Add test to check for orphan yaml variables in the content (#4173)

* Adds a test that checks for orphan variables in the config/values/*.yml files that are not called in the front end content of the site

* Added check for variables in front matter components

* Improved regex

* Rubocop has some opinions

* Remove ruby files from list of files to be scanned

* Refactored spec to keep rubocop happy and format error messages to make them easier to read

* Fix regex using non-capturing group, added an ignore list to reject files with variables we want to keep

* Removed unused variables from front end

* Remove images from components list in markdown template handler as this seems to be handled differently and there is a class with the naming

* Remove calls_to_action from template handlers

* Added another let for merging yaml and component hashes

app/views/content/blog/natural-transition-from-engineering-to-teaching-physics.md

@@ -25,5 +25,3 @@ When I spot the exact barrier stopping a child from learning, I always make sure
 When you train to teach physics, you can get an income. During my teacher training my family and I could afford for me to train for a year.
 
 We’re here to answer your questions and provide advice about teaching, whether you’re just thinking about it or you’re ready to apply. [Find out how you can get help and support with getting into teaching](/help-and-support).
-
-$physics-initial-teacher-training$

app/views/content/life-as-a-teacher/teach-disabled-pupils-and-pupils-with-special-educational-needs.md

@@ -50,5 +50,3 @@ You need [specific qualifications to teach a class of pupils with hearing impair
 A special educational needs coordinator (SENCO) assesses, plans and monitors the progress of disabled pupils and pupils with special educational needs.
 
 Once you’re a qualified teacher you'll need to complete the [National Award in Special Educational Needs Coordination (NASENCo)](https://nasen.org.uk/page/nasenco) when you take up your SENCO post.
-
-$get-an-adviser$

spec/views/identify_orphan_variables_spec.rb

@@ -0,0 +1,95 @@
+require "rails_helper"
+
+YAML_FILES = Dir.glob("config/values/*.yml")
+
+CONTENT_FILES = [
+  PageLister.all_md_files,
+  PageLister.all_erb_files,
+  PageLister.all_locale_files,
+].flatten
+
+VARIABLE_REGEX = /<%=\s*(?:v|value)\s*:(?<content>[a-zA-Z_0-9-]+?)\s*%> |
+                  \$(?<content>[a-zA-Z_0-9-]+?)\$ |
+                  %\{(?<content>[a-zA-Z_0-9-]+?)\} |
+                  \#\{\s*v\s*:(?<content>[a-zA-Z_0-9-]+>?)\}/x
+
+# Matches <%= v :thing %> or <%= value :thing %>
+# Matches $thing$
+# Matches %{thing}
+# Matches #{v :thing}
+
+COMPONENT_TYPES = TemplateHandlers::Markdown::COMPONENT_TYPES + %w[calls_to_action images]
+
+IGNORE_VARIABLES = {
+  "config/locales/loaf.yml" => %w[invalid valid],
+}.freeze
+
+RSpec.describe "orphan variables checker" do
+  include I18n::Backend::Flatten
+
+  let!(:yaml_files_and_values) do
+    YAML_FILES.to_h do |file|
+      yaml = YAML.load(File.read(file))
+      flattened_yaml = flatten_translations(nil, yaml, nil, false).transform_keys { |key| key.to_s.gsub(".", "_") }
+
+      [file, flattened_yaml.keys]
+    end
+  end
+
+  let!(:content_files_and_values) do
+    CONTENT_FILES.index_with do |file|
+      File.read(file).scan(VARIABLE_REGEX).map(&:compact).flatten
+    end
+  end
+
+  let!(:component_files_and_keys) do
+    PageLister.all_md_files.index_with do |file|
+      content = File.open(file).read
+      front_matter = FrontMatterParser::Parser.new(:md).call(content).front_matter
+
+      COMPONENT_TYPES.map { |type| front_matter[type] }.compact.map(&:keys).flatten
+    end
+  end
+
+  let!(:frontend_files_and_values) { (yaml_files_and_values.values + component_files_and_keys.values).flatten }
+
+  let!(:orphan_frontend_variables) do
+    orphan_frontend_variables = content_files_and_values.transform_values { |values|
+      values.reject { |value| frontend_files_and_values.include?(value) }
+    }.compact_blank
+
+    remove_ignored_variables(orphan_frontend_variables)
+  end
+
+  let!(:orphan_yaml_variables) do
+    orphan_yaml_variables = yaml_files_and_values.transform_values { |values|
+      values.reject { |value| content_files_and_values.values.flatten.include?(value) }
+    }.compact_blank
+
+    remove_ignored_variables(orphan_yaml_variables)
+  end
+
+  it "does not find any variables in the config/values/*.yml files that are not references in any markdown, erb, or locale file" do
+    expect(orphan_yaml_variables).to be_empty, error_message(orphan_yaml_variables)
+  end
+
+  it "does not find any variables in the frontend that are not referenced in any front matter component or /config/values/*.yml file" do
+    expect(orphan_frontend_variables).to be_empty, error_message(orphan_frontend_variables)
+  end
+
+  def error_message(file_variables_hash)
+    file_variables_hash.map.with_index { |(file, orphan), index|
+      "#{index.next}: #{file} contains orphan key(s) '#{orphan.join(', ')}'"
+    }.join("\n")
+  end
+
+  def remove_ignored_variables(hash, ignore_hash = IGNORE_VARIABLES)
+    ignore_hash.map do |file, orphans|
+      next if hash[file].nil?
+
+      hash[file] = hash[file] - orphans
+    end
+
+    hash.compact_blank
+  end
+end