🎁 Improve word coordinates generator

This commit will ensure that the word coordinates generator will have
unique values.  This has been observed in IIIF Print's version of the
same generator where the same word coordinates appear multiple times and
results in the UV having multiple annotations for the same word at the
same place.  This also will set up the text generator for a more
Tesseract like text file where extra spaces are omitted for an overall
cleaner output.  This also will set up the alto generator to not have
duplicate word coordinates as well.

- Resolves: #6

lib/derivative_rodeo/services/extract_word_coordinates_from_hocr_sgml_service.rb

@@ -121,6 +121,7 @@ def end_word
           # add trailing space to plaintext buffer for between words:
           @text += ' '
           @words.push(@current) if word_complete?
+          @current = nil # clear the current word
         end
 
         def end_line
@@ -156,10 +157,13 @@ def characters(value)
         # Callback for element end; at this time, flush word coordinate state
         #   for current word, and append line endings to plain text:
         #
-        # @param _name [String] element name.
-        def end_element(_name)
-          end_line if @element_class_name == 'ocr_line'
-          end_word if @element_class_name == 'ocrx_word'
+        # @param name [String] element name.
+        def end_element(name)
+          if name == 'span'
+            end_word if @element_class_name == 'ocrx_word'
+            @text += "\n" if @element_class_name.nil?
+          end
+          @element_class_name = nil
         end
 
         # Callback for completion of parsing hOCR, used to normalize generated

spec/derivative_rodeo/generators/word_coordinates_generator_spec.rb

@@ -13,6 +13,7 @@
         json = JSON.parse(File.read(generated_file.file_path))
         expect(json.keys).to match_array(["width", "height", "coords"])
         expect(generated_file.exist?).to be_truthy
+        expect(generated_file.file_path).to end_with("/ocr_mono_text_hocr.coordinates.json")
       end
 
       expect(generated_file.exist?).to be_falsey

spec/derivative_rodeo/services/extract_word_coordinates_from_hocr_sgml_service_spec.rb

@@ -20,16 +20,19 @@
     subject { hocr.text }
 
     it 'outputs plain text' do
-      expect(subject.slice(0, 40)).to eq "_A  FEARFUL  ADVENTURE.\n‘The  Missouri. "
-      expect(subject.size).to eq 831
+      expect(subject.slice(0, 40)).to eq "_A FEARFUL ADVENTURE.\n‘The Missouri. Rep"
+      expect(subject.size).to eq 723
     end
   end
 
   describe '#to_json' do
     subject { hocr.to_json }
-    it 'outputs JSON that includes coords key' do
+    it 'outputs JSON that includes coords key with unique coordinate values' do
       parsed = JSON.parse(subject)
       expect(parsed['coords'].length).to be > 1
+      parsed['coords'].values.each do |value|
+        expect(value.uniq.length).to eq(value.length)
+      end
     end
   end
 end