Frontend Developer at Lastsecond

Introduction

Contentlayer

Contentlayer is a content SDK that validates and transforms your content into type-safe JSON data that you can easily import into your application. It makes working with local markdown or MDX files a breeze. This replaces MDX-bundler and our own markdown processing workflow.

First, a content source is defined, specifiying the name of the document type, the source where it is located along with the frontmatter fields and any additional computed fields that should be generated as part of the process.

export const Blog = defineDocumentType(() => ({
  name: 'Blog',
  filePathPattern: 'blog/**/*.mdx',
  contentType: 'mdx',
  fields: {
    title: { type: 'string', required: true },
    date: { type: 'date', required: true },
    tags: { type: 'list', of: { type: 'string' }, default: [] },
    ...
  },
  computedFields: {
    readingTime: { type: 'json', resolve: (doc) => readingTime(doc.body.raw) },
    slug: {
      type: 'string',
      resolve: (doc) => doc._raw.flattenedPath.replace(/^.+?(\/)/, ''),
    }
    ...
  },
}))

Contentlayer then processes the MDX files with our desired markdown remark or rehype plugins, validates the schema, generate type definitions and output json files that can be easily imported in our pages. Hot reloading comes out of the box, so edits to the markdown files will be reflected in the browser immediately!

Pliny

A large reason for the popularity of the template was its customizability and integration with other services from analytics providers to commenting solutions. However, this means that a lot of boilerplate code has to be co-located within the template even if the user does not use the feature. Updates and bug fixes had to be copied manually to the user's codebase.

To solve this, I have abstracted the logic to a separate repository - Pliny. Pliny provides out of the box Next.js components to enhance static sites:

• Analytics
  • Google Analytics
  • Plausible Analytics
  • Simple Analytics
  • Umami Analytics
  • Posthog
• Comments
  • Disqus
  • Giscus
  • Utterances
• Newsletter (uses Next 13 API Routes)
  • Buttondown
  • Convertkit
  • Email Octopus
  • Klaviyo
  • Mailchimp
  • Revue
• Command palette search with tailwind style sheet
  • Algolia
  • Kbar (local search)
• UI utility components
  • Bleed
  • Newsletter / Blog Newsletter
  • Pre / Code block
  • Table of Contents

Choose your preferred service by modifying siteMetadata.js and changing the appropriate fields. For example to change from Umami Analytics to Plausible, we can change the following fields:

analytics: {
-   umamiAnalytics: {
-     // We use an env variable for this site to avoid other users cloning our analytics ID
-     umamiWebsiteId: process.env.NEXT_UMAMI_ID, // e.g. 123e4567-e89b-12d3-a456-426614174000
-   },
+    plausibleAnalytics: {
+      plausibleDataDomain: '', // e.g. tailwind-nextjs-starter-blog.vercel.app
+    },
},

Changes in the configuration file gets propagated to the components automatically. No modification to the template is required.

Under the hood, Pliny exports high level components such as <Analytics analyticsConfig={analyticsConfig}/> and <Comments commentsConfig={commentsConfig}/> which takes in a configuration object and renders the appropriate component. Since the layouts are defined on the server side, Next.js is able to use the configuration object to determine which component to render and send only the required component bundle to the client.

New Search Component

What's a blog in 2023 without a command palette search bar?

One of the most highly requested features have been added 🎉! The search component supports 2 search providers - Algolia and Kbar local search.

Algolia

Algolia Docsearch is popular free service used across many documentation websites. It automatically scrapes the website that has is submitted for indexing and makes the search result available via a beautiful dialog modal. The pliny component is greatly inspired by the Docusaurus implementation and comes with a stylesheet that is compatible with the Tailwind CSS theme.

Kbar

Kbar is a fast, portable, and extensible cmd+k interface. The pliny implementation uses kbar to create a local search dialog box. The component loads a JSON file, default search.json, that was created in the contentlayer build process. Try pressing ⌘-k or ctrl-k to see the search bar in action!

Styling and Layout Updates

Theming

tailwind.config.js has been updated to use tailwind typography defaults where possible and to use the built-in support for dark mode via the prose-invert class. This replaces the previous prose-dark class and configuration.

The primary theme color is updated from teal to pink and the primary gray theme from neutral to gray.

Inter is now replaced with Space Grotesk as the default font.

New Layouts

Layout components available in the layouts directory, provide a simple way to customize the look and feel of the blog.¹

The downside of building a popular template is that you start seeing multiple similar sites everywhere 😆. While users are encouraged to customized the layouts to their liking, having more layout options that are easily switchable promotes diversity and perhaps can be a good starting point for further customizations.

In v2, I added a new post layout - PostBanner. It features a large banner image and a centered content container. Check out "Pictures of Canada" blog post which has been updated to use the new layout.

The default blog listing layout has also been updated to include a side bar with blog tags. The search bar in the previous layout has been replace with the new command palette search. To switch back to the old layout, simply change the pages that use the ListLayoutWithTags component back to the original ListLayout.

Migration Recommendations

Due to the large changes in directory structure, setup and tooling, I recommend starting from a fresh template and copying existing content, followed by incrementally migrating changes over to the new template.

Styling changes should be relatively minor and can be copied over from the old tailwind.config.js to the new one. If copying over, you might need to add back the prose-dark class to components that opt into tailwind typography styling. Do modify the font import in the root layout component to use the desired font of choice.

Changes to the MDX processing pipeline and schema can be easily ported to the new Contentlayer setup. If there are changes to the frontmatter fields, you can modify the document type in contentlayer.config.ts to include the new fields. Custom plugins can be added to the remarkPlugins and rehypePlugins properties in the makeSource export of contentlayer.config.ts.

Markdown layouts are no longer sourced automatically from the layouts directory. Instead, they have to be specified in the layouts object defined in blog/[...slug]/page.tsx.²

To port over larger components or pages, I recommend first specificing it as a client component by using the "use client" directive. Once it renders correctly, you can split the interactive components (parts that rely on use hooks) as a client component and keep the remaining code as a server component. Consult the comprehensive Next.js migration guide for more details.

Conclusion

I hope you enjoy the new features and improvements in V2. If you have any feedback or suggestions, feel free to open an issue or reach out to me on Twitter.

Support

Using the template? Support this effort by giving a star on GitHub, sharing your own blog and giving a shoutout on Twitter or be a project sponsor.

Licence

MIT © Timothy Lin