Unescape the special hex code escape sequences (#73)

Unescape the special hex code escape sequences (#73)

Author: jarredholman

lib/creek/shared_strings.rb

@@ -31,9 +31,9 @@ def self.parse_shared_string_from_document(xml)
       xml.css('si').each_with_index do |si, idx|
         text_nodes = si.css('t')
         if text_nodes.count == 1 # plain text node
-          dictionary[idx] = text_nodes.first.content
+          dictionary[idx] = Creek::Styles::Converter.unescape_string(text_nodes.first.content)
         else # rich text nodes with text fragments
-          dictionary[idx] = text_nodes.map(&:content).join('')
+          dictionary[idx] = text_nodes.map { |n| Creek::Styles::Converter.unescape_string(n.content) }.join('')
         end
       end
 

lib/creek/styles/converter.rb

@@ -4,6 +4,10 @@ module Creek
   class Styles
     class Converter
       include Creek::Styles::Constants
+
+      # Excel non-printable character escape sequence
+      HEX_ESCAPE_REGEXP = /_x[0-9A-Za-z]{4}_/
+
       ##
       # The heart of typecasting. The ruby type is determined either explicitly
       # from the cell xml or implicitly from the cell style, and this
@@ -45,9 +49,9 @@ def self.call(value, type, style, options = {})
         when 'b'
           value.to_i == 1
         when 'str'
-          value
+          unescape_string(value)
         when 'inlineStr'
-          value
+          unescape_string(value)
 
         ##
         # Type can also be determined by a style,
@@ -112,6 +116,12 @@ def self.convert_bignum(value)
         end
       end
 
+      def self.unescape_string(value)
+        # excel encodes some non-printable characters using a hex code in the format _xHHHH_
+        # e.g. Carriage Return (\r) is encoded as _x000D_
+        value.gsub(HEX_ESCAPE_REGEXP) { |match| match[2, 4].to_i(16).chr(Encoding::UTF_8) }
+      end
+
       private
 
         def self.base_date(options)

spec/fixtures/sst.xml

@@ -75,4 +75,7 @@
             <t>B2</t>
         </r>
     </si>
+    <si>
+        <t>Cell with_x000D_escaped_x000D_characters</t>
+    </si>
 </sst>

spec/shared_string_spec.rb

@@ -7,12 +7,13 @@
     doc = Nokogiri::XML(shared_strings_xml_file)
     dictionary = Creek::SharedStrings.parse_shared_string_from_document(doc)
 
-    expect(dictionary.keys.size).to eq(5)
+    expect(dictionary.keys.size).to eq(6)
     expect(dictionary[0]).to eq('Cell A1')
     expect(dictionary[1]).to eq('Cell B1')
     expect(dictionary[2]).to eq('My Cell')
     expect(dictionary[3]).to eq('Cell A2')
     expect(dictionary[4]).to eq('Cell B2')
+    expect(dictionary[5]).to eq("Cell with\rescaped\rcharacters")
   end
 
 end