Home

Next.js open graph image

A tool for generating open graph images made with Next.js

Frameworks

Languages

This project was heavily inspired by og-image, but I wanted to change it in a couple ways. For one the original architecture was somewhat confusing, so I wanted to write it in a way that makes it easy for me to customize it. I also wanted to make it with Next.js, which allows for greater expandability in the future

API

The place to start was the API, made easy with Next.js as you can just place the functions in the api folder. I wanted to create this as a dynamic API route as it makes the interface easier as the title can also be the route. To do this, I used the dynamic route feature of Next.js, by creating the file as [title].js. Within the function, I could then use the line

const {
  query: { title },
} = req;

And this allowd for me to use the title variable throughout the rest of my code, and it would be whatever was submitted.

Then came the actual capture of the image, for this I used puppeteer, but due to the size limitation of AWS lambda functions, I had to use the library chrome-aws-lambda which brotli compresses the chrome file in order for it to fit within the 50MB limit. An interesting future expansion on this project would be to use a library like jimp which would allow the image to be made using image processing rather then web development.

I also used an interesting library callefd next-absolute-url, which allowed me to fetch the URL of my deployment, with a fallback for localhost, allowing me to make the correct to the right place.

A customisation I wanted to make to Puppeteer was to change the viewport to the right size for an open graph image, from what I can tell 1200x627 is the recommended size, so I could do that using

await page.setViewport({ width: 1200, height: 627 });

Page to take a screenshot of

I also needed to create this page to screenshot, I used the same dynamic method as I did for the API, making the file [title].js, and a slightly different method for getting the query

const router = useRouter();
const { title } = router.query;

I also wanted to implement a dark mode for the images, so used the new Tailwind CSS dark mode by applying "dark" as a class to the outermost div if dark=true was passed in as a query. I could then use classes like dark:bg-black to change the image. This query was then passed through from the API