Merge branch 'bzr/golem' of /Users/distler/Sites/code/instiki

attic/lib/sanitize.rb

@@ -1,207 +1,262 @@
+# == Introduction
+#
+# This module provides sanitization of XHTML+MathML+SVG 
+# and of inline style attributes. Its genesis is {described here}[http://golem.ph.utexas.edu/~distler/blog/archives/001181.html].
+#
+# Uses the {HTML5lib parser}[http://code.google.com/p/html5lib/], so that the parsing behaviour should
+# resemble that of browsers.
+#
+#  sanitize_xhtml() is a case-sensitive sanitizer, suitable for XHTML
+#  sanitize_html() is a case-insensitive sanitizer suitable for HTML
+#  sanitize_rexml() sanitizes a REXML tree, returning a string
+#  safe_sanitize_xhtml() makes extra-sure that the result is well-formed XHTML
+#                        by running the output of sanitize_xhtml() through REXML
+#
+# == Files
+#
+# {sanitize.rb}[http://golem.ph.utexas.edu/~distler/code/instiki/svn/lib/sanitize.rb],
+# {HTML5lib}[http://golem.ph.utexas.edu/~distler/code/instiki/svn/vendor/plugins/HTML5lib/]
+#
+# == Author
+#
+# {Jacques Distler}[http://golem.ph.utexas.edu/~distler/]
+#
+# ==  License
+#
+# Ruby License
+
 module Sanitize
 
-# This module provides sanitization of XHTML+MathML+SVG
-# and of inline style attributes.
+  require 'html5/html5parser'
+  require 'html5/liberalxmlparser'
+  require 'html5/treewalkers'
+  require 'html5/treebuilders'
+  require 'html5/serializer'
+  require 'html5/sanitizer'
+  require 'stringsupport.rb'
+
+  include HTML5
+
+# Sanitize a string, parsed using XHTML parsing rules.
 #
-# Based heavily on Sam Ruby's code in the Universal FeedParser.
-
-  require 'html/tokenizer'
-  require 'node'
-
-  acceptable_elements = ['a', 'abbr', 'acronym', 'address', 'area', 'b',
-      'big', 'blockquote', 'br', 'button', 'caption', 'center', 'cite',
-      'code', 'col', 'colgroup', 'dd', 'del', 'dfn', 'dir', 'div', 'dl', 'dt',
-      'em', 'fieldset', 'font', 'form', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6',
-      'hr', 'i', 'img', 'input', 'ins', 'kbd', 'label', 'legend', 'li', 'map',
-      'menu', 'ol', 'optgroup', 'option', 'p', 'pre', 'q', 's', 'samp',
-      'select', 'small', 'span', 'strike', 'strong', 'sub', 'sup', 'table',
-      'tbody', 'td', 'textarea', 'tfoot', 'th', 'thead', 'tr', 'tt', 'u',
-      'ul', 'var']
-      
-  mathml_elements = ['maction', 'math', 'merror', 'mfrac', 'mi',
-      'mmultiscripts', 'mn', 'mo', 'mover', 'mpadded', 'mphantom',
-      'mprescripts', 'mroot', 'mrow', 'mspace', 'msqrt', 'mstyle', 'msub',
-      'msubsup', 'msup', 'mtable', 'mtd', 'mtext', 'mtr', 'munder',
-      'munderover', 'none']
-      
-  svg_elements = ['a', 'animate', 'animateColor', 'animateMotion',
-      'animateTransform', 'circle', 'defs', 'desc', 'ellipse', 'font-face',
-      'font-face-name', 'font-face-src', 'g', 'glyph', 'hkern', 'image',
-      'linearGradient', 'line', 'marker', 'metadata', 'missing-glyph',
-      'mpath', 'path', 'polygon', 'polyline', 'radialGradient', 'rect',
-      'set', 'stop', 'svg', 'switch', 'text', 'title', 'tspan', 'use']
-      
-  acceptable_attributes = ['abbr', 'accept', 'accept-charset', 'accesskey',
-      'action', 'align', 'alt', 'axis', 'border', 'cellpadding',
-      'cellspacing', 'char', 'charoff', 'charset', 'checked', 'cite', 'class',
-      'clear', 'cols', 'colspan', 'color', 'compact', 'coords', 'datetime',
-      'dir', 'disabled', 'enctype', 'for', 'frame', 'headers', 'height',
-      'href', 'hreflang', 'hspace', 'id', 'ismap', 'label', 'lang',
-      'longdesc', 'maxlength', 'media', 'method', 'multiple', 'name',
-      'nohref', 'noshade', 'nowrap', 'prompt', 'readonly', 'rel', 'rev',
-      'rows', 'rowspan', 'rules', 'scope', 'selected', 'shape', 'size',
-      'span', 'src', 'start', 'style', 'summary', 'tabindex', 'target', 'title',
-      'type', 'usemap', 'valign', 'value', 'vspace', 'width', 'xml:lang']
-
-
-  mathml_attributes = ['actiontype', 'align', 'columnalign', 'columnalign',
-      'columnalign', 'columnlines', 'columnspacing', 'columnspan', 'depth',
-      'display', 'displaystyle', 'equalcolumns', 'equalrows', 'fence',
-      'fontstyle', 'fontweight', 'frame', 'height', 'linethickness', 'lspace',
-      'mathbackground', 'mathcolor', 'mathvariant', 'mathvariant', 'maxsize',
-      'minsize', 'other', 'rowalign', 'rowalign', 'rowalign', 'rowlines',
-      'rowspacing', 'rowspan', 'rspace', 'scriptlevel', 'selection',
-      'separator', 'stretchy', 'width', 'width', 'xlink:href', 'xlink:show',
-      'xlink:type', 'xmlns', 'xmlns:xlink']
-
-      
-  svg_attributes = ['accent-height', 'accumulate', 'additive', 'alphabetic',
-       'arabic-form', 'ascent', 'attributeName', 'attributeType',
-       'baseProfile', 'bbox', 'begin', 'by', 'calcMode', 'cap-height',
-       'class', 'color', 'color-rendering', 'content', 'cx', 'cy', 'd', 'dx',
-       'dy', 'descent', 'display', 'dur', 'end', 'fill', 'fill-rule',
-       'font-family', 'font-size', 'font-stretch', 'font-style', 'font-variant',
-       'font-weight', 'from', 'fx', 'fy', 'g1', 'g2', 'glyph-name', 
-       'gradientUnits', 'hanging', 'height', 'horiz-adv-x', 'horiz-origin-x',
-       'id', 'ideographic', 'k', 'keyPoints', 'keySplines', 'keyTimes',
-       'lang', 'marker-end', 'marker-mid', 'marker-start', 'markerHeight',
-       'markerUnits', 'markerWidth', 'mathematical', 'max', 'min', 'name',
-       'offset', 'opacity', 'orient', 'origin', 'overline-position',
-       'overline-thickness', 'panose-1', 'path', 'pathLength', 'points',
-       'preserveAspectRatio', 'r', 'refX', 'refY', 'repeatCount', 'repeatDur',
-       'requiredExtensions', 'requiredFeatures', 'restart', 'rotate', 'rx',
-       'ry', 'slope', 'stemh', 'stemv', 'stop-color', 'stop-opacity',
-       'strikethrough-position', 'strikethrough-thickness', 'stroke',
-       'stroke-dasharray', 'stroke-dashoffset', 'stroke-linecap',
-       'stroke-linejoin', 'stroke-miterlimit', 'stroke-opacity',
-       'stroke-width', 'systemLanguage', 'target',
-       'text-anchor', 'to', 'transform', 'type', 'u1', 'u2',
-       'underline-position', 'underline-thickness', 'unicode',
-       'unicode-range', 'units-per-em', 'values', 'version', 'viewBox',
-       'visibility', 'width', 'widths', 'x', 'x-height', 'x1', 'x2',
-       'xlink:actuate', 'xlink:arcrole', 'xlink:href', 'xlink:role',
-       'xlink:show', 'xlink:title', 'xlink:type', 'xml:base', 'xml:lang',
-       'xml:space', 'xmlns', 'xmlns:xlink', 'y', 'y1', 'y2', 'zoomAndPan']
-
-  attr_val_is_uri = ['href', 'src', 'cite', 'action', 'longdesc', 'xlink:href']
+# :call-seq:
+#    sanitize_xhtml(string)                    -> string
+#    sanitize_xhtml(string, {:encoding => 'iso-8859-1', :to_tree => true}) -> REXML::Document
+#
+# Unless otherwise specified, the string is assumed to be utf-8 encoded.
+# By default, the output is a string. But, optionally, you can return a REXML tree.
+#
+# The string returned is utf-8 encoded. If you want, you can use iconv to convert it to some other encoding.
+# (REXML trees are always utf-8 encoded.)
+  def sanitize_xhtml(html, options = {})
+    @encoding = 'utf-8'
+    @treebuilder = TreeBuilders::REXML::TreeBuilder
+    @to_tree = false
+    options.each do |name, value|
+      next unless %w(encoding treebuilder to_tree).include? name.to_s
+      if name.to_s == 'treebuilder'
+        @treebuilder =  HTML5lib::TreeBuilders.get_tree_builder(value)
+      else
+        instance_variable_set("@#{name}", value)
+      end
+    end
+    if @encoding == 'utf-8'
+      parsed = XHTMLParser.parse_fragment(html.to_utf8, {:tokenizer => HTMLSanitizer,
+        :lowercase_element_name => false, :lowercase_attr_name => false,
+        :encoding => @encoding, :tree => @treebuilder })
+    else
+      parsed = XHTMLParser.parse_fragment(html.to_ncr, {:tokenizer => HTMLSanitizer,
+        :lowercase_element_name => false, :lowercase_attr_name => false,
+        :encoding => @encoding, :tree => @treebuilder })
+    end      
+    return parsed if @to_tree
+    return parsed.to_s
+  end
   
-  acceptable_css_properties = ['azimuth', 'background-color',
-      'border-bottom-color', 'border-collapse', 'border-color',
-      'border-left-color', 'border-right-color', 'border-top-color', 'clear',
-      'color', 'cursor', 'direction', 'display', 'elevation', 'float', 'font',
-      'font-family', 'font-size', 'font-style', 'font-variant', 'font-weight',
-      'height', 'letter-spacing', 'line-height', 'overflow', 'pause',
-      'pause-after', 'pause-before', 'pitch', 'pitch-range', 'richness',
-      'speak', 'speak-header', 'speak-numeral', 'speak-punctuation',
-      'speech-rate', 'stress', 'text-align', 'text-decoration', 'text-indent',
-      'unicode-bidi', 'vertical-align', 'voice-family', 'volume',
-      'white-space', 'width']
+# Sanitize a string, parsed using XHTML parsing rules. Reparse the result to
+#    ensure well-formedness. 
+#
+# :call-seq:
+#    safe_sanitize_xhtml(string)                    -> string
+#
+# Unless otherwise specified, the string is assumed to be utf-8 encoded.
+#
+# The string returned is utf-8 encoded. If you want, you can use iconv to convert it to some other encoding.
+# (REXML trees are always utf-8 encoded.)
+  def safe_sanitize_xhtml(html, options = {})
+    options[:to_tree] = false
+    sanitized = sanitize_xhtml(html, options)
+    doc = REXML::Document.new("<div xmlns='http://www.w3.org/1999/xhtml'>#{sanitized}</div>")
+    sanitized = doc.to_s.gsub(/\A<div xmlns='http:\/\/www.w3.org\/1999\/xhtml'>(.*)<\/div>\Z/m, '\1')
+    rescue REXML::ParseException
+      sanitized = sanitized.escapeHTML
+  end 
 
-  acceptable_css_keywords = ['auto', 'aqua', 'black', 'block', 'blue',
-      'bold', 'both', 'bottom', 'brown', 'center', 'collapse', 'dashed',
-      'dotted', 'fuchsia', 'gray', 'green', '!important', 'italic', 'left',
-      'lime', 'maroon', 'medium', 'none', 'navy', 'normal', 'nowrap', 'olive',
-      'pointer', 'purple', 'red', 'right', 'solid', 'silver', 'teal', 'top',
-      'transparent', 'underline', 'white', 'yellow']
+# Sanitize a string, parsed using HTML parsing rules.
+#
+# :call-seq:
+#    sanitize_html( string )                    ->  string
+#    sanitize_html( string, {:encoding => 'iso-8859-1', :to_tree => true} ) ->  REXML::Document
+#
+# Unless otherwise specified, the string is assumed to be utf-8 encoded.
+# By default, the output is a string. But, optionally, you can return a REXML tree.
+#
+# The string returned is utf-8 encoded. If you want, you can use iconv to convert it to some other encoding.
+# (REXML trees are always utf-8 encoded.)
+  def sanitize_html(html, options = {})
+    @encoding = 'utf-8'
+    @treebuilder = TreeBuilders::REXML::TreeBuilder
+    @to_tree = false
+    options.each do |name, value|
+      next unless %w(encoding treebuilder to_tree).include? name.to_s
+      if name.to_s == 'treebuilder'
+        @treebuilder =  HTML5lib::TreeBuilders.get_tree_builder(value)
+      else
+        instance_variable_set("@#{name}", value)
+      end
+    end
+    if @encoding == 'utf-8'
+      parsed = HTMLParser.parse_fragment(html.to_utf8, {:tokenizer => HTMLSanitizer,
+        :encoding => @encoding, :tree => @treebuilder })
+    else
+      parsed = HTMLParser.parse_fragment(html.to_ncr, {:tokenizer => HTMLSanitizer,
+        :encoding => @encoding, :tree => @treebuilder })
+    end 
+    return parsed if @to_tree
+    return parsed.to_s
+  end
 
-  acceptable_svg_properties = [ 'fill', 'fill-opacity', 'fill-rule',
-      'stroke', 'stroke-width', 'stroke-linecap', 'stroke-linejoin',
-      'stroke-opacity']
+# Sanitize a REXML tree. The output is a string.
+#
+# :call-seq:
+#    sanitize_rexml(tree)                    -> string
+#
+  def sanitize_rexml(tree)
+    tokens = TreeWalkers.get_tree_walker('rexml2').new(tree)
+    XHTMLSerializer.serialize(tokens, {:encoding=>'utf-8',
+      :space_before_trailing_solidus => true,
+      :inject_meta_charset => false,
+      :sanitize => true})
+  end
+end
 
-  acceptable_protocols = [ 'ed2k', 'ftp', 'http', 'https', 'irc',
-      'mailto', 'news', 'gopher', 'nntp', 'telnet', 'webcal',
-      'xmpp', 'callto', 'feed', 'urn', 'aim', 'rsync', 'tag',
-      'ssh', 'sftp', 'rtsp', 'afs' ]
+require 'rexml/element'
+module REXML #:nodoc:
+  class Element
 
-      ALLOWED_ELEMENTS = acceptable_elements + mathml_elements + svg_elements  unless defined?(ALLOWED_ELEMENTS)
-      ALLOWED_ATTRIBUTES = acceptable_attributes + mathml_attributes + svg_attributes unless defined?(ALLOWED_ATTRIBUTES)
-      ALLOWED_CSS_PROPERTIES = acceptable_css_properties unless defined?(ALLOWED_CSS_PROPERTIES)
-      ALLOWED_CSS_KEYWORDS = acceptable_css_keywords unless defined?(ALLOWED_CSS_KEYWORDS)
-      ALLOWED_SVG_PROPERTIES = acceptable_svg_properties unless defined?(ALLOWED_SVG_PROPERTIES)
-      ALLOWED_PROTOCOLS = acceptable_protocols unless defined?(ALLOWED_PROTOCOLS)
-      ATTR_VAL_IS_URI = attr_val_is_uri unless defined?(ATTR_VAL_IS_URI)
+# Convert XHTML+MathML Named Entities in a REXML::Element to Numeric Character References
+#
+#  :call-seq:
+#     tree.to_ncr  -> REXML::Element
+#
+# REXML, typically, converts NCRs to utf-8 characters, which is what you'll see when you
+# access the resulting REXML document.
+#
+# Note that this method needs to traverse the entire tree, converting text nodes and attributes
+# for each element. This can be SLOW. It will often be faster to serialize to a string and then
+# use String.to_ncr instead.
+#
+    def to_ncr
+      self.each_element { |el|
+        el.texts.each_index  {|i|
+          el.texts[i].value = el.texts[i].to_s.to_ncr
+        }
+        el.attributes.each { |name,val|
+          el.attributes[name] = val.to_ncr
+        }
+        el.to_ncr if el.has_elements?
+      }
+      return self
+    end
+    
+# Convert XHTML+MathML Named Entities in a REXML::Element to UTF-8
+#
+#  :call-seq:
+#     tree.to_utf8  -> REXML::Element
+#
+# Note that this method needs to traverse the entire tree, converting text nodes and attributes 
+# for each element. This can be SLOW. It will often be faster to serialize to a string and then
+# use String.to_utf8 instead.
+#
+    def to_utf8
+      self.each_element { |el|
+        el.texts.each_index  {|i|
+          el.texts[i].value = el.texts[i].to_s.to_utf8
+        }
+        el.attributes.each { |name,val|
+          el.attributes[name] = val.to_utf8
+        }
+        el.to_utf8 if el.has_elements?
+      }
+      return self
+    end
 
-      # Sanitize the +html+, escaping all elements not in ALLOWED_ELEMENTS, and stripping out all
-      # attributes not in ALLOWED_ATTRIBUTES. Style attributes are parsed, and a restricted set,
-      # specified by ALLOWED_CSS_PROPERTIES and ALLOWED_CSS_KEYWORDS, are allowed through.
-      # attributes in ATTR_VAL_IS_URI are scanned, and only URI schemes specified in
-      # ALLOWED_PROTOCOLS are allowed.
-      # You can adjust what gets sanitized, by defining these constant arrays before this Module is loaded. 
-      #
-      #   sanitize_html('<script> do_nasty_stuff() </script>')
-      #    => &lt;script> do_nasty_stuff() &lt;/script>
-      #   sanitize_html('<a href="javascript: sucker();">Click here for $100</a>')
-      #    => <a>Click here for $100</a>
-      def sanitize_html(html)
-        if html.index("<")
-          tokenizer = HTML::Tokenizer.new(html)
-          new_text = ""
+  end
+end
 
-          while token = tokenizer.next
-            node = XHTML::Node.parse(nil, 0, 0, token, false)
-            new_text << case node.tag?
-              when true
-                if ALLOWED_ELEMENTS.include?(node.name)
-                  if node.closing != :close
-                    node.attributes.delete_if { |attr,v| !ALLOWED_ATTRIBUTES.include?(attr) }
-                    ATTR_VAL_IS_URI.each do |attr|
-                      val_unescaped = CGI.unescapeHTML(node.attributes[attr].to_s).gsub(/[\000-\040\177\s]+|\302*[\200-\240]/,'').downcase
-                      if val_unescaped =~ /^[a-z0-9][-+.a-z0-9]*:/ and !ALLOWED_PROTOCOLS.include?(val_unescaped.split(':')[0]) 
-                        node.attributes.delete attr 
-                      end
-                    end
-                    if node.attributes['style']
-                      node.attributes['style'] = sanitize_css(node.attributes['style']) 
-                    end
-                  end
-                  node.to_s
-                else
-                  node.to_s.gsub(/</, "&lt;")
-                end
-              else
-                node.to_s.gsub(/</, "&lt;")
-            end
-          end
+module HTML5 #:nodoc: all
+  module TreeWalkers
 
-          html = new_text
+    private
+
+    class << self
+      def [](name)
+        case name.to_s.downcase
+        when 'rexml'
+          require 'html5/treewalkers/rexml'
+          REXML::TreeWalker
+        when 'rexml2'
+          REXML2::TreeWalker
+        else
+          raise "Unknown TreeWalker #{name}"
         end
-        html
       end
-      
-      def sanitize_css(style)
-          # disallow urls
-          style = style.to_s.gsub(/url\s*\(\s*[^\s)]+?\s*\)\s*/, ' ')
 
-          # gauntlet
-          if style !~ /^([:,;#%.\sa-zA-Z0-9!]|\w-\w|\'[\s\w]+\'|\"[\s\w]+\"|\([\d,\s]+\))*$/
-             style = ''
-             return style
-          end
-          if style !~ /^(\s*[-\w]+\s*:\s*[^:;]*(;|$))*$/
-             style = ''
-             return style
-          end
+      alias :get_tree_walker :[]
+    end
 
-          clean = []
-          style.scan(/([-\w]+)\s*:\s*([^:;]*)/) do |prop,val|
-            if ALLOWED_CSS_PROPERTIES.include?(prop.downcase)
-              clean <<  prop + ': ' + val + ';'
-            elsif ['background','border','margin','padding'].include?(prop.split('-')[0].downcase) 
-              goodval = true
-              val.split().each do |keyword|
-                if !ALLOWED_CSS_KEYWORDS.include?(keyword) and 
-                   keyword !~ /^(#[0-9a-f]+|rgb\(\d+%?,\d*%?,?\d*%?\)?|\d{0,2}\.?\d{0,2}(cm|em|ex|in|mm|pc|pt|px|%|,|\))?)$/
-                  goodval = false
-                end
-              end
-              if goodval 
-                clean <<  prop + ': ' + val + ';'
-              end
-            elsif ALLOWED_SVG_PROPERTIES.include?(prop.downcase)
-               clean <<  prop + ': ' + val + ';'
+    module REXML2
+      class TreeWalker < HTML5::TreeWalkers::NonRecursiveTreeWalker
+
+        private
+
+        def node_details(node)
+          case node
+          when ::REXML::Document
+            [:DOCUMENT]
+          when ::REXML::Element
+            if !node.name
+              [:DOCUMENT_FRAGMENT]
+            else
+              [:ELEMENT, node.name,
+                node.attributes.map {|name,value| [name,value.to_utf8]},
+                node.has_elements? || node.has_text?]
             end
+          when ::REXML::Text
+            [:TEXT, node.value.to_utf8]
+          when ::REXML::Comment
+            [:COMMENT, node.string]
+          when ::REXML::DocType
+            [:DOCTYPE, node.name, node.public, node.system]
+          when ::REXML::XMLDecl
+            [nil]
+          else
+            [:UNKNOWN, node.class.inspect]
           end
+        end
 
-          style = clean.join(' ')
+        def first_child(node)
+          node.children.first
+        end
+
+        def next_sibling(node)
+          node.next_sibling
+        end
+
+        def parent(node)
+          node.parent
+        end
       end
-end      
+    end
+  end
+end

attic/test/unit/sanitize_test.rb

@@ -1,187 +1,189 @@
 #!/usr/bin/env ruby
 
-require File.expand_path(File.join(File.dirname(__FILE__), '/../test_helper'))
+require File.expand_path(File.dirname(__FILE__) + '/../test_helper')
 require 'sanitize'
+require 'json'
+
 
 class SanitizeTest < Test::Unit::TestCase
+
   include Sanitize
 
   def setup
 
   end
 
-  Sanitize::ALLOWED_ELEMENTS.each do |tag_name|
+  def do_sanitize_xhtml stream
+    safe_sanitize_xhtml(stream)
+  end
+
+  def check_sanitization(input, htmloutput, xhtmloutput, rexmloutput)
+    assert_equal htmloutput, do_sanitize_xhtml(input)
+  end
+  
+  def rexml_doc(string)
+    REXML::Document.new(
+      "<div xmlns='http://www.w3.org/1999/xhtml'>#{string}</div>")
+  end
+  
+  def my_rex(string)
+    sanitize_rexml(rexml_doc(string.to_utf8)).gsub(/\A<div xmlns="http:\/\/www.w3.org\/1999\/xhtml">(.*)<\/div>\Z/m, '\1')
+  end
+
+  def test_sanitize_named_entities
+    input = '<p>Greek &phis; &phi;, double-struck &Aopf;, numeric &#x1D538; &#8279;, uppercase &TRADE; &LT;</p>'
+    output = "<p>Greek \317\225 \317\206, double-struck \360\235\224\270, numeric \360\235\224\270 \342\201\227, uppercase \342\204\242 &lt;</p>"
+    output2 = "<p>Greek \317\225 \317\206, double-struck \360\235\224\270, numeric &#x1D538; &#8279;, uppercase \342\204\242 &lt;</p>"
+    assert_equal(output, sanitize_xhtml(input))
+    assert_equal(output, sanitize_html(input))
+    assert_equal(output, my_rex(input))
+    assert_equal(output2, input.to_utf8)
+  end
+  
+  def test_sanitize_malformed_utf8
+    input = "<p>\357elephant &AMP; \302ivory</p>"
+    output = "<p>\357\277\275elephant &amp; \357\277\275ivory</p>"
+    check_sanitization(input, output, output, output)
+  end    
+
+  Sanitizer::ALLOWED_ELEMENTS.each do |tag_name|
     define_method "test_should_allow_#{tag_name}_tag" do
-      assert_equal "<#{tag_name} title=\"1\">foo &lt;bad>bar&lt;/bad> baz</#{tag_name}>",
-        sanitize_html("<#{tag_name} title='1'>foo <bad>bar</bad> baz</#{tag_name}>")
+      input       = "<#{tag_name} title='1'>foo <bad>bar</bad> baz</#{tag_name}>"
+      htmloutput  = "<#{tag_name.downcase} title='1'>foo &lt;bad&gt;bar&lt;/bad&gt; baz</#{tag_name.downcase}>"
+      xhtmloutput = "<#{tag_name} title='1'>foo &lt;bad&gt;bar&lt;/bad&gt; baz</#{tag_name}>"
+      rexmloutput = xhtmloutput
+      
+      if %w[caption colgroup optgroup option tbody td tfoot th thead tr].include?(tag_name)
+        htmloutput = "foo &lt;bad&gt;bar&lt;/bad&gt; baz"
+        xhtmloutput = htmloutput
+      elsif tag_name == 'col'
+        htmloutput = "foo &lt;bad&gt;bar&lt;/bad&gt; baz"
+        xhtmloutput = htmloutput
+        rexmloutput = "<col title='1' />"
+      elsif tag_name == 'table'
+        htmloutput = "foo &lt;bad&gt;bar&lt;/bad&gt;baz<table title='1'> </table>"
+        xhtmloutput = htmloutput
+      elsif tag_name == 'image'
+        htmloutput = "<img title='1'/>foo &lt;bad&gt;bar&lt;/bad&gt; baz"
+        xhtmloutput = htmloutput
+        rexmloutput = "<image title='1'>foo &lt;bad&gt;bar&lt;/bad&gt; baz</image>"
+      elsif VOID_ELEMENTS.include?(tag_name)
+        htmloutput = "<#{tag_name} title='1'/>foo &lt;bad&gt;bar&lt;/bad&gt; baz"
+        xhtmloutput = htmloutput
+        htmloutput += '<br/>' if tag_name == 'br'
+        rexmloutput =  "<#{tag_name} title='1' />"
+      end
+      check_sanitization(input, xhtmloutput, xhtmloutput, rexmloutput)
     end
   end
 
-  Sanitize::ALLOWED_ELEMENTS.each do |tag_name|
+  Sanitizer::ALLOWED_ELEMENTS.each do |tag_name|
     define_method "test_should_forbid_#{tag_name.upcase}_tag" do
-      assert_equal "&lt;#{tag_name.upcase} title=\"1\">foo &lt;bad>bar&lt;/bad> baz&lt;/#{tag_name.upcase}>",
-        sanitize_html("<#{tag_name.upcase} title='1'>foo <bad>bar</bad> baz</#{tag_name.upcase}>")
+      input = "<#{tag_name.upcase} title='1'>foo <bad>bar</bad> baz</#{tag_name.upcase}>"
+      output = "&lt;#{tag_name.upcase} title=\"1\"&gt;foo &lt;bad&gt;bar&lt;/bad&gt; baz&lt;/#{tag_name.upcase}&gt;"
+      xhtmloutput = "&lt;#{tag_name.upcase} title='1'&gt;foo &lt;bad&gt;bar&lt;/bad&gt; baz&lt;/#{tag_name.upcase}&gt;"
+      check_sanitization(input, output, xhtmloutput, output)
     end
   end
 
-  Sanitize::ALLOWED_ATTRIBUTES.each do |attribute_name|
-    if attribute_name != 'style'
-      define_method "test_should_allow_#{attribute_name}_attribute" do
-        assert_equal "<p #{attribute_name}=\"foo\">foo &lt;bad>bar&lt;/bad> baz</p>",
-          sanitize_html("<p #{attribute_name}='foo'>foo <bad>bar</bad> baz</p>")
+  Sanitizer::ALLOWED_ATTRIBUTES.each do |attribute_name|
+    next if attribute_name == 'style' || attribute_name.include?(':')
+    define_method "test_should_allow_#{attribute_name}_attribute" do
+      input = "<p #{attribute_name}='foo'>foo <bad>bar</bad> baz</p>"
+      output = "<p #{attribute_name}='foo'>foo &lt;bad&gt;bar&lt;/bad&gt; baz</p>"
+      htmloutput = "<p #{attribute_name.downcase}='foo'>foo &lt;bad&gt;bar&lt;/bad&gt; baz</p>"
+      check_sanitization(input, output, output, output)
+    end
+  end
+
+  Sanitizer::ALLOWED_ATTRIBUTES.each do |attribute_name|
+    define_method "test_should_forbid_#{attribute_name.upcase}_attribute" do
+      input = "<p #{attribute_name.upcase}='display: none;'>foo <bad>bar</bad> baz</p>"
+      output =  "<p>foo &lt;bad&gt;bar&lt;/bad&gt; baz</p>"
+      check_sanitization(input, output, output, output)
+    end
+  end
+
+  Sanitizer::ALLOWED_PROTOCOLS.each do |protocol|
+    define_method "test_should_allow_#{protocol}_uris" do
+      input = %(<a href="#{protocol}">foo</a>)
+      output = "<a href='#{protocol}'>foo</a>"
+      check_sanitization(input, output, output, output)
+    end
+  end
+
+  Sanitizer::ALLOWED_PROTOCOLS.each do |protocol|
+    define_method "test_should_allow_uppercase_#{protocol}_uris" do
+      input = %(<a href="#{protocol.upcase}">foo</a>)
+      output = "<a href='#{protocol.upcase}'>foo</a>"
+      check_sanitization(input, output, output, output)
+    end
+  end
+
+  Sanitizer::SVG_ALLOW_LOCAL_HREF.each do |tag_name|
+    next unless Sanitizer::ALLOWED_ELEMENTS.include?(tag_name)
+    define_method "test_#{tag_name}_should_allow_local_href_with_ns_decl" do
+      input = %(<#{tag_name} xlink:href="#foo" xmlns:xlink='http://www.w3.org/1999/xlink'/>)
+      output = "<#{tag_name.downcase} xlink:href='#foo' xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      xhtmloutput = "<#{tag_name} xlink:href='#foo' xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      check_sanitization(input, xhtmloutput, xhtmloutput, xhtmloutput)
+    end
+
+    define_method "test_#{tag_name}_should_allow_local_href_with_newline_and_ns_decl" do
+      input = %(<#{tag_name} xlink:href="\n#foo" xmlns:xlink='http://www.w3.org/1999/xlink'/>)
+      output = "<#{tag_name.downcase} xlink:href='\n#foo' xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      xhtmloutput = "<#{tag_name} xlink:href='\n#foo' xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      check_sanitization(input, xhtmloutput, xhtmloutput, xhtmloutput)
+    end
+
+    define_method "test_#{tag_name}_should_forbid_local_href_without_ns_decl" do
+      input = %(<#{tag_name} xlink:href="#foo"/>)
+      output = "&lt;#{tag_name.downcase} xlink:href='#foo'/>"
+      xhtmloutput = "&lt;#{tag_name} xlink:href=&#39;#foo&#39;&gt;&lt;/#{tag_name}&gt;"
+      check_sanitization(input, xhtmloutput, xhtmloutput, xhtmloutput)
+    end
+
+    define_method "test_#{tag_name}_should_forbid_local_href_with_newline_without_ns_decl" do
+      input = %(<#{tag_name} xlink:href="\n#foo"/>)
+      output = "&lt;#{tag_name.downcase} xlink:href='\n#foo'/>"
+      xhtmloutput = "&lt;#{tag_name} xlink:href=&#39;\n#foo&#39;&gt;&lt;/#{tag_name}&gt;"
+      check_sanitization(input, xhtmloutput, xhtmloutput, xhtmloutput)
+    end
+
+    define_method "test_#{tag_name}_should_forbid_nonlocal_href_with_ns_decl" do
+      input = %(<#{tag_name} xlink:href="http://bad.com/foo" xmlns:xlink='http://www.w3.org/1999/xlink'/>)
+      output = "<#{tag_name.downcase} xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      xhtmloutput = "<#{tag_name} xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      check_sanitization(input, xhtmloutput, xhtmloutput, xhtmloutput)
+    end
+
+    define_method "test_#{tag_name}_should_forbid_nonlocal_href_with_newline_and_ns_decl" do
+      input = %(<#{tag_name} xlink:href="\nhttp://bad.com/foo" xmlns:xlink='http://www.w3.org/1999/xlink'/>)
+      output = "<#{tag_name.downcase} xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      xhtmloutput = "<#{tag_name} xmlns:xlink='http://www.w3.org/1999/xlink'/>"
+      check_sanitization(input, xhtmloutput, xhtmloutput, xhtmloutput)
+    end
+  end
+
+  def test_should_handle_astral_plane_characters
+    input = "<p>&#x1d4b5; &#x1d538;</p>"
+    output = "<p>\360\235\222\265 \360\235\224\270</p>"
+    check_sanitization(input, output, output, output)
+
+    input = "<p><tspan>\360\235\224\270</tspan> a</p>"
+    output = "<p><tspan>\360\235\224\270</tspan> a</p>"
+    check_sanitization(input, output, output, output)
+  end
+  
+    JSON::parse(open(File.expand_path(File.join(File.dirname(__FILE__), '/../sanitizer.dat'))).read).each do |test|
+      define_method "test_#{test['name']}" do
+        check_sanitization(
+          test['input'],
+          test['output'],
+          test['xhtml'] || test['output'],
+          test['rexml'] || test['output']
+        )
       end
     end
-  end
-
-  Sanitize::ALLOWED_ATTRIBUTES.each do |attribute_name|
-    define_method "test_should_forbid_#{attribute_name.upcase}_attribute" do
-      assert_equal "<p>foo &lt;bad>bar&lt;/bad> baz</p>",
-        sanitize_html("<p #{attribute_name.upcase}='display: none;'>foo <bad>bar</bad> baz</p>")
-    end
-  end
-
-  Sanitize::ALLOWED_PROTOCOLS.each do |protocol|
-    define_method "test_should_allow_#{protocol}_uris" do
-      assert_equal "<a href=\"#{protocol}\">foo</a>",
-        sanitize_html(%(<a href="#{protocol}">foo</a>))
-    end
-  end
-
-  Sanitize::ALLOWED_PROTOCOLS.each do |protocol|
-    define_method "test_should_allow_uppercase_#{protocol}_uris" do
-      assert_equal "<a href=\"#{protocol.upcase}\">foo</a>",
-        sanitize_html(%(<a href="#{protocol.upcase}">foo</a>))
-    end
-  end
-
-  def test_should_allow_anchors
-    assert_equal "<a href=\"foo\">&lt;script>baz&lt;/script></a>",
-     sanitize_html("<a href='foo' onclick='bar'><script>baz</script></a>")
-  end
-
-  # RFC 3986, sec 4.2
-  def test_allow_colons_in_path_component
-    assert_equal "<a href=\"./this:that\">foo</a>",
-      sanitize_html("<a href=\"./this:that\">foo</a>")
-  end
-
-  %w(src width height alt).each do |img_attr|
-    define_method "test_should_allow_image_#{img_attr}_attribute" do
-      assert_equal "<img #{img_attr}=\"foo\" />",
-        sanitize_html("<img #{img_attr}='foo' onclick='bar' />")
-    end
-  end
-
-  def test_should_handle_non_html
-    assert_equal 'abc',  sanitize_html("abc")
-  end
-
-  def test_should_handle_blank_text
-    assert_equal '', sanitize_html('')
-  end
-
-  [%w(img src), %w(a href)].each do |(tag, attr)|
-    define_method "test_should_strip_#{attr}_attribute_in_#{tag}_with_bad_protocols" do
-      assert_equal %(<#{tag} title="1">boo</#{tag}>), sanitize_html(%(<#{tag} #{attr}="javascript:XSS" title="1">boo</#{tag}>))
-    end
-  end
-
-  [%w(img src), %w(a href)].each do |(tag, attr)|
-    define_method "test_should_strip_#{attr}_attribute_in_#{tag}_with_bad_protocols_and_whitespace" do
-      assert_equal %(<#{tag} title="1">boo</#{tag}>), sanitize_html(%(<#{tag} #{attr}=" javascript:XSS" title="1">boo</#{tag}>))
-    end
-  end
-
-  [%(<img src="javascript:alert('XSS');" />), 
-   %(<img src=javascript:alert('XSS') />), 
-   %(<img src="JaVaScRiPt:alert('XSS')" />), 
-   %(<img src='javascript:alert(&quot;XSS&quot;)' />),
-   %(<img src='javascript:alert(String.fromCharCode(88,83,83))' />),
-   %(<img src='&#106;&#97;&#118;&#97;&#115;&#99;&#114;&#105;&#112;&#116;&#58;&#97;&#108;&#101;&#114;&#116;&#40;&#39;&#88;&#83;&#83;&#39;&#41;' />),
-   %(<img src='&#0000106;&#0000097;&#0000118;&#0000097;&#0000115;&#0000099;&#0000114;&#0000105;&#0000112;&#0000116;&#0000058;&#0000097;&#0000108;&#0000101;&#0000114;&#0000116;&#0000040;&#0000039;&#0000088;&#0000083;&#0000083;&#0000039;&#0000041' />),
-   %(<img src='&#x6A;&#x61;&#x76;&#x61;&#x73;&#x63;&#x72;&#x69;&#x70;&#x74;&#x3A;&#x61;&#x6C;&#x65;&#x72;&#x74;&#x28;&#x27;&#x58;&#x53;&#x53;&#x27;&#x29' />),
-   %(<img src="jav\tascript:alert('XSS');" />),
-   %(<img src="jav&#x09;ascript:alert('XSS');" />),
-   %(<img src="jav&#x0A;ascript:alert('XSS');" />),
-   %(<img src="jav&#x0D;ascript:alert('XSS');" />),
-   %(<img src=" &#14;  javascript:alert('XSS');" />),
-   %(<img src="&#x20;javascript:alert('XSS');" />),
-   %(<img src="&#xA0;javascript:alert('XSS');" />)].each_with_index do |img_hack, i|
-    define_method "test_should_not_fall_for_xss_image_hack_#{i}" do
-      assert_equal "<img />", sanitize_html(img_hack)
-    end
-  end
-
-  def test_should_sanitize_tag_broken_up_by_null
-    assert_equal "&lt;scr>alert(\"XSS\")&lt;/scr>", sanitize_html(%(<scr\0ipt>alert(\"XSS\")</scr\0ipt>))
-  end
-  
-  def test_should_sanitize_invalid_script_tag
-    assert_equal "&lt;script />&lt;/script>", sanitize_html(%(<script/XSS SRC="http://ha.ckers.org/xss.js"></script>))
-  end
-  
-  def test_should_sanitize_script_tag_with_multiple_open_brackets
-    assert_equal "&lt;&lt;script>alert(\"XSS\");//&lt;&lt;/script>", sanitize_html(%(<<script>alert("XSS");//<</script>))
-    assert_equal %(&lt;iframe src="http:" />&lt;), sanitize_html(%(<iframe src=http://ha.ckers.org/scriptlet.html\n<))
-  end
-  
-  def test_should_sanitize_unclosed_script
-    assert_equal "&lt;script src=\"http:\" /><b>", sanitize_html(%(<script src=http://ha.ckers.org/xss.js?<b>))
-  end
-  
-  def test_should_sanitize_half_open_scripts
-    assert_equal  "<img>", sanitize_html(%(<img src="javascript:alert('XSS')"))
-  end
-  
-  def test_should_not_fall_for_ridiculous_hack
-    img_hack = %(<img\nsrc\n=\n"\nj\na\nv\na\ns\nc\nr\ni\np\nt\n:\na\nl\ne\nr\nt\n(\n'\nX\nS\nS\n'\n)\n"\n />)
-    assert_equal "<img />", sanitize_html(img_hack)
-  end
-
-  def test_platypus
-    assert_equal %(<a href=\"http://www.ragingplatypus.com/\" style=\"display: block; width: 100%; height: 100%; background-color: black; background-image: ; background-x: center; background-y: center;\">never trust your upstream platypus</a>),
-       sanitize_html(%(<a href="http://www.ragingplatypus.com/" style="display:block; position:absolute; left:0; top:0; width:100%; height:100%; z-index:1; background-color:black; background-image:url(http://www.ragingplatypus.com/i/cam-full.jpg); background-x:center; background-y:center; background-repeat:repeat;">never trust your upstream platypus</a>))
-  end
-
-  def test_xul
-    assert_equal %(<p style="">fubar</p>),
-     sanitize_html(%(<p style="-moz-binding:url('http://ha.ckers.org/xssmoz.xml#xss')">fubar</p>))
-  end
-
-  def test_input_image
-    assert_equal %(<input type="image" />),
-      sanitize_html(%(<input type="image" src="javascript:alert('XSS');" />))
-  end
-
-  def test_non_alpha_non_digit
-    assert_equal "&lt;script />&lt;/script>",
-      sanitize_html(%(<script/XSS src="http://ha.ckers.org/xss.js"></script>))
-    assert_equal "<a>foo</a>",
-      sanitize_html('<a onclick!#$%&()*~+-_.,:;?@[/|\]^`=alert("XSS")>foo</a>')
-    assert_equal "<img />",
-      sanitize_html('<img/src="http://ha.ckers.org/xss.js"/>')
-  end
-
-  def test_img_dynsrc_lowsrc
-     assert_equal "<img />",
-       sanitize_html(%(<img dynsrc="javascript:alert('XSS')" />))
-     assert_equal "<img />",
-       sanitize_html(%(<img lowsrc="javascript:alert('XSS')" />))
-  end
-
-  def test_div_background_image_unicode_encoded
-    assert_equal '<div style="">foo</div>',
-      sanitize_html(%(<div style="background-image:\0075\0072\006C\0028'\006a\0061\0076\0061\0073\0063\0072\0069\0070\0074\003a\0061\006c\0065\0072\0074\0028.1027\0058.1053\0053\0027\0029'\0029">foo</div>))
-  end
-
-  def test_div_expression
-    assert_equal '<div style="">foo</div>',
-      sanitize_html(%(<div style="width: expression(alert('XSS'));">foo</div>))
-  end
-
-  def test_img_vbscript
-     assert_equal '<img />',
-       sanitize_html(%(<img src='vbscript:msgbox("XSS")' />))
-  end
 
 end