Merge pull request #1437 from alphagov/faq-page-arbitrary-depth

Make FAQs look deeper in the page for questions

CHANGELOG.md

@@ -7,6 +7,10 @@
   useful summary for people upgrading their application, not a replication
   of the commit log.
 
+## Unreleased
+
+* Make FAQs look deeper in the page for questions ([PR #1437](https://github.com/alphagov/govuk_publishing_components/pull/1437))
+
 ## 21.41.1
 
 * Fix button alignment on cookie-banner ([PR #1450](https://github.com/alphagov/govuk_publishing_components/pull/1450))

lib/govuk_publishing_components/presenters/machine_readable/faq_page_schema.rb

@@ -57,22 +57,42 @@ def question_and_answers(html)
 
         question = page.title
 
-        # rubocop:disable Style/IfInsideElse
         doc.xpath("html/body").children.each_with_object({}) do |element, q_and_as|
-          if question_element?(element)
-            question = element.text
-            q_and_as[question] = { anchor: element["id"] }
-          else
-            if question_hash_is_unset?(q_and_as, question)
-              q_and_as[question] = { answer: element.to_s }
-            elsif answer_is_unset?(q_and_as, question)
-              q_and_as[question][:answer] = element.to_s
-            else
-              q_and_as[question][:answer] << element.to_s
+          _q_and_as, question = recursive_question_and_answers(element, question, q_and_as)
+        end
+      end
+
+      def recursive_question_and_answers(element, question, q_and_as)
+        if is_a_question?(element)
+          question = element.text
+          q_and_as[question] = { anchor: element["id"] }
+        else
+          q_and_as = add_answer_to_question(q_and_as, element, question)
+          element.children.each do |child_element|
+            if child_element.element?
+              q_and_as, question = recursive_question_and_answers(child_element, question, q_and_as)
             end
           end
         end
-        # rubocop:enable Style/IfInsideElse
+
+        [q_and_as, question]
+      end
+
+      def add_answer_to_question(q_and_as, element, question)
+        if no_questions_in_subtree?(element)
+          if question_hash_is_unset?(q_and_as, question)
+            q_and_as[question] = { answer: element.to_s }
+          elsif answer_is_unset?(q_and_as, question)
+            q_and_as[question][:answer] = element.to_s
+          else
+            q_and_as[question][:answer] << element.to_s
+          end
+        end
+        q_and_as
+      end
+
+      def no_questions_in_subtree?(element)
+        element.search(QUESTION_TAG).none?
       end
 
       def question_hash_is_unset?(q_and_as, question)
@@ -86,7 +106,7 @@ def answer_is_unset?(q_and_as, question)
       # we use H2 tags as the "question" and the html between them as the "answer"
       QUESTION_TAG = "h2".freeze
 
-      def question_element?(element)
+      def is_a_question?(element)
         element.name == QUESTION_TAG
       end
 

spec/lib/govuk_publishing_components/presenters/schema_org_spec.rb

@@ -309,6 +309,42 @@ def transaction_content_item(organisations = [], ordered_related_items = [])
         expect(q_and_a.map { |faq| faq["name"] }).to_not include("Step two")
       end
 
+      it "parses nested questions" do
+        part_body = "<div>
+                     <main>
+                     <h2 id='step-one'>Step one</h2>
+                     <p>First catch your dragon</p>
+                     <h2 id='step-two'>Step two</h2>
+                     <p>Give it a treat</p>
+                     <h2 id='step-three'>Step three</h2>
+                     <p>Give it a pat (wear gloves)</p>
+                     </main>
+                     </div>
+                    "
+
+        content_item = dragon_guide
+
+        q_and_a = generate_structured_data(
+          content_item: content_item,
+          schema: :faq,
+          body: part_body,
+          ).structured_data["mainEntity"]
+
+        expect(q_and_a.count).to eq(3)
+
+        expect(q_and_a.first["url"]).to eq("http://www.dev.gov.uk/how-to-train-your-dragon#step-one")
+        expect(q_and_a.second["url"]).to eq("http://www.dev.gov.uk/how-to-train-your-dragon#step-two")
+
+        expect(q_and_a.first["name"]).to eq("Step one")
+        expect(q_and_a.first["acceptedAnswer"]["text"].strip).to eq("<p>First catch your dragon</p>")
+
+        expect(q_and_a.second["name"]).to eq("Step two")
+        expect(q_and_a.second["acceptedAnswer"]["text"].strip).to eq("<p>Give it a treat</p>")
+
+        expect(q_and_a.third["name"]).to eq("Step three")
+        expect(q_and_a.third["acceptedAnswer"]["text"].strip).to eq("<p>Give it a pat (wear gloves)</p>")
+      end
+
       it "handles an empty body to ensure that preview works OK" do
         empty_part_body = ""
         content_item = dragon_guide