datocms-plugin-yoast-seo

Run Yoast SEO analysis on your frontend everytime you make a change to the content of a record

Usage no npm install needed!

<script type="module">
  import datocmsPluginYoastSeo from 'https://cdn.skypack.dev/datocms-plugin-yoast-seo';
</script>

README

Yoast SEO DatoCMS plugin

This plugin uses Yoast to generate in real-time interesting SEO and readability metrics about a specific record you're editing in DatoCMS.

It shows potential SEO problems, improvements, considerations, and content that is already optimized or "Good".

🚨 Important: This is not a drag & drop plugin! It requires some work on your frontend website in order to function. Read more in the following sections!

Installation and configuration

Once the plugin is installed, please configure your Frontend metadata endpoint URL in the plugin settings:

Demo

This plugin is meant to be used on JSON fields, and will store the following information:

{
  "keyword": "food shows",
  "synonyms": "cooking shows, culinary demonstrations",
  "relatedKeywords": [
    {
      "keyword": "food",
      "keyword": ""
    }
  ]
}

So the next step is to assign this plugin to a JSON field in one of your models.

The Frontend metadata endpoint

In order to work, this plugin needs a CORS-ready endpoint API that is able to return, given the ID of a DatoCMS record, a number of information related to its canonical page (that is, the page on the frontend that presents the content stored inside the record).

The plugin performs a GET request to the URL specified in the settings, passing down the following query string parameters:

<ENDPOINT_URL>?itemId=89274&itemTypeId=544589&itemTypeApiKey=blog_post&environmentId=main&locale=en
  • itemId the ID of the DatoCMS record
  • itemTypeId the ID of the record's model
  • itemTypeApiKey the API key of the record's model
  • sandboxEnvironmentId the environment ID (only passed if the record belongs to a sandbox environment)
  • locale the preferred locale

The endpoint is expected to return a 200 response, with the following JSON structure:

{
  "locale": "en",
  "slug": "hello-world",
  "permalink": "https://www.yourwebsite.com/blog/hello-world",
  "title": "This is the SEO title of the page",
  "description": "This is the SEO description of the page",
  "content": "<p>This is the main content of the page/article</p>..."
}

To better serve the content writer, the information returned should be related to the latest version of the record's content — which could be unpublished.

An example implementation for Next.js apps

This is a complete implementation of such endpoint for Next.js websites that are already configured to fetch draft content from DatoCMS when Preview Mode is activated.

To learn how to setup Preview Mode, please read the documentation, or take a look at this example website.

The endpoint is implemented as a Next.js API Route. It uses the res.setPreviewData method to obtain the proper cookies for Preview Mode, which are immediately used to fetch the webpage related to the DatoCMS record. Once the complete HTML of the page is fetched, it uses the jsdom package to pick the interesting parts of the page and build the response for the plugin:

// Put this code in the following path of your Next.js website:
// /pages/api/get-frontend-metadata.js

import { SiteClient } from 'datocms-client';
import got from 'got';
import { JSDOM } from 'jsdom';

// this "routing" function knows how to convert a DatoCMS record
// into its slug and canonical URL within the website
const findSlugAndPermalink = async ({ item, itemTypeApiKey }) => {
  switch (itemTypeApiKey) {
    case 'blog_post':
      return [item.slug, `/blog/${item.slug}`];
    case 'landing_page':
      return [item.slug, `/cms/${item.slug}`];
    case 'changelog_entry':
      return [item.slug, `/product-updates/${item.slug}`];
    default:
      return null;
  }
};

const handler = async (req, res) => {
  // setup relaxed CORS permissions
  res.setHeader('Content-Type', 'text/html');
  res.setHeader('Access-Control-Allow-Origin', '*');
  res.setHeader('Access-Control-Allow-Methods', 'GET');

  // the following parameters are required and represent the
  // the record we want to get the preview for
  const missingParams = [
    'itemId',
    'itemTypeId',
    'itemTypeApiKey',
    'locale',
  ].filter((paramName) => !req.query[paramName]);

  if (missingParams.length > 0) {
    res.status(422).json({
      message: `Missing required parameters! ${missingParams.join(', ')}`,
    });
    return;
  }

  const { itemId, itemTypeId, itemTypeApiKey, sandboxEnvironmentId, locale } = req.query;

  // retrieve the complete record from the DatoCMS API
  const client = new SiteClient(process.env.DATOCMS_READONLY_TOKEN, { environment: sandboxEnvironmentId });
  const item = await client.items.find(itemId);

  // this "routing" function knows which record is linked to which URL
  // in the website
  const [slug, permalink] = await findSlugAndPermalink({
    item,
    itemTypeId,
    itemTypeApiKey,
    locale,
  });

  if (!permalink) {
    res.status(422).json({
      message: `Don\'t know which route corresponds to record #${itemId}!`,
    });
    return;
  }

  // let's start a Next.js Preview Mode, and get the authentication cookies
  // (fill in the preview data object with whatever you need)
  res.setPreviewData({});

  const cookie = res
    .getHeader('Set-Cookie')
    .map((cookie) => cookie.split(';')[0])
    .join(';');

  res.clearPreviewData();

  // get the HTML of the page associated with the record (in Preview Mode)
  const { body } = await got(
    new URL(permalink, process.env.BASE_URL).toString(),
    {
      headers: { cookie },
    },
  );

  const { document } = new JSDOM(body).window;

  // here we're taking the content of the div with id="main-content"
  // as the page main-content, but this heavily depends on your layout!
  const content = document.getElementById('main-content').innerHTML;

  // get the page locale by looking at the "lang" attribute on the <html> tag
  const locale = document.querySelector('html').getAttribute('lang') || 'en';

  // get the <title> of the page
  const title = document.querySelector('title').textContent;

  // get the description meta of the page
  const description = document
    .querySelector('meta[name="description"]')
    .getAttribute('content');

  res.status(200).json({
    locale,
    slug,
    permalink,
    title,
    description,
    content,
  });
};

export default handler;

If you have built alternative endpoint implementations for other frameworks/SSGs, please open up a PR to this plugin and share it with the community!