diff --git a/activesupport/lib/active_support/xml_mini/jdom.rb b/activesupport/lib/active_support/xml_mini/jdom.rb new file mode 100644 index 0000000000..d795d55690 --- /dev/null +++ b/activesupport/lib/active_support/xml_mini/jdom.rb @@ -0,0 +1,162 @@ +raise "JRuby is required to use the JDOM backend for XmlMini" unless RUBY_PLATFORM =~ /java/ + +require 'jruby' +include Java + +import javax.xml.parsers.DocumentBuilder unless defined? DocumentBuilder +import javax.xml.parsers.DocumentBuilderFactory unless defined? DocumentBuilderFactory +import java.io.StringReader unless defined? StringReader +import org.xml.sax.InputSource unless defined? InputSource +import org.xml.sax.Attributes unless defined? Attributes +import org.w3c.dom.Node unless defined? Node + +# = XmlMini JRuby JDOM implementation +module ActiveSupport + module XmlMini_JDOM #:nodoc: + extend self + + CONTENT_KEY = '__content__'.freeze + + NODE_TYPE_NAMES = %w{ATTRIBUTE_NODE CDATA_SECTION_NODE COMMENT_NODE DOCUMENT_FRAGMENT_NODE + DOCUMENT_NODE DOCUMENT_TYPE_NODE ELEMENT_NODE ENTITY_NODE ENTITY_REFERENCE_NODE NOTATION_NODE + PROCESSING_INSTRUCTION_NODE TEXT_NODE} + + node_type_map = {} + NODE_TYPE_NAMES.each { |type| node_type_map[Node.send(type)] = type } + + # Parse an XML Document string into a simple hash using Java's jdom. + # string:: + # XML Document string to parse + def parse(string) + if string.blank? + {} + else + @dbf = DocumentBuilderFactory.new_instance + xml_string_reader = StringReader.new(string) + xml_input_source = InputSource.new(xml_string_reader) + doc = @dbf.new_document_builder.parse(xml_input_source) + merge_element!({}, doc.document_element) + end + end + + private + + # Convert an XML element and merge into the hash + # + # hash:: + # Hash to merge the converted element into. + # element:: + # XML element to merge into hash + def merge_element!(hash, element) + merge!(hash, element.tag_name, collapse(element)) + end + + # Actually converts an XML document element into a data structure. + # + # element:: + # The document element to be collapsed. + def collapse(element) + hash = get_attributes(element) + + child_nodes = element.child_nodes + if child_nodes.length > 0 + for i in 0...child_nodes.length + child = child_nodes.item(i) + merge_element!(hash, child) unless child.node_type == Node.TEXT_NODE + end + merge_texts!(hash, element) unless empty_content?(element) + hash + else + merge_texts!(hash, element) + end + end + + # Merge all the texts of an element into the hash + # + # hash:: + # Hash to add the converted emement to. + # element:: + # XML element whose texts are to me merged into the hash + def merge_texts!(hash, element) + text_children = texts(element) + if text_children.join.empty? + hash + else + # must use value to prevent double-escaping + merge!(hash, CONTENT_KEY, text_children.join) + end + end + + # Adds a new key/value pair to an existing Hash. If the key to be added + # already exists and the existing value associated with key is not + # an Array, it will be wrapped in an Array. Then the new value is + # appended to that Array. + # + # hash:: + # Hash to add key/value pair to. + # key:: + # Key to be added. + # value:: + # Value to be associated with key. + def merge!(hash, key, value) + if hash.has_key?(key) + if hash[key].instance_of?(Array) + hash[key] << value + else + hash[key] = [hash[key], value] + end + elsif value.instance_of?(Array) + hash[key] = [value] + else + hash[key] = value + end + hash + end + + # Converts the attributes array of an XML element into a hash. + # Returns an empty Hash if node has no attributes. + # + # element:: + # XML element to extract attributes from. + def get_attributes(element) + attribute_hash = {} + attributes = element.attributes + for i in 0...attributes.length + attribute_hash[attributes.item(i).name] = attributes.item(i).value + end + attribute_hash + end + + # Determines if a document element has text content + # + # element:: + # XML element to be checked. + def texts(element) + texts = [] + child_nodes = element.child_nodes + for i in 0...child_nodes.length + item = child_nodes.item(i) + if item.node_type == Node.TEXT_NODE + texts << item.get_data + end + end + texts + end + + # Determines if a document element has text content + # + # element:: + # XML element to be checked. + def empty_content?(element) + text = '' + child_nodes = element.child_nodes + for i in 0...child_nodes.length + item = child_nodes.item(i) + if item.node_type == Node.TEXT_NODE + text << item.get_data.strip + end + end + text.strip.length == 0 + end + end +end diff --git a/activesupport/test/xml_mini/jdom_engine_test.rb b/activesupport/test/xml_mini/jdom_engine_test.rb new file mode 100644 index 0000000000..b745228994 --- /dev/null +++ b/activesupport/test/xml_mini/jdom_engine_test.rb @@ -0,0 +1,153 @@ +require 'abstract_unit' +require 'active_support/xml_mini' + +if RUBY_PLATFORM =~ /java/ + +class JDOMEngineTest < Test::Unit::TestCase + include ActiveSupport + + def setup + @default_backend = XmlMini.backend + XmlMini.backend = 'JDOM' + end + + def teardown + XmlMini.backend = @default_backend + end + + # def test_file_from_xml + # hash = Hash.from_xml(<<-eoxml) + # + # + # + # + # eoxml + # assert hash.has_key?('blog') + # assert hash['blog'].has_key?('logo') + # + # file = hash['blog']['logo'] + # assert_equal 'logo.png', file.original_filename + # assert_equal 'image/png', file.content_type + # end + + def test_exception_thrown_on_expansion_attack + assert_raise NativeException do + attack_xml = <<-EOT + + + + + + + + + ]> + + &a; + + EOT + Hash.from_xml(attack_xml) + end + end + + def test_setting_JDOM_as_backend + XmlMini.backend = 'JDOM' + assert_equal XmlMini_JDOM, 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) + + + a post + another post + + + eoxml + end + + def test_one_node_document_as_hash + assert_equal_rexml(<<-eoxml) + + eoxml + end + + def test_one_node_with_attributes_document_as_hash + assert_equal_rexml(<<-eoxml) + + eoxml + end + + def test_products_node_with_book_node_as_hash + assert_equal_rexml(<<-eoxml) + + + + eoxml + end + + def test_products_node_with_two_book_nodes_as_hash + assert_equal_rexml(<<-eoxml) + + + + + eoxml + end + + def test_single_node_with_content_as_hash + assert_equal_rexml(<<-eoxml) + + hello world + + eoxml + end + + def test_children_with_children + assert_equal_rexml(<<-eoxml) + + + + + + eoxml + end + + def test_children_with_text + assert_equal_rexml(<<-eoxml) + + + hello everyone + + + eoxml + end + + def test_children_with_non_adjacent_text + assert_equal_rexml(<<-eoxml) + + good + + hello everyone + + morning + + eoxml + end + + private + def assert_equal_rexml(xml) + hash = XmlMini.with_backend('REXML') { XmlMini.parse(xml) } + assert_equal(hash, XmlMini.parse(xml)) + end +end + +else + # don't run these test because we aren't running in JRuby +end