PeerPrep — Personalized DSA Preparation & Reflection Platform

A full-stack web application that helps developers track DSA problem-solving progress, write structured reflections, and receive AI-powered adaptive review scheduling based on sentiment and confidence analysis.

---

Features

Authentication

• OTP-based two-factor email verification at signup and login (bcrypt + nodemailer)
• Password reset flow with OTP validation and strength enforcement
• JWT-based session management with 7-day token expiry
• Auto-expiring OTP cleanup every 10 minutes

Problem Tracking

• Browse curated DSA sheets (NeetCode 150, Striver SDE Sheet, and more)
• Mark problems as Solved, Attempted, or Unsolved with toggle support
• Paginated problem lists with difficulty filtering
• Per-sheet and overall progress summaries

Reflection & NLP Analysis

• Write post-solve reflections for any problem
• Sentiment classification using HuggingFace (cardiffnlp/twitter-roberta-base-sentiment)
• Keyword extraction from reflection text
• Confidence scoring per reflection

Adaptive Spaced Repetition

• Review date computed from sentiment + problem difficulty + confidence score
• Negative sentiment → review in 1 day; Neutral → 3 days; Positive → 5 days
• Difficulty adjustment: Easy (+5 days), Medium (+2–4 days), Hard (+2 days)
• Low confidence (<0.5) adds 1 extra day
• Today's review queue with completion tracking

Analytics Dashboard

• Time-series progress charts with moving-average smoothing (configurable window)
• Sentiment trend analysis over configurable date ranges
• Upcoming review load view (next N days)
• Overall solved/attempted/unsolved breakdown

Real-Time

• WebSocket server with JWT-authenticated connections
• Invalid token → immediate connection rejection

---

Tech Stack

Layer	Technology
Frontend	React (Vite), React Router, Context API, Axios
Backend	Node.js, Express.js
Database	MySQL (connection pool, limit 10)
Auth	bcryptjs, jsonwebtoken, nodemailer
NLP	HuggingFace Inference API (@huggingface/inference)
Real-Time	WebSocket (ws)
Other	keyword-extractor, dotenv, cors

---

Project Structure

Peerprep/
├── new_frontend/
│   └── src/
│       ├── components/
│       │   ├── Auth/          # Signup, Login, VerifyOTP, ForgotPassword
│       │   ├── Dashboard/     # Dashboard, DSASheets, ProgressChart, NudgeCard, ReviewCard
│       │   ├── Problems/      # Problems list with filtering and progress toggle
│       │   ├── Analysis/      # Analytics and sentiment dashboards
│       │   └── Shared/        # Navbar, ProtectedRoute
│       ├── context/           # AuthContext (global auth state)
│       └── api/               # Axios instance and API helpers
│
└── peerprep-backend/
    ├── controllers/           # authController, otpController, reflectionController,
    │                          # userProgressController, analysisController, sheetsController
    ├── routes/
    │   ├── private/           # JWT-protected: auth, progress, reflections, analysis
    │   └── public/            # Open: sheets, problems
    ├── middleware/            # authMiddleware (JWT verification)
    ├── Database/
    │   ├── Schema.sql         # Core tables
    │   └── Users_Schmea.sql   # Users, progress, reflections, OTP tables
    ├── db.js                  # MySQL connection pool
    └── server.js              # Express app + WebSocket server

---

Database Schema

-- Core tables
problems        (problem_id, title, URL, platform, difficulty, category)
sheets          (sheet_id, name, description)
problem_sheets  (id, problem_id, sheet_id, sheet_order, sheet_title)

-- User tables
users           (user_id, username, email, user_password, first_name, last_name,
                 user_role ENUM, acc_status ENUM, time_created, time_updated)
user_progress   (progress_id, user_id, problem_id,
                 problem_status ENUM('Solved','Attempted','Unsolved'),
                 time_created, time_updated)
                 UNIQUE (user_id, problem_id)

-- Reflection & review
reflections     (reflection_id, user_id, problem_id, reflection_text,
                 sentiment ENUM('Positive','Neutral','Negative'),
                 keywords, confidence_score, review_date, completed,
                 created_at, updated_at)
                 UNIQUE (user_id, problem_id)

otp_verifications (id, email, otp, created_at, expires_at)

---

API Endpoints

Public

Method	Route	Description
GET	/sheets	Get all DSA sheets
GET	/problems	Get problems (public)

Auth (/auth)

Method	Route	Description
POST	/auth/send-otp	Send OTP to email
POST	/auth/verify-otp	Verify OTP
POST	/auth/reset-password	Reset password with OTP
POST	/auth/signup	Register (requires OTP)
POST	/auth/login	Login (issues JWT)
GET	/auth/profile	Get profile (protected)

Progress (/progress) — JWT required

Method	Route	Description
GET	/progress/sheet/:sheet_id	Problems + user progress for a sheet
POST	/progress/update	Update problem status
GET	/progress/summary	Per-sheet progress summary
GET	/progress/user/:user_id/sheets	Sheet progress for user

Reflections (/reflections) — JWT required

Method	Route	Description
POST	/reflections/	Add/update reflection (triggers NLP)
GET	/reflections/	Get all reflections for user
GET	/reflections/:problem_id	Get reflection for one problem
PUT	/reflections/:problem_id	Re-analyze and update reflection
DELETE	/reflections/:problem_id	Delete reflection
GET	/reflections/today	Get today's review queue
GET	/reflections/all	Full review history
PATCH	/reflections/completion	Update review completion status

Analysis (/analysis) — JWT required

Method	Route	Description
GET	/analysis/progress	Time-series progress (filter by sheet/days)
GET	/analysis/sentiment	Sentiment trend with moving average
GET	/analysis/review-status	Upcoming review load

---

Getting Started

Prerequisites

• Node.js v18+
• MySQL 8+
• HuggingFace account (for NLP API token)
• Gmail account (for OTP emails)

Backend Setup

cd peerprep-backend
npm install

Create a .env file:

DB_HOST=localhost
DB_USER=your_mysql_user
DB_PASS=your_mysql_password
DB_NAME=peerprep

JWT_SECRET=your_jwt_secret

EMAIL_USER=your_gmail@gmail.com
EMAIL_PASS=your_gmail_app_password

HF_TOKEN=your_huggingface_token

Run the database schema files in order:

mysql -u root -p peerprep < Database/Schema.sql
mysql -u root -p peerprep < Database/Users_Schmea.sql
mysql -u root -p peerprep < Database/Data/data.sql

Start the backend:

npm start
# Server runs at http://localhost:5000

Frontend Setup

cd new_frontend
npm install
npm run dev
# App runs at http://localhost:5173

---

How Spaced Repetition Works

When a user submits a reflection, the system computes a review_date automatically:

base_days          = sentiment score  → Negative: 1, Neutral: 3, Positive: 5
difficulty_days    = problem difficulty → Easy: +5, Medium: +2-4, Hard: +2
confidence_penalty = confidence < 0.5  → +1 day

review_date = today + base_days + difficulty_days + confidence_penalty

On the review date, the problem appears in the user's daily review queue. Completing it marks it done; incomplete reviews stay visible until resolved.

---