Extracting a helper method to define interface

Author: jeremyf

lib/newspaper_works/text_extraction/alto_reader.rb

@@ -112,10 +112,11 @@ def isxml?(xml)
       # @return [String] JSON serialization of flattened word coordinates
       def json
         words = @doc_stream.words
-        builder = NewspaperWorks::TextExtraction::WordCoordsBuilder.new(words,
-                                                                        @image_width,
-                                                                        @image_height)
-        builder.to_json
+        NewspaperWorks::TextExtraction::WordCoordsBuilder.json_coordinates_for(
+          words: words,
+          width:@image_width,
+          height: @image_height
+        )
       end
     end
   end

lib/newspaper_works/text_extraction/hocr_reader.rb

@@ -161,12 +161,11 @@ def isxml?(xml)
       # @return [String] JSON serialization of flattened word coordinates
       def json
         words = @doc_stream.words
-        builder = NewspaperWorks::TextExtraction::WordCoordsBuilder.new(
-          words,
-          @doc_stream.width,
-          @doc_stream.height
+        NewspaperWorks::TextExtraction::WordCoordsBuilder.json_coordinates_for(
+          words: words,
+          width: @doc_stream.width,
+          height: @doc_stream.height
         )
-        builder.to_json
       end
     end
   end

lib/newspaper_works/text_extraction/page_ocr.rb

@@ -40,12 +40,11 @@ def words
       end
 
       def word_json
-        builder = NewspaperWorks::TextExtraction::WordCoordsBuilder.new(
-          words,
-          width,
-          height
+        NewspaperWorks::TextExtraction::WordCoordsBuilder.json_coordinates_for(
+          words: words,
+          width: width,
+          height: height
         )
-        builder.to_json
       end
 
       def plain

lib/newspaper_works/text_extraction/word_coords_builder.rb

@@ -2,6 +2,14 @@ module NewspaperWorks
   # Module for text extraction (OCR or otherwise)
   module TextExtraction
     class WordCoordsBuilder
+      # @params words [Array<Hash>] an array of hash objects that have the keys `:word` and `:coordinates`.
+      # @params width [Integer] the width of the "canvas" on which the words appear.
+      # @params height [Integer] the height of the "canvas" on which the words appear.
+      # @return [String] a JSON encoded string.
+      def self.json_coordinates_for(words:, width: nil, height: nil)
+        new(words, width, height).to_json
+      end
+
       def initialize(words, width = nil, height = nil)
         @words = words
         @width = width

spec/lib/newspaper_works/text_extraction/word_coords_builder_spec.rb

@@ -11,10 +11,24 @@
   end
   let(:image_width) { 1_234 }
   let(:image_height) { 5_678 }
-  let(:wcb) { described_class.new(words, image_width, image_height) }
+
+  describe '.json_coordinates_for' do
+    let(:wcb_to_json) { described_class.json_coordinates_for(words: words, width: image_width, height: image_height) }
+    it 'has the correct structure' do
+      expect(wcb_to_json['height']).to eq image_height
+      expect(wcb_to_json['width']).to eq image_width
+      expect(wcb_to_json['coords'].length).to eq 3
+      expect(wcb_to_json['coords']['foo']).not_to be_falsey
+    end
+
+    it 'combines coordinates for the same word' do
+      expect(wcb_to_json['coords']['foo']).to eq [[1, 2, 3, 4], [13, 14, 15, 16]]
+    end
+  end
 
   describe '#to_json' do
     let(:wcb_to_json) { JSON.parse(wcb.to_json) }
+    let(:wcb) { described_class.new(words, image_width, image_height) }
 
     it 'has the correct structure' do
       expect(wcb_to_json['height']).to eq image_height