From f7ee8779fd1b00ab7ce06cd148cb7a31ed528de5 Mon Sep 17 00:00:00 2001
From: Timothy Elliott <tle@holymonkey.com>
Date: Fri, 1 May 2009 14:50:05 -0700
Subject: [PATCH] Add Hpricot backend for XmlMini

Based on the Nokogiri implementation. Two test cases from the Nokogiri implementation are omitted because Hpricot does not support entities or file references.
---
 .../lib/active_support/xml_mini/hpricot.rb         |   64 ++++++++
 activesupport/test/xml_mini/hpricot_engine_test.rb |  158 ++++++++++++++++++++
 2 files changed, 222 insertions(+), 0 deletions(-)
 create mode 100644 activesupport/lib/active_support/xml_mini/hpricot.rb
 create mode 100644 activesupport/test/xml_mini/hpricot_engine_test.rb

diff --git a/activesupport/lib/active_support/xml_mini/hpricot.rb b/activesupport/lib/active_support/xml_mini/hpricot.rb
new file mode 100644
index 0000000..22e9508
--- /dev/null
+++ b/activesupport/lib/active_support/xml_mini/hpricot.rb
@@ -0,0 +1,64 @@
+require 'hpricot'
+
+# = XmlMini Hpricot implementation. Based on the Nokogiri implementation.
+module ActiveSupport
+  module XmlMini_Hpricot #:nodoc:
+    extend self
+
+    # Parse an XML Document string into a simple hash using Hpricot.
+    # string::
+    #   XML Document string to parse
+    def parse(string)
+      if string.blank?
+        {}
+      else
+        doc = Hpricot.XML(string)
+        doc.to_hash
+      end
+    end
+
+    module Conversions #:nodoc:
+      module Document #:nodoc:
+        def to_hash
+          root.to_hash
+        end
+      end
+
+      module Node #:nodoc:  class Hpricot::Elem
+        CONTENT_ROOT = '__content__'
+
+        def to_hash(hash = {})
+          hash[name] ||= attributes
+          walker = lambda { |memo, child, callback|
+
+            if [Hpricot::Text, Hpricot::CData].include? child.class
+              unless child.inner_text.strip.empty?
+                (memo[CONTENT_ROOT] ||= '') << child.inner_text
+              end
+              next
+            end
+
+            name = child.name
+
+            child_hash = child.attributes
+            if memo[name]
+              memo[name] = [memo[name]].flatten
+              memo[name] << child_hash
+            else
+              memo[name] = child_hash
+            end
+
+            # Recusively walk children
+            child.each_child{ |c| callback.call(child_hash, c, callback) }
+          }
+
+          each_child{ |c| walker.call(hash[name], c, walker) }
+          hash
+        end
+      end
+    end
+
+    Hpricot::Doc.send(:include, Conversions::Document)
+    Hpricot::Elem.send(:include, Conversions::Node)
+  end
+end
diff --git a/activesupport/test/xml_mini/hpricot_engine_test.rb b/activesupport/test/xml_mini/hpricot_engine_test.rb
new file mode 100644
index 0000000..e0b6cdb
--- /dev/null
+++ b/activesupport/test/xml_mini/hpricot_engine_test.rb
@@ -0,0 +1,158 @@
+require 'abstract_unit'
+require 'active_support/xml_mini'
+require 'active_support/core_ext/hash/conversions'
+
+begin
+  gem 'hpricot'
+rescue Gem::LoadError
+  # Skip hpricot tests
+else
+
+require 'hpricot'
+
+class HpricotEngineTest < Test::Unit::TestCase
+  include ActiveSupport
+
+  def setup
+    @default_backend = XmlMini.backend
+    XmlMini.backend = 'Hpricot'
+  end
+
+  def teardown
+    XmlMini.backend = @default_backend
+  end
+
+#  def test_file_from_xml
+#    hash = Hash.from_xml(<<-eoxml)
+#      <blog>
+#        <logo type="file" name="logo.png" content_type="image/png">
+#        </logo>
+#      </blog>
+#    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 Hpricot::XML::SyntaxError do
+#      attack_xml = <<-EOT
+#      <?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>
+#      EOT
+#      Hash.from_xml(attack_xml)
+#    end
+#  end
+
+  def test_setting_hpricot_as_backend
+    XmlMini.backend = 'Hpricot'
+    assert_equal XmlMini_Hpricot, 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
+
+  private
+  def assert_equal_rexml(xml)
+    hash = XmlMini.with_backend('REXML') { XmlMini.parse(xml) }
+    assert_equal(hash, XmlMini.parse(xml))
+  end
+end
+
+end
-- 
1.5.4.3