cpt-lib

Short for competitive programming tools library, cpt-lib is a collection of API wrappers to request and upload data to various competitive programming websites, enabling the extraction and processing of a myriad of data with relative ease.

Make sure to star ⭐ the project if you found it useful. 😄

Table of Contents

• Overview
• Supported Websites
• Getting Started
  • Installation
  • Usage
• Contributing
• FAQ

Overview

cpt-lib is a library that uses browser automation, to extract, process and automate various processes related to competitive programming websites. Built as an API wrapper, it can perform many different tasks, of which notable ones are:

• Fetching sample tests of problems.
• Submitting solution to remote judge.
• Returning dynamic status of submissions.
• Extracting public details of contests.
• Fetching submissions and its solution code.

Obviously, some websites have more features, while some have only a subset of these. To know all available functions, refer to the respective package documentation (refer below).

Supported Websites

Website	Support	Status
CodeForces	⭐⭐⭐⭐⭐	Is supported
Atcoder	⭐⭐⭐⭐	In development
USACO	⭐⭐⭐	In development
Codechef	⭐⭐⭐	Future milestone

Legend

• Support:

  • ⭐⭐⭐⭐⭐ high priority and long term support.
  • ⭐⭐⭐⭐ medium priority with long term support.
  • ⭐⭐⭐ medium priority with long term support.
  • ⭐⭐ low priority with only bug fixes.
  • ⭐ low priority, development stalled.

• Status:

  • Is supported - Website support available on master branch.
  • In development - Website support is in development branch.
  • Future milestone - Development underway in the near future.

Getting Started

For complete usage examples, view tests of the corresponding packages. Refer godoc badges above for corresponding documentation.

Installation

Usage is simple. First use go get to install the latest version of the library.

go get -u github.com/cp-tools/cpt-lib/v2

Next, include cpt-lib in your application.

import "github.com/cp-tools/cpt-lib/v2"

Usage

The examples below use the codeforces module, for illustration purposes.

The core functionalities of the library are achieved using browser automation, through the DevTools protocol. The package rod is used to control the automated browser.

To use the methods provided by the library, the automated browser must be initiated first. This can be done easily using the function provided in all sub packages - Start().

func main(){
    // Initialization parameters.
    inHeadless := true
    browser := "google-chrome"
    browserProfile := "/home/<username>/.config/google-chrome/"
    codeforces.Start(inHeadless, browserProfile, browser)

    // Do parsing here...
}

At the root, each package implements a Args type. This holds metadata of a contest/problem group, on which the methods are provided. Instantiating a variable of this type is done using the provided Parse() function, which casts the provided specifiers to the variable.

Specifiers supported by Parse() varies between websites, but URLs to the contest/problem are supported by all packages.

arg, err := codeforces.Parse("codeforces.com/contest/1234/problem/c")
if err != nil{
    panic(err)
}

// Run methods on 'arg' next...

The returned variable can then be used to execute the different provided methods, using its metadata. An example of fetching sample tests of the problem (specified in the previous snippet) is as follows:

problems, err := arg.GetProblems()
if err != nil{
    panic(err)
}

// Display information of fetched problems.
for _, problem := range problems {
    fmt.Println("Problem Name:", problem.Name)
    fmt.Println("Time limit:", problem.TimeLimit)
    fmt.Println("Number of sample tests:", len(problem.SampleTests))
}

FAQ

Which browsers are supported?

As the project directly uses rod to control the automated browser, all browsers supported by it are supported by this package. Nevertheless, here is the list of browser support:

• Supported browsers:
  • Google chrome (tested)
  • Chromium (tested)
  • Microsoft Edge (untested)
• Unsupported browsers:
  • Firefox (see issue here)
  • Safari
  • Opera
  • Internet Explorer

Is sensitive data of my browser at risk?

Short answer, No. The library doesn't access or modify any sensitive information, including browser cookies and login credentials.

The functioning of the the browser automation is as follows.

• Starts the specified browser, with the specified user data directory.
• Creates another new browser instance, with a different user data directory.
• Copies cookies data from the first browser instance to the second browser instance.
• Closes the former browser, and uses the latter browser profile to access the websites.

This ensures nothing (history, cookies, bookmarks etc) of your specified user profile are modified.

However, use unofficial versions of this library with extreme caution, as malicious code can be used to extract any private credentials stored on your browser.

How do I make this library work in other languages?

Currently, there is no official support for the same.

However, there are some future plans, which are as follows (sorted by priority):

• A cross platform command line executable.
• An online REST API to return data of public contests.

What are the benefits of using browser automation over fetching web pages with GET requests?

Here are the major plus sides of using browser automation over source code fetching:

1. Uses logged in session of specified browser, doing away with management of login sessions.
2. Improved security since no credentials/login information is given away.
3. Support for websites with dynamic loading of data (done through JavaScript).
4. No major difference in speed between the two methods (view test results for stats).
5. Support for monitoring and returning web socket controlled data (like submission status).
6. Anything that can be done manually, can be done easily with automated browsers.

And thus, the pros of this method clearly outweigh the cons, making this the best method.

---

However, if you wish to fallback to the older method, you may use the archived version:

go get -u github.com/cp-tools/cpt-lib/v1

Note that, there will not be any future updates or support for versions prior to v2.