Engineering

Migrating from Contentlayer to Content Collections

We recently migrated our content framework from Contentlayer to Content Collections. Here's why and how we did it.

Migrating from Contentlayer to Content Collections

Contentlayer is a popular content SDK that transforms your Markdown & MDX content into type-safe JSON data that you can easily import into your React applications.

However, ever since their main sponsor, Stackbit, was acquired by Netlify, Contentlayer has been effectively unmaintained. Couple that with issues like Next.js App Router incompatibility, and it was clear that a drop-in replacement was needed.

Enter Content Collections.

What is Content Collections?

Content Collections is a Contentlayer-inspired content framework created by Sebastian Sdorra. Frustrated by the lack of flexibility in Contentlayer's schema definition (e.g. you can't define an email or url type for a given field), Sebastian decided to create a drop-in replacement for Contentlayer that addresses these limitations.

Why Content Collections?

Here are several reasons why Content Collections is the best drop-in successor to Contentlayer:

1. Zod-powered Schema Validation

If you're already using Zod for schema validation, Content Collections will feel like second nature to you. To define a collection schema, you can easily use the schema attribute, which comes with a Zod parameter z:

content-collections.ts
import { defineCollection, defineConfig } from "@content-collections/core";
 
const BlogPost = defineCollection({
  name: "BlogPost",
  directory: "content/blog",
  include: "*.mdx",
  schema: (z) => ({
    title: z.string(),
    categories: z
      .array(z.enum(["company", "engineering", "education", "customers"]))
      .default(["company"]),
    publishedAt: z.string().datetime(),
    featured: z.boolean().default(false),
  }),
});
 
export default defineConfig({
  collections: [BlogPost],
});

2. Native Fumadocs integration

The creator of Fumadocs – an increasingly popular open-source documentation framework used by projects like Million and Turborepo – recently added native support for Content Collections.

Interestingly, support for Contentlayer has also been deprecated.

3. Compatible with Next.js App Router

Content Collections is compatible with Next.js App Router and React Server Components (RSC).

For example, to render the GitHub repository cards in this blog post, you can create a server component that fetches the latest stats for the repository and renders a React component:

import { getRepo } from "@/lib/github";
import { Suspense } from "react";
 
export function GithubRepo({ url }: { url: string }) {
  return (
    <Suspense fallback={<div>Loading...</div>}>
      <GithubRepoRSC url={url} />
    </Suspense>
  );
}
 
async function GithubRepoRSC({ url }: { url: string }) {
  const { description, stars, forks } = await getRepo(url);
  return (
    <div className="not-prose grid gap-4">
      <h3>{description}</h3>
      <p>{stars} stars</p>
      <p>{forks} forks</p>
    </div>
  );
}

Then, include it in your MDX content:

components/mdx
import { useMDXComponent } from "@content-collections/mdx/react";
import { GithubRepo } from "./github-repo";
 
export function MDX({ code, className }: { code: string; className?: string }) {
  const Component = useMDXComponent(code);
 
  return (
    <article>
      <Component
        components={{
          GithubRepo,
        }}
      />
    </article>
  );
}

4. Actively maintained by an incredible developer

Content Collections is actively maintained by Sebastian Sdorra. If you've enjoyed using Content Collections, we highly recommend sponsoring Sebastian to help him maintain this project.

How to migrate from Contentlayer to Content Collections?

Since Content Collections is a drop-in replacement for Contentlayer, migrating from Contentlayer to Content Collections is fairly straightforward.

First, install the required Content Collections packages.

If you're using Next.js, you'll also need to install @content-collections/next as well as @content-collections/mdx for MDX support.

pnpm add --save-dev @content-collections/core @content-collections/mdx @content-collections/next

Then, adjust your tsconfig.json file:

tsconfig.json
{
  "compilerOptions": {
    // ...
    "paths": {
      "@/*": ["./*"],
      "contentlayer/generated": ["./.contentlayer/generated"],
      "content-collections": ["./.content-collections/generated"] 
    }
  }
}

Next, update your next.config.js file:

next.config.js
// import { withContentlayer } from "next-contentlayer";
import { withContentCollections } from "@content-collections/next";
 
// export default withContentlayer(nextConfig);
export default withContentCollections(nextConfig);

Next, add a content-collections.ts file at the root of your project. As you can see, the main difference is the usage of Zod for schema validation.

import { defineCollection, defineConfig } from "@content-collections/core";
 
const BlogPost = defineCollection({
  name: "BlogPost",
  directory: "content/blog",
  include: "*.mdx",
  schema: (z) => ({
    title: z.string(),
    categories: z
      .array(z.enum(["company", "engineering", "education", "customers"]))
      .default(["company"]),
  }),
});
 
export default defineConfig({
  collections: [BlogPost],
});

If you're using any Remark or Rehype plugins, you'll need to update them accordingly as well. Since Content Collections does not compile or parse the content by default, you'll need to use the transform option to pass the plugins to the compiler:

import { defineCollection, defineConfig } from "@content-collections/core";
import { compileMDX } from "@content-collections/mdx"; 
import { rehypeCode, remarkGfm } from "fumadocs-core/mdx-plugins"; 
 
async function transformMDX(document) { 
  const body = await compileMDX(context, document, { 
    ...options, 
    remarkPlugins: [remarkGfm], 
    rehypePlugins: [[rehypeCode]], 
  }); 
  return { 
    ...document, 
    body, 
  }; 
} 
 
const BlogPost = defineCollection({
  name: "BlogPost",
  directory: "content/blog",
  include: "*.mdx",
  schema: (z) => ({
    title: z.string(),
    categories: z
      .array(z.enum(["company", "engineering", "education", "customers"]))
      .default(["company"]),
  }),
  transform: transformMDX, 
});
 
export default defineConfig({
  collections: [BlogPost],
});
 

The reason why Content Collections does not parse or compile the content by default is for flexibility - you can essentially bring your own Markdown formatter/library and use it with Content Collections.

Finally, update your imports to use the new content-collections source. This can be as simple as doing a CMD+SHIFT+F and replacing contentlayer/generated with content-collections.

blog/page.tsx
// import { allBlogPosts } from "contentlayer/generated";
import { allBlogPosts } from "content-collections";
 
export default function Blog() {
  const posts = allBlogPosts();
  return (
    <div>
      {posts.map((post) => (
        <BlogPost post={post} />
      ))}
    </div>
  );
}

You might also need to update your MDX component's code attribute as well:

blog/[slug]/page.tsx
import { MDX } from "@/components/mdx";
 
export default function BlogPost({ params }: { params: { slug: string } }) {
  const posts = allBlogPosts();
  const post = posts.find((post) => post.slug === params.slug);
  return (
    <MDX
      // code={data.body.code}
      code={data.body}
    />
  );
}

Finally, remove any Contentlayer dependencies and files from your project.

pnpm remove contentlayer next-contentlayer
rm -rf .contentlayer contentlayer.config.ts

And voilà! You've successfully migrated your project from Contentlayer to Content Collections.

Scale your content pipeline with Content Collections

We're very excited about what this migration to Content Collections unlocks for Dub. Not only does it give us a much better developer experience with its Zod schema validation, but it also allows us to ensure our content pipeline is more robust and reliable in the long run.

As always, our codebase is fully open-source, so feel free to check it out and learn more about the latest tech we're using at Dub:

And if you're looking for an actively maintained, modern alternative to Contentlayer, we highly recommend giving Content Collections a try.

Supercharge
your marketing efforts

See why Dub.co is the link management infrastructure of choice for modern marketing teams.