Add structured data for news articles

This implements a minimum of structured data for news articles.

 As per Google:

> Structured data is a standardized format for providing information
about a page and classifying the page content; for example, on a recipe
page, what are the ingredients, the cooking time and temperature, the
calories, and so on.

https://developers.google.com/search/docs/guides/intro-structured-data

We use the JSON-LD (JSON "Linked Data", https://json-ld.org) here. It's
Google's recommended way of implementing this and avoids a lot of the
hassle involved with Microdata, which interleaves the structured data
with HTML.

The output of this is tested with the Structured data testing tool:

https://search.google.com/structured-data/testing-tool

Unfortunately, the "publisher" attribute needs a logo. I've chosen the
Opengraph image for this because it's not likely to change and will
probably stay around for a long time. If we're going to implement more
structured data, we'll have to come up with an alternative (and a
better image).

Author: tijmenb

app/presenters/content_item/updatable.rb

@@ -13,6 +13,14 @@ def history
       reverse_chronological_change_history
     end
 
+    def first_public_at
+      content_item["details"]["first_public_at"]
+    end
+
+    def public_updated_at
+      content_item["public_updated_at"]
+    end
+
   private
 
     def change_history
@@ -39,13 +47,5 @@ def any_updates?
         false
       end
     end
-
-    def first_public_at
-      content_item["details"]["first_public_at"]
-    end
-
-    def public_updated_at
-      content_item["public_updated_at"]
-    end
   end
 end

app/presenters/news_article_presenter.rb

@@ -10,4 +10,8 @@ class NewsArticlePresenter < ContentItemPresenter
   def image
     content_item["details"]["image"]
   end
+
+  def structured_data
+    NewsArticleStructured.new(self).structured_data
+  end
 end

app/presenters/news_article_structured.rb

@@ -0,0 +1,64 @@
+class NewsArticleStructured
+  def initialize(presenter)
+    @presenter = presenter
+  end
+
+  def structured_data
+    return {} unless enough_structured_data?
+
+    # http://schema.org/NewsArticle
+    {
+      "@context" => "http://schema.org",
+      "@type" => "NewsArticle",
+      "mainEntityOfPage" => {
+        "@type" => "WebPage",
+        "@id" => page_url,
+      },
+      "headline" => presenter.title,
+      "datePublished" => presenter.first_public_at,
+      "dateModified" => presenter.public_updated_at,
+      "description" => presenter.description,
+      "publisher" => {
+        "@type" => "Organization",
+        "name" => "GOV.UK",
+        "url" => "https://www.gov.uk",
+        "logo" => {
+          "@type" => "ImageObject",
+          # TODO: change this to a better image, without the URL hard coded.
+          "url" => "https://assets.publishing.service.gov.uk/static/opengraph-image-a1f7d89ffd0782738b1aeb0da37842d8bd0addbd724b8e58c3edbc7287cc11de.png",
+        },
+      },
+      "image" => [
+        image["url"],
+      ],
+      "author" => {
+        "@type" => "Organization",
+        "name" => publishing_organisation["title"],
+        "url" => Plek.current.website_root + publishing_organisation["base_path"],
+      },
+    }
+  end
+
+private
+
+  attr_reader :presenter
+
+  def enough_structured_data?
+    # The author (for which we use the publishing org) and image are required
+    # fields. If the news article doesn't have them, don't use structured data
+    # at all.
+    publishing_organisation && image
+  end
+
+  def publishing_organisation
+    presenter.content_item.dig("links", "primary_publishing_organisation").to_a.first
+  end
+
+  def page_url
+    Plek.current.website_root + presenter.content_item["base_path"]
+  end
+
+  def image
+    presenter.image
+  end
+end

app/views/content_items/news_article.html.erb

@@ -1,3 +1,7 @@
+<script type="application/ld+json">
+  <%= @content_item.structured_data.to_json.html_safe %>
+</script>
+
 <div class="grid-row">
   <div class="column-two-thirds">
     <%= render 'govuk_component/title', @content_item.title_and_context %>

test/presenters/news_article_presenter_test.rb

@@ -45,6 +45,22 @@ class PresentedNewsArticleTest < NewsArticlePresenterTestCase
     test 'presents the locale' do
       assert_equal schema_item['locale'], presented_item.locale
     end
+
+    test 'has no structured data if no publishing org' do
+      expected = {}
+      assert_equal expected, presented_item.structured_data
+    end
+
+    test 'has structured data' do
+      item = { "links" => { "primary_publishing_organisation" => [{ "title" => "Ministry of Magic", "base_path" => "/government/organisations/magic" }] } }
+
+      structured_data = presented_item("news_article", item).structured_data
+
+      assert_equal "https://www.test.gov.uk/government/news/christmas-2016-prime-ministers-message", structured_data["mainEntityOfPage"]["@id"]
+      assert_equal "2016-12-25T00:15:02.000+00:00", structured_data["datePublished"]
+      assert_equal "Ministry of Magic", structured_data["author"]["name"]
+      assert_equal "https://www.test.gov.uk/government/organisations/magic", structured_data["author"]["url"]
+    end
   end
 
   class HistoryModePresentedNewsArticle < NewsArticlePresenterTestCase