handle namespaced xml in metadata import. #165

datacite · Dec 27, 2018 · df42577 · df42577
1 parent 4e9abf8
commit df42577
Show file tree

Hide file tree

Showing 3 changed files with 318 additions and 1 deletion.
diff --git a/app/models/concerns/crosscitable.rb b/app/models/concerns/crosscitable.rb
@@ -101,7 +101,8 @@ def clean_xml(string)
       # remove leading and trailing whitespace
       string = string.strip
 
-      return nil unless string.start_with?('<?xml version=') || string.start_with?('<resource ')
+      # handle missing <?xml version="1.0" ?> and additional namespace
+      return nil unless string.start_with?('<?xml version=') || string.start_with?('<resource ') || /\A<.+:resource/.match(string)
 
       # make sure xml is valid
       doc = Nokogiri::XML(string) { |config| config.strict.noblanks }

diff --git a/spec/concerns/crosscitable_spec.rb b/spec/concerns/crosscitable_spec.rb
@@ -16,6 +16,11 @@
       expect { subject.clean_xml(string) }.to raise_error(Nokogiri::XML::SyntaxError, "39:18: FATAL: Premature end of data in tag resource line 2")
     end
 
+    it "clean_xml namespace" do
+      string = file_fixture('datacite_namespace.xml').read
+      expect(subject.clean_xml(string)).to start_with('<?xml version="1.0"')
+    end
+
     it "clean_xml utf-8 bom" do
       string = file_fixture('utf-8_bom.xml').read
       expect(subject.clean_xml(string)).to start_with('<?xml version="1.0" encoding="UTF-8"?>')