Add DOI normalizer

Closes #113
inukshuk · Sep 14, 2018 · 964f896 · 964f896
1 parent fcf355f
commit 964f896
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 1 deletion.
diff --git a/lib/anystyle/normalizer/locator.rb b/lib/anystyle/normalizer/locator.rb
@@ -3,7 +3,7 @@ module AnyStyle
 
   class Normalizer
     class Locator < Normalizer
-      @keys = [:isbn, :url]
+      @keys = [:isbn, :url, :doi]
 
       def normalize(item, **opts)
         map_values(item) do |key, value|
@@ -12,6 +12,8 @@ def normalize(item, **opts)
             value[/[\d-]+/]
           when :url
             URI.extract(value)
+          when :doi
+            value[/10\.(\d{4,9}\/[-._;()\/:A-Z0-9]+|1002\/\S+)/i] || value
           else
             value
           end

diff --git a/spec/anystyle/normalizer/locator_spec.rb b/spec/anystyle/normalizer/locator_spec.rb
@@ -0,0 +1,14 @@
+module AnyStyle
+  describe "Locator Normalizer" do
+    let(:n) { Normalizer::Locator.new }
+
+    it "extracts valid DOIs" do
+      ({
+        'doi:10.1002/mpr.33.' => { doi: ['10.1002/mpr.33.'] },
+        'https://doi.org/10.1000/182' => { doi: ['10.1000/182'] }
+      }).each do |(a, b)|
+        expect(n.normalize(doi: [a])).to include(b)
+      end
+    end
+  end
+end