* lib/rexml/encoding.rb (REXML::Encoding#encoding=): store @encoding

a String which means the name of the encoding.
  this partially revert r29646.

* lib/rexml/document.rb: follow above.

* lib/rexml/output.rb: ditto.

* lib/rexml/parsers/baseparser.rb: ditto.

* lib/rexml/source.rb: ditto.

* lib/rexml/xmldecl.rb: ditto.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@31008 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,19 @@
+Thu Mar  3 00:36:29 2011  NARUSE, Yui  <naruse@ruby-lang.org>
+
+	* lib/rexml/encoding.rb (REXML::Encoding#encoding=): store @encoding
+	  a String which means the name of the encoding.
+	  this partially revert r29646.
+
+	* lib/rexml/document.rb: follow above.
+
+	* lib/rexml/output.rb: ditto.
+
+	* lib/rexml/parsers/baseparser.rb: ditto.
+
+	* lib/rexml/source.rb: ditto.
+
+	* lib/rexml/xmldecl.rb: ditto.
+
 Wed Mar  2 23:19:56 2011  Nobuyoshi Nakada  <nobu@ruby-lang.org>
 
 	* string.c (str_byte_substr): return nil for negative length.

lib/rexml/document.rb

@@ -184,7 +184,7 @@ module REXML
     #   that IE's limited abilities can handle.  This hack inserts a space
     #   before the /> on empty tags.  Defaults to false
     def write( output=$stdout, indent=-1, transitive=false, ie_hack=false )
-      if xml_decl.encoding != ::Encoding::UTF_8 && !output.kind_of?(Output)
+      if xml_decl.encoding != 'UTF-8' && !output.kind_of?(Output)
         output = Output.new( output, xml_decl.encoding )
       end
       formatter = if indent > -1

lib/rexml/encoding.rb

@@ -1,8 +1,9 @@
 module REXML
   module Encoding
-    # ID ---> Encoding object
+    # ID ---> Encoding name
     attr_reader :encoding
     def encoding=(encoding)
+      encoding = encoding.name if encoding.is_a?(Encoding)
       if encoding.is_a?(String)
         original_encoding = encoding
         encoding = find_encoding(encoding)
@@ -11,35 +12,25 @@ module REXML
         end
       end
       return false if defined?(@encoding) and encoding == @encoding
-      if encoding and encoding != ::Encoding::UTF_8
+      if encoding
-        @encoding = encoding
+        @encoding = encoding.upcase
       else
-        @encoding = ::Encoding::UTF_8
+        @encoding = 'UTF-8'
       end
       true
     end
 
     def check_encoding(xml)
-      # We have to recognize UTF-16, LSB UTF-16, and UTF-8
+      # We have to recognize UTF-16BE, UTF-16LE, and UTF-8
       if xml[0, 2] == "\xfe\xff"
         xml[0, 2] = ""
-        ::Encoding::UTF_16BE
+        return 'UTF-16BE'
       elsif xml[0, 2] == "\xff\xfe"
         xml[0, 2] = ""
-        ::Encoding::UTF_16LE
+        return 'UTF-16LE'
-      else
-        if /\A\s*<\?xml\s+version\s*=\s*(['"]).*?\1
-            \s+encoding\s*=\s*(["'])(.*?)\2/mx =~ xml
-          encoding_name = $3
-          if /\Autf-16\z/i =~ encoding_name
-            ::Encoding::UTF_16BE
-          else
-            find_encoding(encoding_name)
-          end
-        else
-          ::Encoding::UTF_8
-        end
       end
+      xml =~ /^\s*<\?xml\s+version\s*=\s*(['"]).*?\1\s+encoding\s*=\s*(["'])(.*?)\2/m
+      return $3 ? $3.upcase : 'UTF-8'
     end
 
     def encode(string)
@@ -53,14 +44,19 @@ module REXML
     private
     def find_encoding(name)
       case name
-      when "UTF-16"
-        name = "UTF-16BE"
       when /\Ashift-jis\z/i
-        name = "Shift_JIS"
+        return "SHIFT_JIS"
       when /\ACP-(\d+)\z/
         name = "CP#{$1}"
+      when /\AUTF-8\z/i
+        return name
       end
-      ::Encoding.find(name)
+      begin
+        ::Encoding::Converter.search_convpath(name, 'UTF-8')
+      rescue ::Encoding::ConverterNotFoundError
+        return nil
+      end
+      name
     end
   end
 end

lib/rexml/formatters/default.rb

@@ -22,7 +22,7 @@ module REXML
         case node
 
         when Document
-          if node.xml_decl.encoding != ::Encoding::UTF_8 && !output.kind_of?(Output)
+          if node.xml_decl.encoding != 'UTF-8' && !output.kind_of?(Output)
             output = Output.new( output, node.xml_decl.encoding )
           end
           write_document( node, output )

lib/rexml/output.rb

@@ -10,7 +10,7 @@ module REXML
       @output = real_IO
       self.encoding = encd
 
-      @to_utf = (@encoding != ::Encoding::UTF_8)
+      @to_utf = encd != 'UTF-8'
     end
 
     def <<( content )

lib/rexml/parsers/baseparser.rb

@@ -248,7 +248,7 @@ module REXML
             @document_status = :after_doctype
             @source.read if @source.buffer.size<2
             md = @source.match(/\s*/um, true)
-            if @source.encoding == ::Encoding::UTF_8
+            if @source.encoding == "UTF-8"
               @source.buffer.force_encoding(::Encoding::UTF_8)
             end
           end

lib/rexml/source.rb

@@ -54,14 +54,12 @@ module REXML
     def encoding=(enc)
       return unless super
       @line_break = encode( '>' )
-      if @encoding != ::Encoding::UTF_8
+      if @encoding != 'UTF-8'
         @buffer = decode(@buffer)
         @to_utf = true
       else
         @to_utf = false
-        if @buffer.respond_to? :force_encoding
+        @buffer.force_encoding ::Encoding::UTF_8
-          @buffer.force_encoding ::Encoding::UTF_8
-        end
       end
     end
 

lib/rexml/xmldecl.rb

@@ -109,20 +109,9 @@ module REXML
     end
 
     private
-    def normalized_encoding_name(_encoding)
-      if _encoding == ::Encoding::UTF_16BE
-        "UTF-16"
-      else
-        return _encoding.name
-      end
-    end
-
     def content(enc)
       rv = "version='#@version'"
-      if @writeencoding || enc.to_s !~ /\Autf-8\z/i
+      rv << " encoding='#{enc}'" if @writeencoding || enc !~ /utf-8/i
-        encoding_name = normalized_encoding_name(enc)
-        rv << " encoding='#{encoding_name}'"
-      end
       rv << " standalone='#@standalone'" if @standalone
       rv
     end

test/rexml/test_contrib.rb

@@ -241,7 +241,7 @@ DELIMITER
     end
 
     doc = REXML::Document.new(source_iso)
-    assert_equal('ISO-8859-1', doc.xml_decl.encoding.to_s)
+    assert_equal('ISO-8859-1', doc.xml_decl.encoding)
     assert_equal(koln_utf, doc.root.text)
     doc.write(out="")
     assert_equal(source_iso, out )
@@ -255,23 +255,21 @@ DELIMITER
 <position><aktuell datum="01-10-11">Technik</aktuell></position>
 <hauptspalte>
 <headline>Technik</headline>
-Die Technik ist das Rückgrat der meisten Geschäftsprozesse bei Home of the Brave. Deshalb sollen hier alle relevanten technischen Abläufe, Daten und Einrichtungen beschrieben werden, damit jeder im Bedarfsfall die nötigen Informationen, Anweisungen und Verhaltensempfehlungen nachlesen und/oder abrufen kann.
+Die Technik ist das R\xFCckgrat der meisten Gesch\xFCftsprozesse bei Home of the Brave. Deshalb sollen hier alle relevanten technischen Abl\xFCufe, Daten und Einrichtungen beschrieben werden, damit jeder im Bedarfsfall die n\xFCtigen Informationen, Anweisungen und Verhaltensempfehlungen nachlesen und/oder abrufen kann.
 </hauptspalte>
 <nebenspalte>
   <link ziel="Flash/">Flash</link><umbruch/>
-  Nützliches von Flashern für Flasher.<umbruch/>
+  N\xFCtzliches von Flashern f\xFCr Flasher.<umbruch/>
   <link neu="ja" ziel="Cvs/">CVS-FAQ</link><umbruch/>
   FAQ zur Benutzung von CVS bei HOB
 </nebenspalte>
 </intranet>
 EOF
     tn = XPath.first(doc, "//nebenspalte/text()[2]")
-    expected_iso = "Nützliches von Flashern für Flasher."
+    expected_iso = "N\xFCtzliches von Flashern f\xFCr Flasher."
-                expected_utf = expected_iso.unpack('C*').pack('U*')
+    expected_utf = expected_iso.unpack('C*').pack('U*')
-                if expected_utf.respond_to? :encode
+    expected_iso.force_encoding(::Encoding::ISO_8859_1)
-      expected_iso.force_encoding("iso-8859-1")
+    expected_utf.force_encoding(::Encoding::UTF_8)
-      expected_utf.force_encoding(::Encoding::UTF_8)
-                end
     assert_equal(expected_utf, tn.to_s.strip)
     f = REXML::Formatters::Default.new
     f.write( tn, Output.new(o = "", "ISO-8859-1") )

test/rexml/test_core.rb

@@ -230,34 +230,12 @@ class Tester < Test::Unit::TestCase
     doc = Document.new(docin)
     doc.write(test="")
     assert_equal(31, doc.doctype.size)
-
-    # Here's a little ditty from Tobias...
-    src = <<-EOL
-    <!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.0//EN"
-    "http://www.w3.org/TR/SVG/DTD/svg10.dtd"
-    [
-    <!-- <!ENTITY % fast-slow "0 0  .5 1">-->
-    <!--<!ENTITY % slow-fast ".5 0  1 1">-->
-    <!ENTITY hover_ani
-    '<animateTransform attributeName="transform"
-    type="scale" restart="whenNotActive" values="1;0.96"
-    dur="0.5s" calcMode="spline" keySplines="0 0  .5 1"
-    fill="freeze" begin="mouseover"/>
-    <animateTransform  attributeName="transform"
-    type="scale" restart="whenNotActive" values="0.96;1"
-    dur="0.5s" calcMode="spline" keySplines=".5 0  1 1"
-    fill="freeze" begin="mouseover+0.5s"/>'
-    >
-    ]
-    > <a/>
-    EOL
   end
 
   def test_document
     # Testing cloning
     source = "<element/>"
     doc = Document.new source
-    doc2 = Document.new doc
 
     # Testing Root
     assert_equal doc.root.name.to_s, "element"
@@ -642,11 +620,10 @@ class Tester < Test::Unit::TestCase
   end
 
   def test_line
-    doc = Document.new File.new(fixture_path("bad.xml"))
+    Document.new File.new(fixture_path("bad.xml"))
     assert_fail "There should have been an error"
   rescue Exception
     # We should get here
-    er = $!
     assert($!.line == 5, "Should have been an error on line 5, "+
       "but was reported as being on line #{$!.line}" )
   end
@@ -664,13 +641,11 @@ class Tester < Test::Unit::TestCase
   def test_exception
     source = SourceFactory.create_from "<a/>"
     p = ParseException.new( "dummy message", source )
-    s = p.to_s
     begin
       raise "dummy"
     rescue Exception
       p.continued_exception = $!
     end
-    s = p.to_s
   end
 
   def test_bad_content
@@ -682,7 +657,7 @@ class Tester < Test::Unit::TestCase
     assert_equal "content>content", tree_gt.elements[1].text
     # This isn't
     begin
-      tree_lt = Document.new in_lt
+      Document.new in_lt
       assert_fail "Should have gotten a parse error"
     rescue ParseException
     end
@@ -856,8 +831,6 @@ EOL
   def test_attlist_write
     file=File.new(fixture_path("foo.xml"))
     doc=Document.new file
-    root = doc.root 
-
     out = ''
     doc.write(out)
   end
@@ -865,7 +838,7 @@ EOL
   def test_more_namespaces
     assert_raise( REXML::UndefinedNamespaceException,
                    %Q{Should have gotten an Undefined Namespace error} )  {
-      doc1 = Document.new("<r><p><n:c/></p></r>")
+      Document.new("<r><p><n:c/></p></r>")
     }
     doc2 = Document.new("<r xmlns:n='1'><p><n:c/></p></r>")
     es = XPath.match(doc2, '//c')
@@ -916,7 +889,7 @@ EOL
   end
 
   def test_oses_with_bad_EOLs
-    d = Document.new("\n\n\n<?xml version='1.0'?>\n\n\n<a/>\n\n")
+    Document.new("\n\n\n<?xml version='1.0'?>\n\n\n<a/>\n\n")
   end
 
   # Contributed (with patch to fix bug) by Kouhei
@@ -1024,7 +997,6 @@ EOL
     document.write(s)
   end
 
-  
   def test_write_cdata
     src = "<a>A</a>"
     doc = REXML::Document.new( src )
@@ -1045,7 +1017,7 @@ EOL
       <x:b x:n="foo"/>
     </a>
     EOL
-    d = REXML::Document.new( source )
+    d = Document.new( source )
     assert_equal( 'foo', REXML::XPath.first(d.root, "//x:b/@x:n").value )
     assert_equal( nil, REXML::XPath.first(d.root, "//x:b/@x:n", {}))
   end
@@ -1233,17 +1205,17 @@ EOL
   def test_ticket_21
     src = "<foo bar=value/>"
     assert_raise( ParseException, "invalid XML should be caught" ) {
-      d = REXML::Document.new(src)
+      Document.new(src)
     }
     begin
-      d = REXML::Document.new(src)
+      Document.new(src)
     rescue
       assert_match( /missing attribute quote/, $!.message )
     end
   end
 
   def test_ticket_63
-    d = REXML::Document.new(File.new(fixture_path("t63-1.xml")))
+    Document.new(File.new(fixture_path("t63-1.xml")))
   end
 
   def test_ticket_75
@@ -1275,9 +1247,9 @@ EOL
 
   def test_ticket_88
     doc = REXML::Document.new("<?xml version=\"1.0\" encoding=\"shift_jis\"?>")
-    assert_equal("<?xml version='1.0' encoding='Shift_JIS'?>", doc.to_s)
+    assert_equal("<?xml version='1.0' encoding='SHIFT_JIS'?>", doc.to_s)
     doc = REXML::Document.new("<?xml version = \"1.0\" encoding = \"shift_jis\"?>")
-    assert_equal("<?xml version='1.0' encoding='Shift_JIS'?>", doc.to_s)
+    assert_equal("<?xml version='1.0' encoding='SHIFT_JIS'?>", doc.to_s)
   end
 
   def test_ticket_85
@@ -1295,8 +1267,6 @@ ENDXML
   </bar>
 </foo>"
 
-    zml = "<foo><bar><bob name='jimmy'/></bar></foo>"
-    
     # The pretty printer ignores all whitespace, anyway so output1 == output2
     f = REXML::Formatters::Pretty.new( 2 )
     d = Document.new( xml, :ignore_whitespace_nodes=>:all )
@@ -1358,7 +1328,7 @@ ENDXML
     # Per .2.5 Node Tests of XPath spec
     assert_raise( REXML::UndefinedNamespaceException,
                    %Q{Should have gotten an Undefined Namespace error} )  {
-      d = Document.new("<a><n:b/></a>") 
+      Document.new("<a><n:b/></a>")
     }
   end
 

test/rexml/test_encoding.rb

@@ -18,7 +18,7 @@ class EncodingTester < Test::Unit::TestCase
   def test_encoded_in_encoded_out
     doc = Document.new( @encoded )
     doc.write( out="" )
-    out.force_encoding('binary') if out.respond_to? :force_encoding
+    out.force_encoding(::Encoding::ASCII_8BIT)
     assert_equal( @encoded, out )
   end
 
@@ -26,12 +26,12 @@ class EncodingTester < Test::Unit::TestCase
   def test_encoded_in_change_out
     doc = Document.new( @encoded )
     doc.xml_decl.encoding = "UTF-8"
-    assert_equal( ::Encoding::UTF_8, doc.encoding )
+    assert_equal("UTF-8", doc.encoding)
     REXML::Formatters::Default.new.write( doc.root, out="" )
-    out.force_encoding('binary') if out.respond_to? :force_encoding
+    out.force_encoding(::Encoding::ASCII_8BIT)
     assert_equal( @not_encoded, out )
     char = XPath.first( doc, "/a/b/text()" ).to_s
-    char.force_encoding('binary') if char.respond_to? :force_encoding
+    char.force_encoding(::Encoding::ASCII_8BIT)
     assert_equal( "ĉ", char )
   end
 
@@ -39,7 +39,7 @@ class EncodingTester < Test::Unit::TestCase
   def test_encoded_in_different_out
     doc = Document.new( @encoded )
     REXML::Formatters::Default.new.write( doc.root, Output.new( out="", "UTF-8" ) )
-    out.force_encoding('binary') if out.respond_to? :force_encoding
+    out.force_encoding(::Encoding::ASCII_8BIT)
     assert_equal( @not_encoded, out )
   end
 
@@ -47,9 +47,9 @@ class EncodingTester < Test::Unit::TestCase
   def test_in_change_out
     doc = Document.new( @not_encoded )
     doc.xml_decl.encoding = "ISO-8859-3"
-    assert_equal( ::Encoding::ISO_8859_3, doc.encoding )
+    assert_equal("ISO-8859-3", doc.encoding)
     doc.write( out="" )
-    out.force_encoding('binary') if out.respond_to? :force_encoding
+    out.force_encoding(::Encoding::ASCII_8BIT)
     assert_equal( @encoded, out )
   end
 
@@ -57,7 +57,7 @@ class EncodingTester < Test::Unit::TestCase
   def test_in_different_out
     doc = Document.new( @not_encoded )
     doc.write( Output.new( out="", "ISO-8859-3" ) )
-    out.force_encoding('binary') if out.respond_to? :force_encoding
+    out.force_encoding(::Encoding::ASCII_8BIT)
     assert_equal( @encoded, out )
   end
 
@@ -66,10 +66,10 @@ class EncodingTester < Test::Unit::TestCase
   def test_in_different_access
     doc = Document.new <<-EOL
     <?xml version='1.0' encoding='ISO-8859-1'?>
-    <a a="ÿ">ÿ</a>
+    <a a="\xFF">\xFF</a>
     EOL
     expect = "\303\277"
-    expect.force_encoding('UTF-8') if expect.respond_to? :force_encoding
+    expect.force_encoding(::Encoding::UTF_8)
     assert_equal( expect, doc.elements['a'].attributes['a'] )
     assert_equal( expect, doc.elements['a'].text )
   end
@@ -86,7 +86,7 @@ class EncodingTester < Test::Unit::TestCase
 
   def test_ticket_110
     utf16 = REXML::Document.new(File.new(fixture_path("ticket_110_utf16.xml")))
-    assert_equal( ::Encoding::UTF_16BE, utf16.encoding )
+    assert_equal(utf16.encoding, "UTF-16")
     assert( utf16[0].kind_of?(REXML::XMLDecl))
   end
 end