Introduction

Hello everyone, I'm Apoorv and in this small blog, I'll be explaining how to create a pre-commit hook to compress images.

As we all know there are multiple ways to compress images before pushing the code to production. One of them is to set up a pre-commit hook. At work, developers in my team had to manually compress the image before pushing it to production, so I came up with a solution to compress images before committing, so it becomes easy for them and saves time as well.

So without any delay, let's get our hands dirty

Steps

1. Install husky and set up pre-commit hook file. For setting up, you can refer to husky docs.

2. Now we'll write the bash script for compressing images.

   In this blog, I'm only covering SVG formats, you can definitely add a compressor for different types of image formats (png, jpeg, jpg, etc).

3. In the pre-commit file, we'll check if svgo is installed for compressing images.

     #!/bin/sh
     . "$(dirname "$0")/_/husky.sh"
   
     command -v ./node_modules/.bin/svgo >/dev/null 2>&1 || { 
         echo "\033[1mPlease install svgo to reduce images size before commit\033[0m" 
         echo "Install svgo as a devDependency using npm install -D svgo" 
         exit 1;
     }
   

4. Now we need to find all the SVG files that are added, renamed, changed or modified in the codebase. we can get that using the command.

   git diff --diff-filter=ACMRT --cached --name-only | grep ".svg\$"

   Continuing with our script, we'll find all those files and loop over them.

     for file in `git diff --diff-filter=ACMRT --cached --name-only | grep ".svg\$"`
     do 
         ...
         ...
         ...
     done
   

5. We will provide an option to the developer if they want to compress a particular svg or not.

       for file in `git diff --diff-filter=ACMRT --cached --name-only | grep ".svg\$"`
       do 
         exec < /dev/tty 
         read -p "Do you want to compress $file? [yN]: " yn 
         case $yn in 
             ...
             ...
         esac 
       done
   

   This is a switch case, where we ask the user to input either Y or y for compressing the image.

6. Now we have to write the code for compressing the SVG file if the user inputs y.

    case $yn in 
        [Yy]* ) 
            echo "Compressing $file" 
            ./node_modules/.bin/svgo $file -o ${file%.svg}.new 
            mv -f ${file%.svg}.new $file
            git add $file;;
        * ) echo "Skipping $file" && continue;; 
    esac

In the first line, we are compressing the image and storing it with the name ${file}.svg.new Then we'll force move the file to its original extension ${file}. And then add it to the commit.

If the user inputs anything other than Y or y, we'll skip the compression for that particular file and continue with other files.

That's all, you have built a simple pre-commit hook to compress the images before committing them.

Checkout the code here - Code link

Enhancements

1. Adding support for multiple image formats.
2. Publish it as a package.

---

Thanks for reading this mini blog. Connect with me on - Twitter, GitHub