Fix a bug that SAX2 parser doesn't expand the predefined entities for "characters"

lib/rexml/parsers/sax2parser.rb

@@ -157,25 +157,8 @@ def parse
               end
             end
           when :text
-            #normalized = @parser.normalize( event[1] )
-            #handle( :characters, normalized )
-            copy = event[1].clone
-
-            esub = proc { |match|
-              if @entities.has_key?($1)
-                @entities[$1].gsub(Text::REFERENCE, &esub)
-              else
-                match
-              end
-            }
-
-            copy.gsub!( Text::REFERENCE, &esub )
-            copy.gsub!( Text::NUMERICENTITY ) {|m|
-              m=$1
-              m = "0#{m}" if m[0] == ?x
-              [Integer(m)].pack('U*')
-            }
-            handle( :characters, copy )
+            unnormalized = @parser.unnormalize( event[1], @entities )
+            handle( :characters, unnormalized )
           when :entitydecl
             handle_entitydecl( event )
           when :processing_instruction, :comment, :attlistdecl,

lib/rexml/parsers/streamparser.rb

@@ -36,8 +36,8 @@ def parse
             @listener.tag_end( event[1] )
             @tag_stack.pop
           when :text
-            normalized = @parser.unnormalize( event[1] )
-            @listener.text( normalized )
+            unnormalized = @parser.unnormalize( event[1] )
+            @listener.text( unnormalized )
           when :processing_instruction
             @listener.instruction( *event[1,2] )
           when :start_doctype

test/test_pullparser.rb

@@ -82,6 +82,22 @@ def test_character_references
       assert_equal("B", events['b'])
     end
 
+    def test_text_entity_references
+      source = '<root><a>&lt;P&gt; &lt;I&gt; &lt;B&gt; Text &lt;/B&gt;  &lt;/I&gt;</a></root>'
+      parser = REXML::Parsers::PullParser.new( source )
+
+      events = []
+      while parser.has_next?
+        event = parser.pull
+        case event.event_type
+        when :text
+          events << event[1]
+        end
+      end
+
+      assert_equal(["<P> <I> <B> Text </B>  </I>"], events)
+    end
+
     def test_text_content_with_line_breaks
       source = "<root><a>A</a><b>B\n</b><c>C\r\n</c></root>"
       parser = REXML::Parsers::PullParser.new( source )

test/test_sax.rb

@@ -31,6 +31,17 @@ def test_entity_replacement
       assert_equal '--1234--', results[1]
     end
 
+    def test_characters_predefined_entities
+      source = '<root><a>&lt;P&gt; &lt;I&gt; &lt;B&gt; Text &lt;/B&gt;  &lt;/I&gt;</a></root>'
+
+      sax = Parsers::SAX2Parser.new( source )
+      results = []
+      sax.listen(:characters) {|x| results << x }
+      sax.parse
+
+      assert_equal(["<P> <I> <B> Text </B>  </I>"], results)
+    end
+
     def test_sax2
       File.open(fixture_path("documentation.xml")) do |f|
         parser = Parsers::SAX2Parser.new( f )