AtCoderStudyBooster
Overview
🚧 This project is still in experimental stage. We are continuously adding features to help with daily AtCoder practice.
AtCoderStudyBooster is a CLI tool designed to accelerate your AtCoder learning journey. It supports downloading problems locally, testing, submitting, and generating solutions. Python installation is required. If you have Python installed, you can install this tool with:
pip install AtCoderStudyBooster
(Python 3.8 or higher is required)
Even after CAPTCHA authentication was introduced, you can still login and submit semi-automatically from CLI. However, you need to manually solve the CAPTCHA through GUI. This tool does not bypass CAPTCHA authentication.
This project is strongly influenced by:
Use Cases
Let's start by using the download command to download problems locally.
1. Download a Specific Contest
Examples for downloading problems from a single contest.
Download all problems from ABC350
❯ atcdr download abc350
Download problems A-D from ABC350
❯ atcdr download abc350 {A..D}
Download Typical 90 Problems
❯ atcdr download typical90
2. Batch Download Multiple Contests
Examples for downloading multiple contests at once. Utilizes bash brace expansion.
All problems from ABC001 to ABC010
❯ atcdr download abc{001..010}
Specific problems from multiple contests
❯ atcdr download abc{301..310} {A..C}
3. Download Specific Difficulty Problems
Examples for collecting problems of the same difficulty across different contests.
Download all A problems from ABC301-310
❯ atcdr download A abc{301..310}
This creates the following directory structure:
A/
├── abc301/
│ ├── Problem.html
│ └── Problem.md
├── abc302/
│ ├── Problem.html
│ └── Problem.md
└── ...
Download all B problems from ABC300-302
❯ atcdr download B abc{300..302}
Creates:
B/
├── abc300/
│ ├── Problem.html
│ └── Problem.md
├── abc301/
│ ├── Problem.html
│ └── Problem.md
└── abc302/
├── Problem.html
└── Problem.md
This directory structure allows you to efficiently practice problems of the same difficulty level in one place.
Solving Problems
You can view problems by opening Markdown or HTML files with VS Code's HTML Preview or Markdown Preview. In VS Code, you can display the text editor on the left and work on problems while viewing them on the right.
Testing Samples Locally
Navigate to the folder where you downloaded the problem.
❯ cd abc224/B
After creating your solution file in the folder, run the test command to test against sample cases.
~/.../abc224/B
❯ atcdr t
For Wrong Answer (WA) cases, the display looks like this:
Submitting Solutions
~/.../abc224/B
❯ atcdr s
Running this command will submit your solution. Login to AtCoder website is required for submission.
Generating Solutions with GPT
~/.../abc224/B
❯ atcdr g
This command generates a solution using OpenAI's GPT model. An OpenAI API key is required. On first run, you'll be prompted to input your API key.
Login to AtCoder
❯ atcdr login
Logs into AtCoder. A browser window will open for CAPTCHA verification. After solving the CAPTCHA, login completes automatically.
Create Markdown File
~/.../abc224/B
❯ atcdr m
Creates a Markdown file from the HTML file in the current directory. This command is automatically executed during atcdr download.
Open Problem in Browser
~/.../abc224/B
❯ atcdr o
Opens the problem page in your browser. Convenient for checking detailed problem statements or constraints.
Commands
| Command | Alias | Description |
|---|---|---|
atcdr download |
atcdr d |
Download problems |
atcdr test |
atcdr t |
Test with sample cases |
atcdr submit |
atcdr s |
Submit solution |
atcdr generate |
atcdr g |
Generate solution with GPT |
atcdr login |
- | Login to AtCoder |
atcdr logout |
- | Logout from AtCoder |
atcdr markdown |
atcdr m |
Create Markdown file |
atcdr open |
atcdr o |
Open problem in browser |
GPT Code Generation
atcdr generate command uses GPT to generate solutions. It requires an OpenAI API key.
Generate Solution with Test
By default, generated code is tested against sample cases:
~/.../abc224/B
❯ atcdr generate
Specify Language
Specify the programming language for generation. Default is Python.
~/.../abc224/B
❯ atcdr generate --lang cpp
Supported languages:
python(default)cppjavarust
Specify GPT Model
~/.../abc224/B
❯ atcdr generate --gpt gpt-4o
Available models:
gpt-4o-mini(default) - Fast and cost-effectivegpt-4o- More accurate but slower
Generate Code Only Without Testing
To generate code without testing:
~/.../abc224/B
❯ atcdr generate --lang rust --without_test
Metadata
Release files for AtCoderStudyBooster 0.4.2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| atcoderstudybooster-0.4.2.tar.gz | 28.1 MB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| atcoderstudybooster-0.4.2-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 28.1 MB
Release files / atcoderstudybooster-0.4.2.tar.gz
| Download URL | atcoderstudybooster-0.4.2.tar.gz |
|---|---|
| Size | 28.1 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
fdccf0c41fdaf43b1dff28130151d56b74cdfea96e4d35beb7e0c7cf4a1b3690
|
|
BLAKE2b-256 checksum How to use checksums |
ca52465d47fb7266d7fb5b08cf3b48a8e4af644e268c6fa987c5e51b0e589aac
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/5.1.1 CPython/3.12.4
|
Release files / atcoderstudybooster-0.4.2-py3-none-any.whl
| Download URL | atcoderstudybooster-0.4.2-py3-none-any.whl |
|---|---|
| Size | 36.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
d874fa7952d7d2949d9a6b0896768002c4ad838b48738509f17a5cf05c52c7c3
|
|
BLAKE2b-256 checksum How to use checksums |
c59e706936ed761d2bcc542b4f367870cd96674e44270647af943906926f1b6c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/5.1.1 CPython/3.12.4
|