About

It takes about 4-5 days for leetcode to update the contest ratings of participants. So you have to wait for a long time to know your rating changes. This application predicts accurate leetcode rating changes for all the contestants within a few minutes of completion of the contest.

Getting started

This project consists of two types of user interfaces. You can either use browser extension or the website to get your rating changes.

Chrome extension

You can install the extension from Chrome Web Store. It adds the rating changes on leetcode ranking pages itself.

Website

You can also visit lcpredictor.onrender.com to get your rating changes.

How It Works

This project is written in Node + MongoDB + Redis tech stack. We can divide it into three microservices.

Backgound

It is the most important part of the project. It's job is to fetch the data from leetcode and predict the contest ratings periodically.

Rating prediction

Rating prediction is a cpu intensive task. Therefore a nodejs C++ addon is implemented for this task so that we can utilize threading with better performance using C++. For performance measurement we got these results :

Machine configuration :

It implements leetcode's latest rating prediction algorithm. Rating predictions are very close to the original rating but the accuracy may not be 100% due to changes in contest rankings after the completion of contest (leetcode rejudges some submissons).

These are the results for the predictions of weekly-contest-242:

Job scheduling

Job scheduling is required for processing jobs on desired time. Leetcode contests are weekly and biweekly. We can schedule them by scheduling a repeated job. But for making it more generic, job schedulers are implemented who schedules prediction and data update jobs. These job schedulers are scheduled as a repeated job. It is accomplished by using bull, a redis based queue for Node. A bull dashboard is also integrated using bull-board.

Website

It is built using express framework. Ejs is used for writing templates. It contains a table for contests and ranking pages with predicted rating changes for all the contests. Pagination is added to ranking pages for better user experience and performace.

API

It is also implemented using express framework. It contains an endpoint for fetching users' predicted rating changes which is used in the browser extension.

IP based rate limit is enforced for both the API and the website using express-rate-limit.

Development

Setup

• Clone the repository

  git clone https://github.com/SysSn13/leetcode-rating-predictor

• Install the dependencies

  npm install

• Setup environment variables

  cp .env.example .env

  Fill in the required values in the .env file.

• Build the predict-addon (if you are using different node version)

  npm run buildAddon

• Start the project

  npm start

• Or start the development server by:

  npm run dev

Environment variables

DATABASE_URL: Connection string for mongodb.

# for web
WEB: Whether to run the website or not. (0 or 1)

RATE_LIMIT_WINDOW: Window size for rate limit in milliseconds (default: 10000).

RATE_LIMIT: Number of requests allowed in the window (default: 50).


# for api
API_DISABLED: Whether to disable the API or not. (0 or 1)

API_RATE_LIMIT_WINDOW: Window size for API rate limit in milliseconds (default: 10000).

API_RATE_LIMIT: Number of API requests allowed in the window (default: 20).


# for background
BACKGROUND: Whether to run the background or not. (0 or 1)

REDIS_URL: Connection string for redis.

THREAD_CNT: Number of threads for prediction.(default: 4)

# bull-board auth
BULLBOARD_USERNAME: username for bull-board login

BULLBOARD_PASS: password for bull-board login

SESSION_SECRET: secret to hash the session

Browser extension

Current only chrome browser is supported. It uses manifest V3. See this for getting started with extension development. Source code for the extension is in ./chrome-extension.

Contributing

You can contribute by creating issues, feature/ pull requests. Any meaningful contributions you make are greatly appreciated.

For contributing in the source code, please follow these steps:

• Fork the Project
• Create your new Branch

  git checkout -b feature/AmazingFeature

• Commit your Changes

  git commit -m 'Add some AmazingFeature'
  

• Push to the Branch

  git push origin feature/AmazingFeature
  

• Open a Pull Request

Contributors

Stargazers

License

Distributed under the MIT License. See LICENSE for more information.