marc 1.0.4 → 1.2.0

data/lib/marc/marc8/to_unicode.rb

@@ -1,8 +1,6 @@
-# encoding: UTF-8
-
-require 'marc'
-require 'marc/marc8/map_to_unicode'
-require 'unf/normalizer'
+require "marc"
+require "marc/marc8/map_to_unicode"
+require "unf/normalizer"
 
 module MARC
   module Marc8
@@ -24,8 +22,8 @@ module MARC
       BASIC_LATIN = 0x42
       ANSEL = 0x45
 
-      G0_SET = ['(', ',', '$']
-      G1_SET = [')', '-', '$']
+      G0_SET = ["(", ",", "$"]
+      G1_SET = [")", "-", "$"]
 
       CODESETS = MARC::Marc8::MapToUnicode::CODESETS
 
@@ -63,10 +61,9 @@ module MARC
       # it's not already, if it's Marc8 there's no good reason for it not to
       # be already.
       def transcode(marc8_string, options = {})
-        invalid_replacement     = options.fetch(:replace, "\uFFFD")
-        expand_ncr              = options.fetch(:expand_ncr, true)
-        normalization           = options.fetch(:normalization, :nfc)
-
+        invalid_replacement = options.fetch(:replace, "\uFFFD")
+        expand_ncr = options.fetch(:expand_ncr, true)
+        normalization = options.fetch(:normalization, :nfc)
 
         # don't choke on empty marc8_string
         return "" if marc8_string.nil? || marc8_string.empty?
@@ -82,91 +79,89 @@ module MARC
         combinings = []
         pos = 0
         while pos < marc8_string.length
-            if marc8_string[pos] == "\x1b"
-                next_byte = marc8_string[pos+1]
-                if G0_SET.include? next_byte
-                    if marc8_string.length >= pos + 3
-                        if marc8_string[pos+2] == ',' and next_byte == '$'
-                            pos += 1
-                        end
-                        self.g0 = marc8_string[pos+2].ord
-                        pos = pos + 3
-                        next
-                    else
-                        # if there aren't enough remaining characters, readd
-                        # the escape character so it doesn't get lost; may
-                        # help users diagnose problem records
-                        uni_list.push marc8_string[pos]
-                        pos += 1
-                        next
-                    end
-
-                elsif G1_SET.include? next_byte
-                    if marc8_string[pos+2] == '-' and next_byte == '$'
-                        pos += 1
-                    end
-                    self.g1 = marc8_string[pos+2].ord
-                    pos = pos + 3
-                    next
-                else
-                    charset = next_byte.ord
-                    if CODESETS.has_key? charset
-                        self.g0 = charset
-                        pos += 2
-                    elsif charset == 0x73
-                        self.g0 = BASIC_LATIN
-                        pos += 2
-                        if pos == marc8_string.length
-                            break
-                        end
-                    end
+          if marc8_string[pos] == "\x1b"
+            next_byte = marc8_string[pos + 1]
+            if G0_SET.include? next_byte
+              if marc8_string.length >= pos + 3
+                if (marc8_string[pos + 2] == ",") && (next_byte == "$")
+                  pos += 1
                 end
-            end
-
-            mb_flag = is_multibyte(self.g0)
-
-            if mb_flag
-                code_point = (marc8_string[pos].ord * 65536 +
-                     marc8_string[pos+1].ord * 256 +
-                     marc8_string[pos+2].ord)
+                self.g0 = marc8_string[pos + 2].ord
                 pos += 3
-            else
-                code_point = marc8_string[pos].ord
+              else
+                # if there aren't enough remaining characters, readd
+                # the escape character so it doesn't get lost; may
+                # help users diagnose problem records
+                uni_list.push marc8_string[pos]
                 pos += 1
+              end
+              next
+            elsif G1_SET.include? next_byte
+              if (marc8_string[pos + 2] == "-") && (next_byte == "$")
+                pos += 1
+              end
+              self.g1 = marc8_string[pos + 2].ord
+              pos += 3
+              next
+            else
+              charset = next_byte.ord
+              if CODESETS.has_key? charset
+                self.g0 = charset
+                pos += 2
+              elsif charset == 0x73
+                self.g0 = BASIC_LATIN
+                pos += 2
+                if pos == marc8_string.length
+                  break
+                end
+              end
             end
+          end
 
-            if (code_point < 0x20 or
-                (code_point > 0x80 and code_point < 0xa0))
-                uni = unichr(code_point)
-                next
-            end
+          mb_flag = is_multibyte(g0)
 
-            begin
-              code_set = (code_point > 0x80 and not mb_flag) ? self.g1 : self.g0
-              (uni, cflag) = CODESETS.fetch(code_set).fetch(code_point)
+          if mb_flag
+            code_point = (marc8_string[pos].ord * 65536 +
+              marc8_string[pos + 1].ord * 256 +
+              marc8_string[pos + 2].ord)
+            pos += 3
+          else
+            code_point = marc8_string[pos].ord
+            pos += 1
+          end
 
-              if cflag
-                  combinings.push unichr(uni)
-              else
-                  uni_list.push unichr(uni)
-                  if combinings.length > 0
-                      uni_list.concat combinings
-                      combinings = []
-                  end
-              end
-            rescue KeyError
-              if options[:invalid] == :replace
-                # Let's coallesece multiple replacements
-                uni_list.push invalid_replacement unless uni_list.last == invalid_replacement
-                pos += 1
-              else
-                raise Encoding::InvalidByteSequenceError.new("MARC8, input byte offset #{pos}, code set: 0x#{code_set.to_s(16)}, code point: 0x#{code_point.to_s(16)}, value: #{transcode(marc8_string, :invalid => :replace, :replace => "�")}")
+          if (code_point < 0x20) ||
+              ((code_point > 0x80) && (code_point < 0xa0))
+            uni = unichr(code_point)
+            next
+          end
+
+          begin
+            code_set = (code_point > 0x80) && !mb_flag ? g1 : g0
+            (uni, cflag) = CODESETS.fetch(code_set).fetch(code_point)
+
+            if cflag
+              combinings.push unichr(uni)
+            else
+              uni_list.push unichr(uni)
+              if combinings.length > 0
+                uni_list.concat combinings
+                combinings = []
               end
             end
+          rescue KeyError
+            if options[:invalid] == :replace
+              # Let's coallesece multiple replacements
+              uni_list.push invalid_replacement unless uni_list.last == invalid_replacement
+              pos += 1
+            else
+              raise Encoding::InvalidByteSequenceError.new("MARC8, input byte offset #{pos}, code set: 0x#{code_set.to_s(16)}, code point: 0x#{code_point.to_s(16)}, value: #{transcode(marc8_string, invalid: :replace, replace: "�")}")
+            end
+          end
         end
 
         # what to do if combining chars left over?
-        uni_str = uni_list.join('')
+        uni_str = uni_list.join("")
 
         if expand_ncr
           uni_str.gsub!(/&#x([0-9A-F]{4,6});/) do
@@ -178,7 +173,7 @@ module MARC
           uni_str = UNF::Normalizer.normalize(uni_str, normalization)
         end
 
-        return uni_str
+        uni_str
       end
 
       # from the original python, yeah, apparently
@@ -192,7 +187,6 @@ module MARC
       def unichr(code_point)
         [code_point].pack("U")
       end
-
     end
   end
 end