PyBites Carbon

A small utility to generate beautiful code images using the awesome carbon service.

Install

You can get it from PyPI:

pip install pybites-carbon

Required driver

You need to download the ChromeDriver (same version as your Chrome browser) and make it available in your environment:

export DRIVER_PATH=$HOME/bin/chromedriver

To run this automatically upon starting a new shell, add this line to your .zshrc or .bashrc file.

Slow internet

If you have a slow internet connection you can optionally set SECONDS_SLEEP_BEFORE_DOWNLOAD to a value higher than the default 3. This is the time the script waits between clicking on "export image" and closing the driver connection to the carbon site (ending the script).

export SECONDS_SLEEP_BEFORE_DOWNLOAD=10

Usage

You can load in code from a file, the clipboard or a snippet. You can change the language, the image background and theme. You can also provide a different directory to store the image. Lastly, this tool uses Selenium in headless mode, to see what it does in the foreground, use -i (or --interactive).

$ carbon -h
usage: carbon [-h] (-f FILE | -c | -s SNIPPET | -v) [-i] [-l LANGUAGE] [-b BACKGROUND] [-t THEME] [-d DESTINATION]

Create a carbon code image

optional arguments:
  -h, --help            show this help message and exit
  -f FILE, --file FILE  File with code
  -c, --clipboard       Use code on clipboard
  -s SNIPPET, --snippet SNIPPET
                        Code snippet
  -v, --version         Show version
  -i, --interactive     Run Selenium in interactive (not headless) mode
  -l LANGUAGE, --language LANGUAGE
                        Programming language
  -b BACKGROUND, --background BACKGROUND
                        Background color
  -t THEME, --theme THEME
                        Name of the theme
  -d DESTINATION, --destination DESTINATION
                        Specify folder where image should be stored (defaults to current directory)

Examples

1. Make a hello world snippet carbon image:

   $ carbon -s 'print("hello world")'
   

   Resulting image:

   

2. Make a code image of a file, let's pick a FastAPI app I am working on:

   $ cat $HOME/code/infinite-scroll/main.py
   from fastapi import FastAPI, Query
   from sqlmodel import select, Session
   
   from youtube.models import YouTube, YouTubeRead
   from youtube.db import engine
   
   app = FastAPI()
   
   
   @app.get("/videos/", response_model=list[YouTubeRead])
   def read_videos(offset: int = 0, limit: int = Query(default=100, lte=100)):
   	with Session(engine) as session:
   		videos = session.exec(
   			select(YouTube).offset(offset).limit(limit)
   		).all()
   		return videos
   

   Run the script with the -f option:

   carbon -f $HOME/code/infinite-scroll/main.py
   

   Resulting image:

   

3. Copying the following lines to the clipboard:

   Here is my favorite feature: make an image from code I currently have on my OS clipboard (thanks pyperclip):

   Try it out, copy this code:

   from time import sleep
   
   sleep(2)
   

   Then run the script with -c:

   $ carbon -c
   

   Resulting image:

   

Useful shell aliases

I added this alias to my .zshrc to make it even easier:

(Actually I created this image having this alias line on my clipboard, then I ran: carbon -c -l application/x-sh -t monokai -b #D7D7BE -d $HOME/Downloads)

Developer setup

Make a virtual environment and install the requirements-dev.txt file or just run make setup.

Download the ChromeDriver, and extract it in a folder, then set it's full path in .env, for example:

echo "DRIVER_PATH=$HOME/bin/chromedriver" > .env

Again you can increase the time the script takes to download the image, in case you have a slower internet connection:

echo "SECONDS_SLEEP_BEFORE_DOWNLOAD=10" >> .env

(>> means append (not override) to an existing file)

The script uses Selenium in headless mode. The resulting carbon.png image will be downloaded to your computer unless you specify a different destination directory using -d (or --destination).

To run the tests, type pytest or make test (it uses pytesseract - in the dev requirements - to read the text from the generated carbon image file).

We recommend running black before committing code. To set this up run this after checking out the repo:

$ pre-commit install
pre-commit installed at .git/hooks/pre-commit

Enjoy and feel free to mention me or PyBites when you post one of the created images on Twitter.