Add conditions to validate image cron job (#1467)

* initial commit

* Update integrity_checkable.rb

* Update integrity_checkable_spec.rb

* Add fixture

* Update comments

---------

Co-authored-by: JP Engstrom <[email protected]>

app/models/concerns/integrity_checkable.rb

@@ -5,6 +5,7 @@ module IntegrityCheckable
 
   # rubocop:disable Metrics/AbcSize
   # rubocop:disable Metrics/MethodLength
+  # rubocop:disable Metrics/CyclomaticComplexity
   # rubocop:disable Metrics/PerceivedComplexity
   # rubocop:disable Layout/LineLength
   def integrity_check
@@ -26,8 +27,18 @@ def integrity_check
         add_admin_set_to_bp(sets, co)
 
         if co.access_master_exists?
-          co.parent_object.processing_event("Integrity check complete for Child Object: #{co.oid}", 'review-complete')
-          co.processing_event("Child Object: #{co.oid} - file exists.", 'review-complete')
+          if co&.file_size&.positive?
+            if co&.access_master_checksum_matches?
+              co.parent_object.processing_event("Integrity check complete for Child Object: #{co.oid}", 'review-complete')
+              co.processing_event("Child Object: #{co.oid} - file exists and checksum matches.", 'review-complete')
+            else
+              co.parent_object.processing_event("Integrity check complete for Child Object: #{co.oid}", 'failed')
+              co.processing_event("The Child Object: #{co.oid} - has a checksum mismatch. The checksum of the image file saved to this child oid does not match the checksum of the image file in the database. This may mean that the image has been corrupted. Please verify integrity of image for Child Object: #{co.oid} - by manually comparing the checksum values and update record as necessary.", 'failed')
+            end
+          else
+            co.parent_object.processing_event("Integrity check complete for Child Object: #{co.oid}", 'failed')
+            co.processing_event("Child Object: #{co.oid} - has a file size of 0. Please verify image for Child Object: #{co.oid}.", 'failed')
+          end
         else
           co.parent_object.processing_event("Integrity check complete for Child Object: #{co.oid}", 'failed')
           co.processing_event("Child Object: #{co.oid} - file not found at #{co.access_master_path} on #{ENV['ACCESS_MASTER_MOUNT']}.", 'failed')
@@ -42,6 +53,7 @@ def integrity_check
   end
   # rubocop:enable Metrics/AbcSize
   # rubocop:enable Metrics/MethodLength
+  # rubocop:enable Metrics/CyclomaticComplexity
   # rubocop:enable Metrics/PerceivedComplexity
   # rubocop:enable Layout/LineLength
 end

spec/models/concerns/integrity_checkable_spec.rb

@@ -8,9 +8,11 @@
   let(:parent_object_one) { FactoryBot.create(:parent_object, oid: '111', child_object_count: 1, authoritative_metadata_source: metadata_source, admin_set: admin_set) }
   let(:parent_object_two) { FactoryBot.create(:parent_object, oid: '222', child_object_count: 1, authoritative_metadata_source: metadata_source, admin_set: admin_set) }
   let(:parent_object_three) { FactoryBot.create(:parent_object, oid: '333', child_object_count: 1, authoritative_metadata_source: metadata_source, admin_set: admin_set) }
+  let(:parent_object_four) { FactoryBot.create(:parent_object, oid: '444', child_object_count: 1, authoritative_metadata_source: metadata_source, admin_set: admin_set) }
   let(:child_object_one) { FactoryBot.create(:child_object, oid: '1', parent_object: parent_object_one) }
   let(:child_object_two) { FactoryBot.create(:child_object, oid: '356789', parent_object: parent_object_two, checksum: '78909999999999999') }
-  let(:child_object_three) { FactoryBot.create(:child_object, oid: '456789', parent_object: parent_object_three, checksum: 'f3755c5d9e086b4522a0d3916e9a0bfcbd47564e') }
+  let(:child_object_three) { FactoryBot.create(:child_object, oid: '456789', parent_object: parent_object_three, file_size: 1234, checksum: 'f3755c5d9e086b4522a0d3916e9a0bfcbd47564e') }
+  let(:child_object_four) { FactoryBot.create(:child_object, oid: '567890', parent_object: parent_object_four, file_size: 1234, checksum: 'f3755c5d9e086b4522a0d3916e9a0bfcbd47564et') }
 
   around do |example|
     original_access_master_mount = ENV["ACCESS_MASTER_MOUNT"]
@@ -28,18 +30,21 @@
     before do
       # file not present
       stub_request(:get, File.join(child_object_one.access_master_path)).to_return(status: 200, body: '')
-      # file present but checksum does not match
+      # file present but file size is less than 0
       stub_request(:get, File.join(child_object_two.access_master_path)).to_return(status: 200, body: File.open(File.join(child_object_two.access_master_path)).read)
       # file present and checksum matches
       stub_request(:get, File.join(child_object_three.access_master_path)).to_return(status: 200, body: File.open(File.join(child_object_three.access_master_path)).read)
+      # file present and file size greater than 0 but checksum does not match
+      stub_request(:get, File.join(child_object_four.access_master_path)).to_return(status: 200, body: File.open(File.join(child_object_three.access_master_path)).read)
     end
 
     it 'reflects messages as expected' do
-      expect { ChildObjectIntegrityCheckJob.new.perform }.to change { IngestEvent.count }.by(7)
+      expect { ChildObjectIntegrityCheckJob.new.perform }.to change { IngestEvent.count }.by(9)
       # rubocop:disable Layout/LineLength
       expect(child_object_one.events_for_batch_process(BatchProcess.first)[0].reason).to eq "Child Object: #{child_object_one.oid} - file not found at #{child_object_one.access_master_path} on #{ENV['ACCESS_MASTER_MOUNT']}."
-      expect(child_object_two.events_for_batch_process(BatchProcess.first)[0].reason).to eq "Child Object: #{child_object_two.oid} - file exists."
-      expect(child_object_three.events_for_batch_process(BatchProcess.first)[0].reason).to eq "Child Object: #{child_object_three.oid} - file exists."
+      expect(child_object_two.events_for_batch_process(BatchProcess.first)[0].reason).to eq "Child Object: #{child_object_two.oid} - has a file size of 0. Please verify image for Child Object: #{child_object_two.oid}."
+      expect(child_object_three.events_for_batch_process(BatchProcess.first)[0].reason).to eq "Child Object: #{child_object_three.oid} - file exists and checksum matches."
+      expect(child_object_four.events_for_batch_process(BatchProcess.first)[0].reason).to eq "The Child Object: #{child_object_four.oid} - has a checksum mismatch. The checksum of the image file saved to this child oid does not match the checksum of the image file in the database. This may mean that the image has been corrupted. Please verify integrity of image for Child Object: #{child_object_four.oid} - by manually comparing the checksum values and update record as necessary."
       # rubocop:enable Layout/LineLength
     end
   end