krispykalsi/leet-scrape

A simple CLI to download and create the default solution code snippet or the question statement HTML page from https://leetcode.com

Leetscrape

An unofficial utility that can download Leetcode problems using its GraphQL API.

📜Examples in GIF 👆

• Question

leetscrape --daily-challenge question

leetscrape --name "two sum" question

• Solution

leetscrape --number 58 solution --lang C++

leetscrape --url "leetcode.com/problems/3sum" solution --lang java

More details in the Usage section.

⚙ Installation

• Using Go

  1. Make sure you have Go (1.16 or greater) installed with your $GOROOT environment variable correctly setup. (If you installed Go in the default location, then you don't need to worry about setting up $GOROOT)

  2. Run the following command in your shell

  go install github.com/ISKalsi/leet-scrape/v2/cmd/leetscrape@latest

🧰 Usage

For every problem you can either download its default code snippet or the question statement. This is done by using the following sub-commands -

• question or ques - This downloads the full question (including images) in an HTML file
• solution or sol - This downloads the default code snippet in the Leetcode's editor when you start solving a fresh problem. You have to specify the language using the --lang (or -l) flag (choose any one of the available options given in Leetcode's editor)

This tool provides four ways to find a problem -

Searching Method	Flag	Example
Name	--name or -n	leetscrape -n "two sum" sol python
Number	--number or -N	leetscrape -N 1 sol -l C++
URL	--url or -u	leetscrape -u "https://leetcode.com/problems/two-sum/" ques
Daily Challenge	--daily-challenge or -d	leetscrape -d ques

Note: providing one of the four methods is necessary.

Additional Flags

• --output-dir or -o Command: leetscrape Example: leetscrape -o random/path -n "two sum" ques By default the file is downloaded to the working directory. To provide a custom path, specify the flag and give a ** relative** path.
• --boilerplate or -b Command: solution Example: leetscrape -n "two sum" sol -l C++ -b "#include<string>\nusing namespace std;\n" This will prepend the given string to the code snippet before writing to file.

💭 QnA

• Why did I make this tool?

  Two Reasons -

  First: I solve my Leetcode problems in a local IDE instead of their embedded editor because it's so shitty for non-premium users (no formatting/autocomplete/debugging). So this is what used to happen - Go to the problem, copy the code in editor, make new file, paste, add #include headers and other stuff. I had to repeat all those things for every new problem. Now I can type a simple command get it all done in one go. Second: Well, just for fun :)

• How did I find Leetcode's API?

  TL;DR - Luck + Hit and trial.

  I originally planned to make this tool using web-scrapping (via Colly). I struggled quite a lot in between the process when I needed to scrape the page after entering a search query in the Problemset page of Leetcode. The problem was that Colly doesn't support scrapping in an AJAX loaded site (meaning - the page first loads all the question from 1-50, then executes the search query and displays results. Colly scrapes and stops before the query can even happen). To solve this I would have to use a Selenium like package which would involve chromium driven browser. Didn't wanna do that.

  While I was searching for some solution in the Inspect-Element window, I found a few graphql entries in the network tab. Had no clue what GraphQL was at that time (it is really cool thing, must check it out). Randomly opened a few of them and there I found what I really needed but wasn't looking for. Then I spent a few hours getting to know how GraphQL works and reverse-engineered the queries from the network tab, and it worked!

• While using the --name flag, do I need to give the name exactly how it is on Leetcode?

  No, as long as you enter the correct spellings, it will be alright. Check this out for a few examples of a valid names.

• While using the --url flag, do I need to give the full url?

  Again, no. Just make sure it is in double quotes to avoid any unexpected errors. Check this out for a few examples of a valid urls.