ├── models/                   # 8 Mongoose model files
│   │   ├── chatmessage.model.js  # Live chat messages
│   │   ├── comment.model.js      # Video comments
│   │   ├── dout.model.js         # AI doubt Q&A records
│   │   ├── like.model.js         # Video likes
│   │   ├── subscription.model.js # Channel subscriptions
│   │   ├── transcriptChunk.model.js  # Video transcript chunks + embeddings
│   │   ├── user.model.js         # Users with auth tokens
│   │   └── video.model.js        # Videos with metadata
│   ├── route/                    # 7 route files
│   │   ├── ai.route.js           # /api/v1/ai
│   │   ├── chat.route.js         # /api/v1/chat
│   │   ├── comment.route.js      # /api/v1/commnets
│   │   ├── like.route.js         # /api/v1/likes
│   │   ├── subscribe.route.js    # /api/v1/subscriptions
│   │   ├── user.route.js         # /api/v1/users
│   │   └── video.route.js        # /api/v1/video
│   ├── service/                  # 5 service files
│   │   ├── ai.service.js         # Groq LLaMA 3.3 integration
│   │   ├── audio.service.js      # FFmpeg audio extraction
│   │   ├── embedding.service.js  # OpenAI text embeddings
│   │   ├── pdf.service.js        # PDFKit note generation
│   │   └── transcript.service.js # OpenAI Whisper transcription
│   ├── middleware/               # Auth, Multer, Rate limiters
│   │   ├── auth.middleware.js    # JWT verification
│   │   ├── multer.js             # File upload configuration
│   │   └── ratelimmiter/
│   │       └── ratelimmiter.js   # Rate limiting rules
│   └── utils/                    # Shared utilities
│       ├── ApiError.js           # Custom error class
│       ├── ApiResponse.js        # Standardized response class
│       ├── asynchandler.js       # Async error wrapper
│       ├── cloudinary.js         # Upload to Cloudinary
│       ├── cosineSimilarity.js   # Vector similarity for RAG
│       ├── deletefilefromCloudinary.js  # Delete from Cloudinary
│       └── splitTranscript.js    # Transcript chunking for RAG
└── Readme/                       # Dev notes

3. API Reference

Base URL: http://localhost:4000/api/v1
Auth: JWT Bearer token in Authorization header or accessToken cookie
Response Format: { statuscode, data, message, success }

3.1 Users — `/api/v1/users`

Method	Endpoint	Auth	Rate Limit	Middleware	Description
`POST`	`/register`	❌	5 req / 15 min	Multer (avtar, coverImage)	Register new user
`POST`	`/login`	❌	10 req / 5 min	—	Login, returns tokens in cookies
`POST`	`/logout`	✅	—	—	Clear tokens, logout
`POST`	`/refresh-token`	❌	—	—	Refresh access token
`POST`	`/change-password`	✅	—	—	Change password
`GET`	`/current-user`	✅	—	—	Get logged-in user profile
`PATCH`	`/update-account`	✅	—	—	Update fullname & email
`POST`	`/update-avtar`	✅	—	Multer (avtar)	Update avatar image
`POST`	`/cover-image`	✅	—	Multer (coverImage)	Update cover image
`GET`	`/c/:username`	✅	—	—	Get channel profile + subscriber counts
`GET`	`/history`	✅	—	—	Get watch history

Register — Request Body (multipart/form-data)

{
  "fullname": "John Doe",
  "username": "johndoe",
  "email": "john@example.com",
  "password": "secret123",
  "avtar": "<file>",
  "coverImage": "<file>"
}

Login — Request Body

{
  "email": "john@example.com",
  "username": "johndoe",
  "password": "secret123"
}

Login — Response

{
  "statuscode": 200,
  "data": {
    "user": { "_id": "...", "username": "johndoe", "email": "...", "avtar": {} },
    "accessToken": "eyJhbGciOiJIUzI1NiI...",
    "refreshToken": "eyJhbGciOiJIUzI1NiI..."
  },
  "message": "user login successfully",
  "success": true
}

Change Password — Request Body

{
  "oldpassword": "secret123",
  "newpassword": "newsecret456"
}

Update Account — Request Body

{
  "fullname": "John Updated",
  "email": "john.updated@example.com"
}

3.2 Videos — `/api/v1/video`

Method	Endpoint	Auth	Middleware	Description
`POST`	`/upload`	✅	Multer (videoupload, thumbnailupload)	Upload video + thumbnail
`GET`	`/getallvideos?page=1&limit=10`	❌	—	Paginated video feed
`GET`	`/:videoId`	❌	—	Get video by ID (auto-increments views)

Upload Video — Request Body (multipart/form-data)

{
  "title": "Learn JavaScript",
  "description": "A complete JS tutorial",
  "videoupload": "<file>",
  "thumbnailupload": "<file>"
}

Upload Video — Response

{
  "statuscode": 201,
  "data": {
    "_id": "...",
    "title": "Learn JavaScript",
    "description": "A complete JS tutorial",
    "videoFile": { "url": "https://res.cloudinary.com/...", "public_id": "..." },
    "thumbnail": { "url": "https://res.cloudinary.com/...", "public_id": "..." },
    "owner": "...",
    "views": 0,
    "isPublished": true
  },
  "message": "video uploaded successfully",
  "success": true
}

Get All Videos — Response

{
  "statuscode": 200,
  "data": [
    {
      "_id": "...",
      "title": "Learn JavaScript",
      "videoFile": { "url": "..." },
      "thumbnail": { "url": "..." },
      "views": 42,
      "owner": { "username": "johndoe", "avtar": {} },
      "createdAt": "2026-03-20T..."
    }
  ],
  "message": "Video fetch successfully",
  "success": true
}

3.3 Likes — `/api/v1/likes`

Method	Endpoint	Auth	Description
`POST`	`/video/:videoId`	✅	Toggle like/unlike on a video

3.4 Comments — `/api/v1/commnets`

Method	Endpoint	Auth	Description
`POST`	`/:videoId`	✅	Add a comment to a video
`GET`	`/:videoId`	❌	Get all comments on a video
`DELETE`	`/delete/:commentId`	✅	Delete own comment

Add Comment — Request Body

{
  "content": "Great video!"
}

Get Comments — Response

{
  "statuscode": 200,
  "data": [
    {
      "_id": "...",
      "content": "Great video!",
      "video": "...",
      "owner": { "username": "johndoe", "avtar": {} },
      "createdAt": "2026-03-20T..."
    }
  ],
  "message": "Comments fetched successfully",
  "success": true
}

3.5 Subscriptions — `/api/v1/subscriptions`

Method	Endpoint	Auth	Description
`POST`	`/:channelId`	✅	Toggle subscribe/unsubscribe

3.6 Chat — `/api/v1/chat`

Method	Endpoint	Auth	Description
`GET`	`/:videoId`	❌	Get all chat messages for a video

Note: Live chat messages are sent via Socket.io (see Section 7). This REST endpoint only retrieves message history.

3.7 AI — `/api/v1/ai`

Method	Endpoint	Auth	Description
`POST`	`/ask`	✅	Ask a general AI question
`GET`	`/history`	✅	Get user's doubt history
`GET`	`/notes-pdf`	✅	Download doubt history as PDF
`POST`	`/video-doubt`	✅	Ask AI about a specific video

Ask AI — Request Body

{
  "question": "What is a closure in JavaScript?"
}

Ask AI — Response

{
  "statuscode": 200,
  "data": {
    "_id": "...",
    "user": "...",
    "question": "What is a closure in JavaScript?",
    "answer": "A closure is a function that has access to...",
    "topic": "general",
    "createdAt": "..."
  },
  "message": "AI response generated successfully",
  "success": true
}

Video Doubt — Request Body

{
  "videoId": "60d21b4667d0d8992e610c85",
  "question": "What did the speaker say about React hooks?"
}

4. Database Schema (MongoDB)

4.1 Entity-Relationship Diagram

erDiagram
    USER ||--o{ VIDEO : owns
    USER ||--o{ COMMENT : writes
    USER ||--o{ LIKE : gives
    USER ||--o{ SUBSCRIPTION : subscribes
    USER ||--o{ SUBSCRIPTION : "is subscribed to"
    USER ||--o{ CHATMESSAGE : sends
    USER ||--o{ DOUBT : asks
    VIDEO ||--o{ COMMENT : has
    VIDEO ||--o{ LIKE : receives
    VIDEO ||--o{ CHATMESSAGE : has
    VIDEO ||--o{ TRANSCRIPTCHUNK : has

    USER {
        ObjectId _id PK
        String username UK
        String email UK
        String password
        String fullname
        Object avtar "url + public_id"
        Object coverImage "url + public_id"
        ObjectId[] watchHistory "ref Video"
        String refreshToken
    }

    VIDEO {
        ObjectId _id PK
        Object videoFile "url + public_id"
        Object thumbnail "url + public_id"
        String title
        String description
        Number duration
        Number views
        ObjectId owner FK "ref User"
        String transcript
        Boolean isPublished
    }

    COMMENT {
        ObjectId _id PK
        ObjectId video FK "ref Video"
        ObjectId owner FK "ref User"
        String content
    }

    LIKE {
        ObjectId _id PK
        ObjectId video FK "ref Video"
        ObjectId likes FK "ref User"
    }

    SUBSCRIPTION {
        ObjectId _id PK
        ObjectId subscriber FK "ref User"
        ObjectId channel FK "ref User"
    }

    CHATMESSAGE {
        ObjectId _id PK
        ObjectId video FK "ref Video"
        ObjectId user FK "ref User"
        String message
    }

    DOUBT {
        ObjectId _id PK
        ObjectId user FK "ref User"
        String question
        String answer
        String topic "default: general"
    }

    TRANSCRIPTCHUNK {
        ObjectId _id PK
        ObjectId video FK "ref Video"
        String text
        Number[] embedding
    }

4.2 Model Details

Model	File	Key Fields	Indexes
User	`src/models/user.model.js`	username, email, password (bcrypt hashed), avtar, coverImage, watchHistory, refreshToken	`username` (unique, indexed)
Video	`src/models/video.model.js`	videoFile, thumbnail, title, description, duration, views, owner, transcript	`owner: 1`, `createdAt: -1`
Comment	`src/models/comment.model.js`	video, owner, content	—
Like	`src/models/like.model.js`	video, likes (user ref)	—
Subscription	`src/models/subscription.model.js`	subscriber, channel (both User refs)	—
ChatMessage	`src/models/chatmessage.model.js`	video, user, message	—
Doubt	`src/models/dout.model.js`	user, question, answer, topic	—
TranscriptChunk	`src/models/transcriptChunk.model.js`	video, text, embedding (Number[])	—

4.3 User Model — Hooks & Methods

Feature	Description
Pre-save hook	Automatically hashes password with bcrypt (salt rounds = 10) before saving, only when `password` field is modified
`isPasswordCorrect(password)`	Compares plaintext password against stored hash using `bcrypt.compare()`
`generateAccessToken()`	Creates JWT with `{ _id, email, username, fullname }`, signed with `ACCESS_TOKEN_SECRET`, expires in `1d`
`generateRefreshToken()`	Creates JWT with `{ _id }`, signed with `REFRESH_TOKEN_SECRET`, expires in `10d`

5. Authentication Flow

sequenceDiagram
    participant C as Client
    participant S as Server
    participant DB as MongoDB

    Note over C,S: Registration
    C->>S: POST /api/v1/users/register (form-data)
    S->>S: Validate fields, check duplicates
    S->>S: Upload avtar + coverImage to Cloudinary
    S->>DB: Create User (password auto-hashed via pre-save hook)
    S->>C: 201 { user } (no password/refreshToken)

    Note over C,S: Login
    C->>S: POST /api/v1/users/login { email, password }
    S->>DB: Find user by email/username
    S->>S: bcrypt.compare(password)
    S->>S: Generate accessToken (1d) + refreshToken (10d)
    S->>DB: Save refreshToken on User document
    S->>C: 200 { user, accessToken, refreshToken } + httpOnly cookies

    Note over C,S: Authenticated Request
    C->>S: GET /api/v1/users/current-user (Bearer token)
    S->>S: jwtverify middleware decodes token
    S->>DB: Find user by decoded._id
    S->>S: Attach user to req.user
    S->>C: 200 { user }

    Note over C,S: Token Refresh
    C->>S: POST /api/v1/users/refresh-token
    S->>S: Verify refreshToken from cookie/body
    S->>DB: Match against stored refreshToken
    S->>S: Generate new accessToken + refreshToken
    S->>C: 200 { newAccessToken, newRefreshToken }

    Note over C,S: Logout
    C->>S: POST /api/v1/users/logout (Bearer token)
    S->>DB: Unset refreshToken from User document
    S->>S: Clear accessToken & refreshToken cookies
    S->>C: 200 { "user log out" }

Token Config:

Access Token: signed with ACCESS_TOKEN_SECRET, expires in 1d
Refresh Token: signed with REFRESH_TOKEN_SECRET, expires in 10d
Both set as httpOnly, secure cookies

6. AI Pipeline

6.1 General Doubt Resolution (Working ✅)

graph LR
    Q["Student Question"] --> Groq["Groq API (LLaMA 3.3 70B)"]
    Groq --> A["AI Answer"]
    A --> DB["Save to Doubt collection"]
    DB --> Res["Return to student"]

Uses Groq SDK with llama-3.3-70b-versatile model
System prompt: "You are a helpful learning assistant that explains programming concepts clearly."
Each Q&A is persisted in the Doubt collection for history & PDF export

6.2 Video-Specific RAG Pipeline (Planned, not fully wired)

graph TB
    Upload["Video Upload"] --> FFmpeg["Extract Audio (FFmpeg)"]
    FFmpeg --> Whisper["Speech-to-Text (OpenAI Whisper)"]
    Whisper --> Transcript["Full Transcript"]
    Transcript --> Split["Split into 200-word chunks"]
    Split --> Embed["Generate Embeddings (OpenAI)"]
    Embed --> Store["Store TranscriptChunks in MongoDB"]
    
    Question["Student Question"] --> QEmbed["Embed Question"]
    QEmbed --> Search["Cosine Similarity Search"]
    Search --> BestChunk["Most Relevant Chunk"]
    BestChunk --> LLM["Groq LLaMA 3.3"]
    LLM --> Answer["Context-aware Answer"]

Supporting Services:

Service	File	Technology	Purpose
AI	`src/service/ai.service.js`	Groq SDK (LLaMA 3.3)	Generate AI responses
Audio	`src/service/audio.service.js`	fluent-ffmpeg	Extract audio from video files
Transcript	`src/service/transcript.service.js`	OpenAI Whisper	Audio → text transcription
Embedding	`src/service/embedding.service.js`	OpenAI `text-embedding-3-small`	Generate vector embeddings
PDF	`src/service/pdf.service.js`	PDFKit	Export doubt history as PDF

Supporting Utilities:

src/utils/cosineSimilarity.js — Vector similarity calculation
src/utils/splitTranscript.js — Chunk transcript into 200-word blocks

7. Real-time Chat (Socket.io)

File: src/Socket.io/socket.js

Events

Event	Direction	Payload	Description
`connection`	Server ← Client	—	New client connected
`join-video-room`	Server ← Client	`videoId`	Join a video's chat room
`recived-message`	Server → Room	`{ user, message }`	Broadcast message to room
`disconnect`	Server ← Client	—	Client disconnected

Client Usage Example

import { io } from "socket.io-client";

const socket = io("http://localhost:4000");

// Join a video room
socket.emit("join-video-room", "VIDEO_ID_HERE");

// Listen for messages
socket.on("recived-message", (data) => {
  console.log(`${data.user}: ${data.message}`);
});

// Send a message (when implemented)
socket.emit("send-message", { 
  videoId: "VIDEO_ID_HERE", 
  user: "username", 
  message: "Hello!" 
});

⚠️ Warning: The Socket.io implementation is incomplete. The send-message listener is missing — the io.to(videoId).emit(...) call is outside any event handler and references undefined variables. Messages are not being persisted to the ChatMessage model.

8. Middleware Layer

8.1 JWT Authentication — `src/middleware/auth.middleware.js`

Extracts token from cookies.accessToken or Authorization: Bearer <token> header
Verifies with ACCESS_TOKEN_SECRET
Looks up user in DB, excludes password & refreshToken
Attaches user object to req.user

8.2 File Upload — `src/middleware/multer.js`

Uses multer.diskStorage to store uploaded files
Destination: ./public directory
Filename: Original filename preserved
Used for avatar, cover image, video, and thumbnail uploads

8.3 Rate Limiters — `src/middleware/ratelimmiter/ratelimmiter.js`

Limiter	Window	Max Requests	Applied To
`ApiRatelimiter`	15 min	100	General API (not currently used)
`registerratelimt`	15 min	5	`/register`
`logginratelimit`	5 min	10	`/login`

8.4 Security — Helmet

Enabled globally via app.use(helmet()) to set secure HTTP headers
Protects against common vulnerabilities (XSS, clickjacking, etc.)

9. Utilities

Utility	File	Purpose
ApiError	`src/utils/ApiError.js`	Custom error class with `statuscode`, `message`, `errors[]` — extends native `Error`
ApiResponse	`src/utils/ApiResponse.js`	Standardized response: `{ statuscode, data, message, success }` where `success = statuscode < 400`
asynchandler	`src/utils/asynchandler.js`	Higher-order function that wraps async route handlers, catches Promise rejections and forwards to `next(err)`
uploadOnCloudinary	`src/utils/cloudinary.js`	Uploads local file to Cloudinary with `resource_type: "auto"`, deletes local copy after upload, returns Cloudinary response
deletefilefromCloudinary	`src/utils/deletefilefromCloudinary.js`	Delete file from Cloudinary by `public_id`
cosineSimilarity	`src/utils/cosineSimilarity.js`	Compute cosine similarity between two embedding vectors (for RAG search)
splitTranscript	`src/utils/splitTranscript.js`	Split transcript text into 200-word chunks for embedding generation

10. Environment Variables

Create a .env file in the project root with the following variables:

Variable	Purpose	Example
`PORT`	Server port	`4000`
`MONGO_URI`	MongoDB Atlas connection string	`mongodb+srv://user:pass@cluster.mongodb.net`
`CROSS_ORI`	CORS allowed origins	`*`
`ACCESS_TOKEN_SECRET`	JWT access token signing key	`your-secret-key`
`ACCESS_TOKEN_EXPIRY`	Access token TTL	`1d`
`REFRESH_TOKEN_SECRET`	JWT refresh token signing key	`your-secret-key`
`REFRESH_TOKEN_EXPIRY`	Refresh token TTL	`10d`
`CLOUDINARY_API_KEY`	Cloudinary API key	`582355621587992`
`CLOUDINARY_API_SECRET`	Cloudinary API secret	`your-cloudinary-secret`
`CLOUDINARY_URL`	Cloudinary full URL	`cloudinary://key:secret@cloud-name`
`GROQ_API_KEY`	Groq API key for LLaMA 3.3	`gsk_...`
`OPENAI_API_KEY`	OpenAI API key (for Whisper & Embeddings)	`sk-...`

⚠️ Caution: OPENAI_API_KEY is referenced in embedding.service.js and transcript.service.js but is not defined in the current .env. The RAG pipeline will fail without it.

11. Known Bugs & Issues

⚠️ Important: The following issues were identified during static code analysis. They should be addressed before production deployment.

Critical Bugs

#	File	Issue
1	`src/middleware/auth.middleware.js` (L7)	`jwtverify` uses `asynchandler` but signature is `(req, res)` — missing `next` parameter. `next()` is called on line 26 but it's undefined.
2	`src/controller/ai.controller.js` (L83)	`generateNotes` has `(res, req)` — parameters are swapped.
3	`src/controller/ai.controller.js` (L90)	`if(doubts)` should be `if(!doubts.length)` — condition is inverted.
4	`src/controller/ai.controller.js` (L107)	`Video.find(ById(videoId))` — syntax error, should be `Video.findById(videoId)`.
5	`src/controller/ai.controller.js` (L124)	`bestChunk` is referenced but never defined.
6	`src/controller/like.controller.js` (L28)	`res.stauts(200),json(...)` — typo (`stauts` → `status`) and comma instead of dot.
7	`src/controller/like.controller.js` (L18)	Querying `likedBy` field but model uses `likes`.
8	`src/controller/subscription.controller.js` (L46)	Returns `newSubscription` but variable is named `newSubscriber`.
9	`src/controller/video.controller.js` (L98)	`Like` is used but never imported.
10	`src/controller/video.controller.js` (L101)	`console(...)` — missing `.log`.
11	`src/Socket.io/socket.js` (L13)	`io.to(videoId).emmit(...)` — typo (`emmit` → `emit`) and placed outside event handler; `videoId`, `user`, `message` are undefined.
12	`src/utils/splitTranscript.js` (L2)	Splits on empty string `""` instead of space `" "`.
13	`src/utils/splitTranscript.js` (L7)	Infinite loop: `i+chunkSize` doesn't mutate `i` — should be `i+=chunkSize`.
14	`src/service/pdf.service.js` (L17-22)	Uses `iteam` as parameter but `item` in the body — `ReferenceError`.
15	`src/models/user.model.js` (L29)	`reuired` typo — should be `required`.
16	`src/middleware/ratelimmiter/ratelimmiter.js`	`windowsMs` — should be `windowMs` (express-rate-limit API).

Architectural Issues

#	Issue	Recommendation
1	Video upload route field names (`videoupload`, `thumbnailupload`) don't match controller expectations (`video`, `thumbnail`)	Align multer field names with `req.files` access patterns
2	No validation library (Joi, Zod) — all validation is manual	Add schema validation for request bodies
3	User registration stores `avtar.url` as a string but model defines it as `{ url, public_id }` object	Fix registration controller to save the full object
4	`DB_NAME` is extracted but never used in the connection string	Append `/${DB_NAME}` to `MONGO_URI`
5	No `OPENAI_API_KEY` in `.env` — transcript & embedding services will crash	Add the key or disable unused services
6	The `ApiRatelimiter` (general) is defined but never mounted	Mount it globally or remove it
7	Comment route path is `/commnets` (typo)	Fix to `/comments`

12. Dependencies

Package	Version	Purpose
`express`	5.2.1	Web framework
`mongoose`	9.2.0	MongoDB ODM
`socket.io`	4.8.3	Real-time WebSocket communication
`jsonwebtoken`	9.0.3	JWT generation & verification
`bcrypt`	6.0.0	Password hashing
`cloudinary`	2.9.0	Cloud file storage (images, videos)
`multer`	2.0.2	Multipart form-data file uploads
`cors`	2.8.6	Cross-Origin Resource Sharing
`helmet`	8.1.0	HTTP security headers
`cookie-parser`	1.4.7	Parse cookies from requests
`express-rate-limit`	8.2.1	API rate limiting
`groq-sdk`	1.1.1	Groq AI (LLaMA) integration
`@google/generative-ai`	0.24.1	Google Gemini (commented out, unused)
`openai`	6.31.0	OpenAI Whisper + Embeddings
`fluent-ffmpeg`	2.1.3	Audio extraction from video
`ffmpeg-static`	5.3.0	Static FFmpeg binary
`pdfkit`	0.18.0	PDF generation
`axios`	1.13.6	HTTP client (unused currently)
`dotenv`	17.2.4	Environment variable loader
`nodemon`	3.1.11	Dev: auto-restart on file changes

13. How to Run

# 1. Clone the repository
git clone <repo-url>
cd Backend

# 2. Install dependencies
npm install

# 3. Create .env with required variables (see Section 10)
cp .env.example .env  # then fill in your values

# 4. Start development server
npm run dev
# → Server runs on http://localhost:4000

The dev script: nodemon -r dotenv/config --experimental-json-modules src/index.js

Testing with Postman

Register → POST http://localhost:4000/api/v1/users/register (form-data with avatar & cover image files)
Login → POST http://localhost:4000/api/v1/users/login (get access token from response)
Set Auth Header → Authorization: Bearer <accessToken>
Upload Video → POST http://localhost:4000/api/v1/video/upload (form-data with video & thumbnail files)
Ask AI → POST http://localhost:4000/api/v1/ai/ask (JSON body with question)

License

ISC

Documentation generated on March 20, 2026

Name		Name	Last commit message	Last commit date
Latest commit History 59 Commits
Readme		Readme
public		public
src		src
.gitignore		.gitignore
README.md		README.md
constant.js		constant.js
package-lock.json		package-lock.json
package.json		package.json

Folders and files

Latest commit

History

Repository files navigation

VideoTube Backend — Complete Documentation

Table of Contents

1. Product Requirements Document (PRD)

1.1 Vision

1.2 Core Features

1.3 Target Users

2. System Architecture

2.1 High-Level Architecture

2.2 Request Lifecycle

2.3 Directory Structure

3. API Reference

3.1 Users — /api/v1/users

Register — Request Body (multipart/form-data)

Login — Request Body

Login — Response

Change Password — Request Body

Update Account — Request Body

3.2 Videos — /api/v1/video

Upload Video — Request Body (multipart/form-data)

Upload Video — Response

Get All Videos — Response

3.3 Likes — /api/v1/likes

3.4 Comments — /api/v1/commnets

Add Comment — Request Body

Get Comments — Response

3.5 Subscriptions — /api/v1/subscriptions

3.6 Chat — /api/v1/chat

3.7 AI — /api/v1/ai

Ask AI — Request Body

Ask AI — Response

Video Doubt — Request Body

4. Database Schema (MongoDB)

4.1 Entity-Relationship Diagram

4.2 Model Details

4.3 User Model — Hooks & Methods

5. Authentication Flow

6. AI Pipeline

6.1 General Doubt Resolution (Working ✅)

6.2 Video-Specific RAG Pipeline (Planned, not fully wired)

7. Real-time Chat (Socket.io)

Events

Client Usage Example

8. Middleware Layer

8.1 JWT Authentication — src/middleware/auth.middleware.js

8.2 File Upload — src/middleware/multer.js

8.3 Rate Limiters — src/middleware/ratelimmiter/ratelimmiter.js

8.4 Security — Helmet

9. Utilities

10. Environment Variables

11. Known Bugs & Issues

Critical Bugs

Architectural Issues

12. Dependencies

13. How to Run

Testing with Postman

License

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

3.1 Users — `/api/v1/users`

3.2 Videos — `/api/v1/video`

3.3 Likes — `/api/v1/likes`

3.4 Comments — `/api/v1/commnets`

3.5 Subscriptions — `/api/v1/subscriptions`

3.6 Chat — `/api/v1/chat`

3.7 AI — `/api/v1/ai`

8.1 JWT Authentication — `src/middleware/auth.middleware.js`

8.2 File Upload — `src/middleware/multer.js`

8.3 Rate Limiters — `src/middleware/ratelimmiter/ratelimmiter.js`

Packages