Added SAX-based parser for XmlMini, using LibXML

Signed-off-by: Jeremy Kemper <jeremy@bitsweat.net>

2 files changed, 278 insertions, 0 deletions

diff --git a/activesupport/lib/active_support/xml_mini/libxmlsax.rb b/activesupport/lib/active_support/xml_mini/libxmlsax.rb
new file mode 100644
index 0000000000..dba1f8755a
--- /dev/null
+++ b/activesupport/lib/active_support/xml_mini/libxmlsax.rb
@@ -0,0 +1,84 @@
+require 'libxml'
+
+# = XmlMini LibXML implementation using a SAX-based parser
+module ActiveSupport
+  module XmlMini_LibXMLSAX
+    extend self
+
+    # Class that will build the hash while the XML document
+    # is being parsed using SAX events.
+    class HashBuilder
+
+      include LibXML::XML::SaxParser::Callbacks
+
+      CONTENT_KEY   = '__content__'.freeze
+      HASH_SIZE_KEY = '__hash_size__'.freeze
+
+      attr_reader :hash
+
+      def current_hash
+        @hash_stack.last
+      end
+
+      def on_start_document
+        @hash = {}
+        @hash_stack = [@hash]
+      end
+
+      def on_end_document
+        raise "Parse stack not empty!" if @hash_stack.size > 1
+      end
+
+      def on_error(error_message)
+        raise LibXML::XML::Error, error_message
+      end
+
+      def on_start_element(name, attrs = {})
+        new_hash = { CONTENT_KEY => '' }.merge(attrs)
+        new_hash[HASH_SIZE_KEY] = new_hash.size + 1
+
+        case current_hash[name]
+          when Array then current_hash[name] << new_hash
+          when Hash  then current_hash[name] = [current_hash[name], new_hash]
+          when nil   then current_hash[name] = new_hash
+        end
+
+        @hash_stack.push(new_hash)
+      end
+
+      def on_end_element(name)
+        if current_hash.length > current_hash.delete(HASH_SIZE_KEY) && current_hash[CONTENT_KEY].blank? || current_hash[CONTENT_KEY] == ''
+          current_hash.delete(CONTENT_KEY)
+        end
+        @hash_stack.pop
+      end
+
+      def on_characters(string)
+        current_hash[CONTENT_KEY] << string
+      end
+
+      alias_method :on_cdata_block, :on_characters
+    end
+
+    attr_accessor :document_class
+    self.document_class = HashBuilder
+
+    def parse(data)
+      if !data.respond_to?(:read)
+        data = StringIO.new(data || '')
+      end
+
+      char = data.getc
+      if char.nil?
+        {}
+      else
+        data.ungetc(char)
+        document = self.document_class.new
+        parser = LibXML::XML::SaxParser.io(data)
+        parser.callbacks = document
+        parser.parse
+        document.hash
+      end
+    end
+  end
+end
\ No newline at end of file
diff --git a/activesupport/test/xml_mini/libxmlsax_engine_test.rb b/activesupport/test/xml_mini/libxmlsax_engine_test.rb
new file mode 100644
index 0000000000..6d5b3673fa
--- /dev/null
+++ b/activesupport/test/xml_mini/libxmlsax_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 LibXMLSAXEngineTest < Test::Unit::TestCase
+  include ActiveSupport
+
+  def setup
+    @default_backend = XmlMini.backend
+    XmlMini.backend = 'LibXMLSAX'
+
+    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 = 'LibXMLSAX'
+    assert_equal XmlMini_LibXMLSAX, 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