⚡ Leet-Search – Semantic LeetCode Search Engine

Find LeetCode problems by meaning — not just by name.

Have you ever faced a situation where you vaguely remembered a LeetCode problem you once solved, but couldn’t recall its exact title? Or come across a question in an online assessment (OA) that felt familiar — like a slightly tweaked version of a LeetCode problem — but no amount of keyword guessing could help you find it again?

Leet-Search is built to solve exactly that.

A semantic search engine for LeetCode problems — built with FastAPI, Next.js, and pgvector.Leet-Search understands the meaning and context behind queries, allowing users to find relevant problems based on intent rather than exact wording. Whether you're preparing for coding interviews or exploring new problem types, Leet-Search helps you discover the right problems faster and more intuitively.

🔗 Live App: leet-search-gray.vercel.app

📦 Repository: github.com/Eli4479/Leet-Search

---

🧠 Features

• 🔍 Semantic Search — vector-based querying of LeetCode problem descriptions
• 🧩 Modular FastAPI Architecture — clean separation of concerns (routes, services, controllers)
• 🔗 Supabase — hosted PostgreSQL with pgvector

---

📂 Folder Structure

leet-search/
├── backend/
│   ├── app/
│   │   ├── controllers/          # Handles endpoint logic
│   │   ├── database/             # DB connection + Supabase client
│   │   ├── models/               # Pydantic schemas & data models
│   │   ├── routes/               # API route registration
│   │   ├── services/             # Core business logic
│   │   ├── utils/                # Helper utility functions
│   │   └── main.py               # FastAPI app entry point
│   ├── .env                      # Environment variables
│   └── requirements.txt          # Python dependencies
│
├── frontend/
│   ├── .next/                    # Next.js build output
│   ├── node_modules/
│   ├── public/                   # Static assets
│   ├── src/                      # All frontend source code
│   ├── .gitignore
│   ├── components.json
│   ├── eslint.config.mjs
│   ├── next-env.d.ts
│   ├── next.config.ts
│   ├── package-lock.json
│   ├── package.json
│   ├── postcss.config.mjs
│   ├── .env
│   ├── README.md
│   └── tsconfig.json
├── .gitignore
└── README.md                     #  Project documentation

---

🔐 Environment Variables

In your .env file in frontend (excluded via .gitignore), define:

NEXT_PUBLIC_API_URL=http://127.0.0.1:8000/api

In your .env file in backend (excluded via .gitignore), define:

SUPABASE_URL=https://your-project.supabase.co
SUPABASE_KEY=your-supabase-service-role-key

---

🧪 API Example

Endpoint

POST /api/search?page=1

Request Body

{
  "query": "graph dfs",
  "limit": 5
}

Response

[
  {
    "id": "1472",
    "title": "Design Browser History",
    "url": "https://leetcode.com/problems/design-browser-history/",
    "paid_only": false,
    "match_percentage": 77.24,
    "content": "formated HTML content of the problem.",
    "original_content": "original HTML content of the problem."
  },
  ...
]

💡 Explanation of Response Fields:

Field	Description
id	LeetCode problem ID
title	Name of the problem
url	Direct LeetCode URL
paid_only	Whether the problem is paid-only
match_percentage	Similarity score with query (0–100%)
content	Formatted HTML content of the problem
original_content	Original HTML content (for reference)

---

🛠 Tech Stack

Tool/Service	Description
🐍 FastAPI	Fast Python web framework (ASGI-compatible)
🔗 Supabase	PostgreSQL + pgvector for embeddings
🚀 Next.js	React framework for the frontend
💎 shadcn/ui	Beautiful UI components for Next.js

---

🤝 Contributing

We’re thrilled you’re considering contributing to Leet-Search — a fast, intelligent semantic search engine for LeetCode problems!

Whether you're:

• Fixing bugs 🐛
• Improving performance ⚡
• Enhancing the UI/UX 💅
• Refining search algorithms 🔍
• Writing tests or documentation 🧪📝
• Your contribution matters 💪🏻

🙋‍♂️ How to Contribute

1. Fork the repo and create your branch:

   git checkout -b feature/your-feature-name

2. Commit your changes:

   git commit -m "feat: brief description of your change"

3. Push to GitHub:

   git push origin feature/your-feature-name

4. Open a Pull Request here 🚀

Found a bug or have a feature idea? Open an Issue – we’d love to discuss it with you!

🧠 Ideas to Contribute

• 🔥 Add support for more problem platforms (e.g., Codeforces, GFG)
• ✨ Enhance the frontend UX (Next.js + Tailwind)
• 🧠 Improve vector similarity logic or embeddings
• 🧪 Write unit tests for routes/services
• 🛡 Add rate limiting, error logging, or caching

---

🛠 Getting Started (Local Setup)

Want to run Leet-Search on your machine? Here's how:

🔁 Clone the Repository

git clone https://github.com/Eli4479/Leet-Search.git
cd Leet-Search

🔧 Backend Setup (FastAPI + Supabase)

1. Navigate to backend and create a virtual environment:

   cd backend
   python3 -m venv venv
   source venv/bin/activate  # Windows: venv\Scripts\activate

2. Install dependencies:

   pip install -r requirements.txt

3. Create a .env file:

   SUPABASE_URL=https://your-project.supabase.co
   SUPABASE_KEY=your-supabase-service-role-key

4. Start the FastAPI server:

   uvicorn app.main:app --reload

   API will run at: http://127.0.0.1:8000

🌍 Frontend Setup (Next.js + Tailwind)

1. Navigate to frontend and install packages:

   cd ../frontend
   npm install

2. Start the frontend dev server:

   npm run dev

   Your app will run at: http://localhost:3000

📦 Populate the Database

Prepare the LeetCode problem embeddings and load them into Supabase:

1️⃣ Create the Table in Supabase

• Open your Supabase project
• Go to SQL Editor
• Open the file backend/scripts/sql.txt
• Copy its contents and run the query

This will create the problems_bge table with the necessary columns (including vector type for embeddings).

2️⃣ Edit main.py

Open backend/app/main.py and un-comment the following line:

# populate_db()

Change it to:

populate_db()

3️⃣ Run the Backend Server

cd backend
uvicorn app.main:app --reload

This will:

• Fetch LeetCode problems (free + paid)
• Format and clean problem content
• Generate embeddings using OpenAI (or your configured model)
• Save the output to: backend/scripts/problems.csv

4️⃣ Upload CSV to Supabase

• Go to your Supabase Table Editor
• Select the problems_bge table
• Click "Import Data"
• Upload the generated CSV file from:

backend/scripts/problems.csv

✅ NOTE: After populating, you have to comment out populate_db() again to avoid re-triggering on future backend runs.

✅ Test It Works

1. Visit http://localhost:3000
2. Type a query like graph dfs
3. Confirm the results load from your FastAPI backend

🧑‍💻 Dev Tips

• Format code before committing (black, prettier, etc.)
• Use conventional commit messages (e.g. feat:, fix:, docs:)
• Open small, focused PRs with clear descriptions

---

👋 Final Note

Whether you're here to contribute, learn, or get inspired — thank you for checking out Leet-Search!

Built with ❤️ and ☕ by Aryan Patel