From fd317c69c75e3945975111814137c586c92a582f Mon Sep 17 00:00:00 2001
From: Finn Bacall <finn.bacall@manchester.ac.uk>
Date: Wed, 1 Nov 2023 15:35:52 +0000
Subject: [PATCH] Extract `mentions` as "external resources". Fixes #2

---
 lib/tess/rdf/course_extractor.rb   |  2 +-
 lib/tess/rdf/extraction.rb         |  7 +++++++
 lib/tess/rdf/material_extractor.rb |  1 +
 test/field_test.rb                 | 33 ++++++++++++++++++++++++++++++
 4 files changed, 42 insertions(+), 1 deletion(-)

diff --git a/lib/tess/rdf/course_extractor.rb b/lib/tess/rdf/course_extractor.rb
index ed02cee..a41a74b 100644
--- a/lib/tess/rdf/course_extractor.rb
+++ b/lib/tess/rdf/course_extractor.rb
@@ -16,7 +16,7 @@ def extract_params
         end
         # course_params[:difficulty_level] ||=
         #          extract_names_or_values(RDF::Vocab::SCHEMA.educationalLevel, subject: course).first
-
+        course_params[:external_resources_attributes] = extract_mentions
         # ...and override with more specific metadata from CourseInstance
         course_instance_params = super
         course_params.merge(course_instance_params)
diff --git a/lib/tess/rdf/extraction.rb b/lib/tess/rdf/extraction.rb
index 4be1885..61e9eb5 100644
--- a/lib/tess/rdf/extraction.rb
+++ b/lib/tess/rdf/extraction.rb
@@ -248,6 +248,13 @@ def extract_audience(subject: resource)
           [:audience, RDF::RDFS.label, :target_audience, { optional: true }]).map { |a| a[:target_audience] }.compact
       end
 
+      def extract_mentions(subject: resource)
+        query(
+          [subject, RDF::Vocab::SCHEMA.mentions, :mention],
+          [:mention, RDF::Vocab::SCHEMA.name, :name, { optional: true }],
+          [:mention, RDF::Vocab::SCHEMA.url, :url, { optional: true }]).map { |a| { title: a[:name], url: a[:url] } }.compact
+      end
+
       def parse_value(value)
         # Using 'value.class.name' instead of just 'value' here or things like RDF::Literal::DateTime fall into the RDF::Literal block
         # Not using 'value.class' because 'case' uses '===' for comparison and RDF::URI === RDF::URI is false!
diff --git a/lib/tess/rdf/material_extractor.rb b/lib/tess/rdf/material_extractor.rb
index f729a59..119efb9 100644
--- a/lib/tess/rdf/material_extractor.rb
+++ b/lib/tess/rdf/material_extractor.rb
@@ -24,6 +24,7 @@ def extract_params
         params[:contributors] = extract_names(RDF::Vocab::SCHEMA.contributor)
         params[:target_audience] = extract_audience
         params[:resource_type] = extract_values(RDF::Vocab::SCHEMA.learningResourceType)
+        params[:external_resources_attributes] = extract_mentions
 
         remove_blanks(params)
       end
diff --git a/test/field_test.rb b/test/field_test.rb
index b7bc64b..dd841f8 100644
--- a/test/field_test.rb
+++ b/test/field_test.rb
@@ -197,6 +197,33 @@ class FieldTest < Test::Unit::TestCase
     assert_equal '2023-10-02', course_instance_extractor(json).extract_params[:end].to_s
   end
 
+  test 'extract tools from mentions' do
+    json = %(
+[{
+  "@context": "http://schema.org",
+  "@type": "LearningResource",
+  "http://purl.org/dc/terms/conformsTo": {
+    "@id": "https://bioschemas.org/profiles/TrainingMaterial/1.0-RELEASE",
+    "@type": "CreativeWork"
+  },
+  "mentions": [
+    {
+      "@type": "SoftwareApplication",
+      "name": "Galaxy",
+      "url": "https://bio.tools/galaxy",
+      "description": "Open, web-based platform for data intensive biomedical research"
+    },{
+      "@type": "Dataset",
+      "name": "European Genome-phenome Archive",
+      "url": "https://www.ebi.ac.uk/ega/home",
+      "description": "The EGA archives a large number of datasets, the access to which is controlled by a Data Access Committee (DAC)."
+    }
+  ]
+}])
+    assert_equal [{ title: 'Galaxy', url: 'https://bio.tools/galaxy' },
+                  { title: 'European Genome-phenome Archive', url: 'https://www.ebi.ac.uk/ega/home' }], learning_resource_extractor(json).send(:extract_mentions)
+  end
+
   private
 
   def course_extractor(fixture, format: :jsonld, base_uri: 'https://example.com/my.json')
@@ -210,4 +237,10 @@ def course_instance_extractor(fixture, format: :jsonld, base_uri: 'https://examp
     ex.instance_variable_set(:@_temp_resource, ex.resources.first.individual)
     ex
   end
+
+  def learning_resource_extractor(fixture, format: :jsonld, base_uri: 'https://example.com/my.json')
+    ex = Tess::Rdf::LearningResourceExtractor.new(StringIO.new(fixture), format, base_uri: base_uri)
+    ex.instance_variable_set(:@_temp_resource, ex.resources.first.individual)
+    ex
+  end
 end
\ No newline at end of file