owzim / kirby-meta

All-in-one powful solution to all of your SEO/OpenGraph/JSON-LD needs.

Kirby Meta

⚠️ Warning: This plugin is in beta state. Documentation and implementation are still incomplete.

This piece of code handles the generation of meta tags for search engines, social networks, browsers and beyond.

Key features:

• 🔎 All-in-one solution for SEO and social media optimization
• 📱 Support for OpenGraph, Twitter Cards and Schema.org markup
• 🚀 Customizable Metadata for auto-generated metadata from page contents
• 💻 Extensive panel UI including social media previews
• 🦊 Easy-to-understand language in the panel, providing a good middle ground between simplicity and extensive control options.
• 🧙‍♂️ Most features can be enabled/disabled in config, panel UI only shows enabled features (thanks to dynamic blueprints)
• 🪝 Hooks for altering the plugin’s behavior
• 🌍 All blueprints are fully translatable

TO DO

• Generate correct fallback title for homepage in share preview widget.

Future plans:

• 🌍 Add German translation
• 🌍 Full compability with multilanguage setups
• 📋 Dedicated panel area for debugging site-wide meta information
• ✅ Live-check of metadata with hints in the panel

This plugin is completely free and published under the MIT license. However, if you are using it in a commercial project and want to help me keep up with maintenance, please consider to ❤️ sponsor me for securing the continued development of the plugin.

How it works

The plugin tries looks for metadata from a pages content file (e.g. article.txt) by the corrsponding key. If the page does not contain the specific field, it looks for a metadata. Method on the current page model, which can return an array metadata for the current page If that also fails, it will fall back to default metadata, as stored in the site.txt file at the top-level of the content directory.

That way, every page will always be able to serve default values, even if the specific page or its model does not contain information like e.g. a thumbnail or a dedicated description.

Installation & Setup

Install using composer (recommended):

composer require fabianmichael/kirby-meta

Alternative download methods:

You can also download this repository as ZIP or add the whole repo as a submodule.

Available configuration options

The options below have to be set in your config.php. Please note that every option has to be prifixed with the plugin namespace, e.g. sitemap => fabianmichael.meta.sitemap.

Key	Type	Default	Description
sitemap	bool	true	When true, will generate an XML sitemap for search engines. The sitemap includes all listed pages by default. ⚠️ If you disable the robots setting, no robots.txt will be served to sell search engines where your sitemap is located.
sitemap.pages.exclude	array	[]	An array of page IDs to exlude from the sitemap. Values are treated as regular expressions, so they can include wildcards like e.g. about/.*. The error page is always excluded.
sitemap.pages.includeUnlisted	array	[]	An array of page IDs to include in the sitemap, even if their status is unlisted. Values are treated as regular expressions, so they can include wildcards like e.g. about/.*.
sitemap.templates.exclude	array	[]	An array of template names to exlude from the sitemap. Values are treated as regular expressions, so they can include wildcards like e.g. `article-(internal
sitemap.templates.includeUnlisted	array	[]	An array of templates to include in the sitemap, even if their status is unlisted. Values are treated as regular expressions.
schema	bool	true	Generate Schema.org markup as JSON-LD.
social	bool	true	Generate OpenGraph and Twitter Cards markup.
robots	bool	true	Generate the robots metatag and serve robots.txt at http(s)://yourdomain.com/robots.txt.
robots.canonical	bool	true	Generate canonical url meta tag. Requires robots option to be true.
robots.index	bool	true	Allows crawlers to index pages. Can be overriden in global or page-specidic settings from the panel. Requires robots option to be true for having an effect. If a page is excluded from the sitemap or unlisted, the robots meta tag will always contain noindex.
robots.follow	bool	true	Allows crawlers to follow links on pages. Can be overriden in global or page-specidic settings from the panel. Requires robots option to be true for having an effect.
robots.archive	bool	true	Allows crawlers to serve a cached version of pages. Can be overriden in global or page-specidic settings from the panel. Requires robots option to be true for having an effect.
robots.imageindex	bool	true	Allows crawlers to include images to appear in search results. Can be overriden in global or page-specidic settings from the panel. Requires robots option to be true for having an effect.
robots.snippet	bool	true	Allows crawlers to generate snippets from page content. Can be overriden in global or page-specidic settings from the panel. Requires robots option to be true for having an effect.
robots.translate	bool	true	Allows crawlers offer automated translation of your content. Can be overriden in global or page-specidic settings from the panel. Requires robots option to be true for having an effect.
theme.color	string|null	null	If not empty, will generate a corresponding meta tag used by some browsers for coloring the UI.

Blueprint setup

Meta comes with predefined tab blueprints to be added to your site and page blueprints:

# site/blueprints/site.yml
[…]
tabs:
  […]
  meta: tabs/meta/site

# site/blueprints/pages/default.yml
[…]
tabs:
  […]
  meta: tabs/meta/page

Template setup

Include the meta snippet within your <head> element, preferably before loading any scripts or stylesheets:

<!doctype html>
<html>
<head>
  <?php snippet('meta') ?>
  […]
</head>
[…]

Now you are ready to add/editr metadata from the panel.

Advanced usage

Providing metadata from page models

meta_description string og_title_prefix string og_image File @graph array @social array

Using hooks

the meta plugin provides a set of handy hooks, allowing you to further add/remove/modify metadata without overriding the built-in snippets or having to set up a page model for every template.

⚠️ Hooks are a powerful tool that can break the plugin’s expected behavior for editors working on the panel. Use with care!

meta.load:after

After metadata has been loaded by calling the $page->metadata() method on a model. This allows you to inject additional data.

return [
  'meta.load:after' => function (
    array $metadata,
    Kirb\Cms\Page $page,
    ?string $languageCode
  ) {
    // set `thumbnail.png` as default share image for all pages,
    // if not other image was already set by a page model
    if (empty($metadata['og_image']) === true) {
      $metadata['og_image'] = $page->image('thumbnail.png');
    }
    return $metadata;
  },
];

meta.jsonld:after hook

After the Schema.org graph has been generated. This allows you to pass additional data to the array.

return [
  'meta.jsonld:after' => function (
    array $json,
    FabianMichael\Meta\PageMeta $meta,
    Kirb\Cms\Page $page
  ) {
    // add breadcrumb to JSON-LD graph
    $items = [];

    $parents = $page->parents();

    if ($parents->count() === 0) {
      return $json;
    }

    $i = 0;

    foreach ($parents->flip() as $parent) {
      $items[] = [
        '@type' => 'ListItem',
        'position' => ++$i,
        'item' => [
          '@id' => $parent->url(),
          'name' => $parent->title()->toString(),
        ],
      ];
    }

    $json['@graph'][] = [
      '@type' => 'BreadcrumbList',
      'itemListElement' => $items,
    ];

    return $json;
  },
];

meta.social:after

Allows you to alter the OpenGraph/Twitter card data.

return [
  'meta.social:after' => function (
    array $social,
    FabianMichael\Meta\PageMeta $meta,
    Kirby\Cms\Page $page
  ) {
    // add first video file of page to OpenGraph markup
    if ($page->hasVideos()) {
      $social[] = [
        'property' => 'og:video',
        'content'  => $page->videos()->first()->url(),
      ];
    }
    return $social;
  },
];

'meta.sitemap… hooks

These hooks allow you to completely alter the way how the sitemap is being generated. These functions are meant to manipulate the provided DOM document and elements directly and should not return anything.

return [
  'hooks' => [
    'meta.sitemap:before' => function (
      Kirby $kirby,
      DOMDocument $doc,
      DOMElement $root
    ) {
      // add namespace for image sitemap
      $root->setAttributeNS('http://www.w3.org/2000/xmlns/', 'xmlns:image', 'http://www.google.com/schemas/sitemap-image/1.1');
    },

    'meta.sitemap.url' => function (
      Kirby $kirby,
      Page $page,
      PageMeta $meta,
      DOMDocument $doc,
      DOMElement $url,
      ?string $languageCode
    ) {
      foreach ($page->images() as $image) {
        // add all images from page to image sitemap.
        $imageEl = $doc->createElement('image:image');
        $imageEl->appendChild($doc->createElement('image:loc', $image->url()));

        if ($image->alt()->isNotEmpty()) {
          $imageEl->appendChild($doc->createElement('image:caption', $image->alt()));
        }

        $url->appendChild($imageEl);
      }
    },

    'meta.sitemap:after' => function (
      Kirby $kirby,
      DOMDocument $doc,
      DOMElement $root
    ) {
      foreach ($root->getElementsByTagName('url') as $url) {
        if ($lastmod = $url->getElementsByTagName('lastmod')[0] ?? null) {
          // remove lastmod date from sitemap entries for some reason …
          $url->removeChild($lastmod);
        }
      }
    },
  ],
];

Credits

This is partly based on an older version of the meta plugin, that I had initially developed for getkirby.com. I liked the idea so much, that I wanted to adapt it for general use on other websites.

It took a lot of inspiration (and some code) from other existing Kirby plugins, like MetaKnight by diesdas ⚡️ digital and Meta Tags by Pedro Borges.