Added CDATA support to the XmlMini LibXML engine, adjusted whitespace…

… handling to closer match that of the REXML engine, and added a LibXML engine test

Signed-off-by: Michael Koziarski <michael@koziarski.com>

activesupport/lib/active_support/xml_mini/libxml.rb

@@ -13,8 +13,6 @@ def parse(data)
         data = StringIO.new(data || '')
       end
 
-      LibXML::XML.default_keep_blanks = false
-
       char = data.getc
       if char.nil?
         {}
@@ -44,9 +42,9 @@ module Node #:nodoc:
       # hash::
       #   Hash to merge the converted element into.
       def to_hash(hash={})
-        if text?
-          raise LibXML::XML::Error if content.length >= LIB_XML_LIMIT
-          hash[CONTENT_ROOT] = content
+        if text? || cdata?
+          raise LibXML::XML::Error if hash[CONTENT_ROOT].to_s.length + content.length >= LIB_XML_LIMIT
+          hash[CONTENT_ROOT] = hash[CONTENT_ROOT].to_s + content
         else
           sub_hash = insert_name_into_hash(hash, name)
           attributes_to_hash(sub_hash)
@@ -88,6 +86,11 @@ def insert_name_into_hash(hash, name)
         #   Hash to merge the children into.
         def children_to_hash(hash={})
           each { |child| child.to_hash(hash) }
+
+          if hash.length > 1 && hash[CONTENT_ROOT].blank?
+            hash.delete(CONTENT_ROOT)
+          end
+
           attributes_to_hash(hash)
           hash
         end

activesupport/test/xml_mini/libxml_engine_test.rb

@@ -0,0 +1,194 @@
+require 'abstract_unit'
+require 'active_support/xml_mini'
+require 'active_support/core_ext/hash/conversions'
+
+begin
+  require 'libxml'
+rescue LoadError
+  # Skip libxml tests
+else
+
+class LibxmlEngineTest < Test::Unit::TestCase
+  include ActiveSupport
+
+  def setup
+    @default_backend = XmlMini.backend
+    XmlMini.backend = 'LibXML'
+
+    LibXML::XML::Error.set_handler(&lambda { |error| }) #silence libxml, exceptions will do
+  end
+
+  def teardown
+    XmlMini.backend = @default_backend
+  end
+
+  def test_exception_thrown_on_expansion_attack
+    assert_raise LibXML::XML::Error do
+      attack_xml = %{<?xml version="1.0" encoding="UTF-8"?>
+      <!DOCTYPE member [
+        <!ENTITY a "&b;&b;&b;&b;&b;&b;&b;&b;&b;&b;">
+        <!ENTITY b "&c;&c;&c;&c;&c;&c;&c;&c;&c;&c;">
+        <!ENTITY c "&d;&d;&d;&d;&d;&d;&d;&d;&d;&d;">
+        <!ENTITY d "&e;&e;&e;&e;&e;&e;&e;&e;&e;&e;">
+        <!ENTITY e "&f;&f;&f;&f;&f;&f;&f;&f;&f;&f;">
+        <!ENTITY f "&g;&g;&g;&g;&g;&g;&g;&g;&g;&g;">
+        <!ENTITY g "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx">
+      ]>
+      <member>
+      &a;
+      </member>
+     }
+      Hash.from_xml(attack_xml)
+    end
+  end
+
+  def test_setting_libxml_as_backend
+    XmlMini.backend = 'LibXML'
+    assert_equal XmlMini_LibXML, XmlMini.backend
+  end
+
+  def test_blank_returns_empty_hash
+    assert_equal({}, XmlMini.parse(nil))
+    assert_equal({}, XmlMini.parse(''))
+  end
+
+  def test_array_type_makes_an_array
+    assert_equal_rexml(<<-eoxml)
+      <blog>
+        <posts type="array">
+          <post>a post</post>
+          <post>another post</post>
+        </posts>
+      </blog>
+    eoxml
+  end
+
+  def test_one_node_document_as_hash
+    assert_equal_rexml(<<-eoxml)
+    <products/>
+    eoxml
+  end
+
+  def test_one_node_with_attributes_document_as_hash
+    assert_equal_rexml(<<-eoxml)
+    <products foo="bar"/>
+    eoxml
+  end
+
+  def test_products_node_with_book_node_as_hash
+    assert_equal_rexml(<<-eoxml)
+    <products>
+      <book name="awesome" id="12345" />
+    </products>
+    eoxml
+  end
+
+  def test_products_node_with_two_book_nodes_as_hash
+    assert_equal_rexml(<<-eoxml)
+    <products>
+      <book name="awesome" id="12345" />
+      <book name="america" id="67890" />
+    </products>
+    eoxml
+  end
+
+  def test_single_node_with_content_as_hash
+    assert_equal_rexml(<<-eoxml)
+      <products>
+        hello world
+      </products>
+    eoxml
+  end
+
+  def test_children_with_children
+    assert_equal_rexml(<<-eoxml)
+    <root>
+      <products>
+        <book name="america" id="67890" />
+      </products>
+    </root>
+    eoxml
+  end
+
+  def test_children_with_text
+    assert_equal_rexml(<<-eoxml)
+    <root>
+      <products>
+        hello everyone
+      </products>
+    </root>
+    eoxml
+  end
+
+  def test_children_with_non_adjacent_text
+    assert_equal_rexml(<<-eoxml)
+    <root>
+      good
+      <products>
+        hello everyone
+      </products>
+      morning
+    </root>
+    eoxml
+  end
+
+  def test_parse_from_io
+    io = StringIO.new(<<-eoxml)
+    <root>
+      good
+      <products>
+        hello everyone
+      </products>
+      morning
+    </root>
+    eoxml
+    XmlMini.parse(io)
+  end
+
+  def test_children_with_simple_cdata
+    assert_equal_rexml(<<-eoxml)
+    <root>
+      <products>
+         <![CDATA[cdatablock]]>
+      </products>
+    </root>
+    eoxml
+  end
+
+  def test_children_with_multiple_cdata
+    assert_equal_rexml(<<-eoxml)
+    <root>
+      <products>
+         <![CDATA[cdatablock1]]><![CDATA[cdatablock2]]>
+      </products>
+    </root>
+    eoxml
+  end
+
+  def test_children_with_text_and_cdata
+    assert_equal_rexml(<<-eoxml)
+    <root>
+      <products>
+        hello <![CDATA[cdatablock]]>
+        morning
+      </products>
+    </root>
+    eoxml
+  end
+
+  def test_children_with_blank_text
+    assert_equal_rexml(<<-eoxml)
+    <root>
+      <products>   </products>
+    </root>
+    eoxml
+  end
+
+  private
+  def assert_equal_rexml(xml)
+    hash = XmlMini.with_backend('REXML') { XmlMini.parse(xml) }
+    assert_equal(hash, XmlMini.parse(xml))
+  end
+end
+
+end