handle utf-8 bom controller

app/controllers/dois_controller.rb

@@ -519,6 +519,14 @@ def safe_params
     # extract attributes from xml field and merge with attributes provided directly
     xml = p[:xml].present? ? Base64.decode64(p[:xml]).force_encoding("UTF-8") : nil
 
+    if xml.present?
+      # remove optional utf-8 bom
+      xml.gsub!("\xEF\xBB\xBF", '')
+
+      # remove leading and trailing whitespace
+      xml = xml.strip
+    end
+
     meta = xml.present? ? parse_xml(xml, doi: p[:doi]) : {}
     xml = meta["string"]
 

app/models/concerns/crosscitable.rb

@@ -89,7 +89,7 @@ def clean_xml(string)
 
         # enforce utf-8
         string = string.force_encoding("UTF-8")
-      rescue ArgumentError, Encoding::CompatibilityError => error
+      rescue ArgumentError, Encoding::CompatibilityError => exception
         # convert utf-16 to utf-8
         string = string.force_encoding('UTF-16').encode('UTF-8')
         string.gsub!('encoding="UTF-16"', 'encoding="UTF-8"')
@@ -105,7 +105,13 @@ def clean_xml(string)
 
       # make sure xml is valid
       doc = Nokogiri::XML(string) { |config| config.strict.noblanks }
-      doc.to_xml.strip
+      doc.to_xml
+    rescue ArgumentError, Encoding::CompatibilityError => exception
+      logger = Logger.new(STDOUT)
+      logger.error "Error " + exception.message + "."
+      logger.error exception
+
+      nil
     end
 
     def well_formed_xml(string)
@@ -120,7 +126,7 @@ def from_xml(string)
       return nil unless string.start_with?('<?xml version=') || string.start_with?('<resource ')
 
       doc = Nokogiri::XML(string) { |config| config.strict.noblanks }
-      doc.to_xml.strip
+      doc.to_xml
     end
 
     def from_json(string)

spec/fixtures/files/datacite.xml

@@ -37,4 +37,4 @@
   <descriptions>
     <description descriptionType="Abstract">Eating your own dog food is a slang term to describe that an organization should itself use the products and services it provides. For DataCite this means that we should use DOIs with appropriate metadata and strategies for long-term preservation for...</description>
   </descriptions>
-</resource>
+</resource>

spec/requests/metadata_spec.rb

@@ -65,7 +65,7 @@
       before { post "/dois/#{doi.doi}/metadata", params: valid_attributes.to_json, headers: headers }
 
       it 'creates a metadata record' do
-        expect(Base64.decode64(json.dig('data', 'attributes', 'xml'))).to eq(xml)
+        expect(Base64.decode64(json.dig('data', 'attributes', 'xml'))).to eq(xml.rstrip)
         expect(json.dig('data', 'attributes', 'namespace')).to eq("http://datacite.org/schema/kernel-4")
       end