loofah 2.20.0 → 2.21.0.rc1

data/lib/loofah/html5/scrub.rb

@@ -1,4 +1,5 @@
 # frozen_string_literal: true
+
 require "cgi"
 require "crass"
 
@@ -6,9 +7,9 @@ module Loofah
   module HTML5 # :nodoc:
     module Scrub
       CONTROL_CHARACTERS = /[`\u0000-\u0020\u007f\u0080-\u0101]/
-      CSS_KEYWORDISH = /\A(#[0-9a-fA-F]+|rgb\(\d+%?,\d*%?,?\d*%?\)?|-?\d{0,3}\.?\d{0,10}(ch|cm|r?em|ex|in|lh|mm|pc|pt|px|Q|vmax|vmin|vw|vh|%|,|\))?)\z/
+      CSS_KEYWORDISH = /\A(#[0-9a-fA-F]+|rgb\(\d+%?,\d*%?,?\d*%?\)?|-?\d{0,3}\.?\d{0,10}(ch|cm|r?em|ex|in|lh|mm|pc|pt|px|Q|vmax|vmin|vw|vh|%|,|\))?)\z/ # rubocop:disable Layout/LineLength
       CRASS_SEMICOLON = { node: :semicolon, raw: ";" }
-      CSS_IMPORTANT = '!important'
+      CSS_IMPORTANT = "!important"
       CSS_PROPERTY_STRING_WITHOUT_EMBEDDED_QUOTES = /\A(["'])?[^"']+\1\z/
       DATA_ATTRIBUTE_NAME = /\Adata-[\w-]+\z/
 
@@ -26,7 +27,7 @@ module Loofah
               attr_node.node_name
             end
 
-            if attr_name =~ DATA_ATTRIBUTE_NAME
+            if DATA_ATTRIBUTE_NAME.match?(attr_name)
               next
             end
 
@@ -43,10 +44,12 @@ module Loofah
               scrub_attribute_that_allows_local_ref(attr_node)
             end
 
-            if SafeList::SVG_ALLOW_LOCAL_HREF.include?(node.name) && attr_name == "xlink:href" && attr_node.value =~ /^\s*[^#\s].*/m
-              attr_node.remove
-              next
-            end
+            next unless SafeList::SVG_ALLOW_LOCAL_HREF.include?(node.name) &&
+              attr_name == "xlink:href" &&
+              attr_node.value =~ /^\s*[^#\s].*/m
+
+            attr_node.remove
+            next
           end
 
           scrub_css_attribute(node)
@@ -66,29 +69,28 @@ module Loofah
         end
 
         def scrub_css(style)
+          url_flags = [:url, :bad_url]
           style_tree = Crass.parse_properties(style)
           sanitized_tree = []
 
           style_tree.each do |node|
             next unless node[:node] == :property
             next if node[:children].any? do |child|
-              [:url, :bad_url].include?(child[:node])
+              url_flags.include?(child[:node])
             end
 
             name = node[:name].downcase
             next unless SafeList::ALLOWED_CSS_PROPERTIES.include?(name) ||
-                SafeList::ALLOWED_SVG_PROPERTIES.include?(name) ||
-                SafeList::SHORTHAND_CSS_PROPERTIES.include?(name.split("-").first)
+              SafeList::ALLOWED_SVG_PROPERTIES.include?(name) ||
+              SafeList::SHORTHAND_CSS_PROPERTIES.include?(name.split("-").first)
 
             value = node[:children].map do |child|
               case child[:node]
               when :whitespace
                 nil
               when :string
-                if child[:raw] =~ CSS_PROPERTY_STRING_WITHOUT_EMBEDDED_QUOTES
+                if CSS_PROPERTY_STRING_WITHOUT_EMBEDDED_QUOTES.match?(child[:raw])
                   Crass::Parser.stringify(child)
-                else
-                  nil
                 end
               when :function
                 if SafeList::ALLOWED_CSS_FUNCTIONS.include?(child[:name].downcase)
@@ -97,8 +99,8 @@ module Loofah
               when :ident
                 keyword = child[:value]
                 if !SafeList::SHORTHAND_CSS_PROPERTIES.include?(name.split("-").first) ||
-                   SafeList::ALLOWED_CSS_KEYWORDS.include?(keyword) ||
-                   (keyword =~ CSS_KEYWORDISH)
+                    SafeList::ALLOWED_CSS_KEYWORDS.include?(keyword) ||
+                    (keyword =~ CSS_KEYWORDISH)
                   keyword
                 end
               else
@@ -107,6 +109,7 @@ module Loofah
             end.compact
 
             next if value.empty?
+
             value << CSS_IMPORTANT if node[:important]
             propstring = format("%s:%s", name, value.join(" "))
             sanitized_node = Crass.parse_properties(propstring).first
@@ -126,13 +129,9 @@ module Loofah
             when :url
               if node[:value].start_with?("#")
                 node[:raw]
-              else
-                nil
               end
             when :hash, :ident, :string
               node[:raw]
-            else
-              nil
             end
           end.compact
 
@@ -142,7 +141,8 @@ module Loofah
         def scrub_uri_attribute(attr_node)
           # this block lifted nearly verbatim from HTML5 sanitization
           val_unescaped = CGI.unescapeHTML(attr_node.value).gsub(CONTROL_CHARACTERS, "").downcase
-          if val_unescaped =~ /^[a-z0-9][-+.a-z0-9]*:/ && !SafeList::ALLOWED_PROTOCOLS.include?(val_unescaped.split(SafeList::PROTOCOL_SEPARATOR)[0])
+          if val_unescaped =~ /^[a-z0-9][-+.a-z0-9]*:/ &&
+              !SafeList::ALLOWED_PROTOCOLS.include?(val_unescaped.split(SafeList::PROTOCOL_SEPARATOR)[0])
             attr_node.remove
             return true
           elsif val_unescaped.split(SafeList::PROTOCOL_SEPARATOR)[0] == "data"
@@ -184,8 +184,8 @@ module Loofah
         end
 
         def cdata_needs_escaping?(node)
-          # Nokogiri's HTML4 parser on JRuby doesn't flag the child of a `style` or `script` tag as cdata, but it acts that way
-          node.cdata? || (Nokogiri.jruby? && node.text? && (node.parent.name == "style" || node.parent.name == "script"))
+          # Nokogiri's HTML4 parser on JRuby doesn't flag the child of a `style` tag as cdata, but it acts that way
+          node.cdata? || (Nokogiri.jruby? && node.text? && node.parent.name == "style")
         end
 
         def cdata_escape(node)
@@ -198,28 +198,28 @@ module Loofah
         end
 
         TABLE_FOR_ESCAPE_HTML__ = {
-          '<' => '&lt;',
-          '>' => '&gt;',
-          '&' => '&amp;',
+          "<" => "&lt;",
+          ">" => "&gt;",
+          "&" => "&amp;",
         }
 
         def escape_tags(string)
           # modified version of CGI.escapeHTML from ruby 3.1
           enc = string.encoding
-          unless enc.ascii_compatible?
+          if enc.ascii_compatible?
+            string = string.b
+            string.gsub!(/[<>&]/, TABLE_FOR_ESCAPE_HTML__)
+            string.force_encoding(enc)
+          else
             if enc.dummy?
               origenc = enc
               enc = Encoding::Converter.asciicompat_encoding(enc)
               string = enc ? string.encode(enc) : string.b
             end
-            table = Hash[TABLE_FOR_ESCAPE_HTML__.map {|pair|pair.map {|s|s.encode(enc)}}]
+            table = Hash[TABLE_FOR_ESCAPE_HTML__.map { |pair| pair.map { |s| s.encode(enc) } }]
             string = string.gsub(/#{"[<>&]".encode(enc)}/, table)
             string.encode!(origenc) if origenc
             string
-          else
-            string = string.b
-            string.gsub!(/[<>&]/, TABLE_FOR_ESCAPE_HTML__)
-            string.force_encoding(enc)
           end
         end
       end

data/lib/loofah/metahelpers.rb

@@ -1,12 +1,16 @@
 # frozen_string_literal: true
+
 module Loofah
   module MetaHelpers # :nodoc:
-    def self.add_downcased_set_members_to_all_set_constants(mojule)
-      mojule.constants.each do |constant_sym|
-        constant = mojule.const_get constant_sym
-        next unless Set === constant
-        constant.dup.each do |member|
-          constant.add member.downcase
+    class << self
+      def add_downcased_set_members_to_all_set_constants(mojule)
+        mojule.constants.each do |constant_sym|
+          constant = mojule.const_get(constant_sym)
+          next unless Set === constant
+
+          constant.dup.each do |member|
+            constant.add(member.downcase)
+          end
         end
       end
     end

data/lib/loofah/scrubber.rb

@@ -1,4 +1,5 @@
 # frozen_string_literal: true
+
 module Loofah
   #
   #  A RuntimeError raised when Loofah could not find an appropriate scrubber.
@@ -24,7 +25,7 @@ module Loofah
   #
   #  This can then be run on a document:
   #
-  #    Loofah.fragment("<span>foo</span><p>bar</p>").scrub!(span2div).to_s
+  #    Loofah.html5_fragment("<span>foo</span><p>bar</p>").scrub!(span2div).to_s
   #    # => "<div>foo</div><p>bar</p>"
   #
   #  Scrubbers can be run on a document in either a top-down traversal (the
@@ -32,7 +33,6 @@ module Loofah
   #  Scrubber::STOP to terminate the traversal of a subtree.
   #
   class Scrubber
-
     # Top-down Scrubbers may return CONTINUE to indicate that the subtree should be traversed.
     CONTINUE = Object.new.freeze
 
@@ -67,7 +67,9 @@ module Loofah
       unless [:top_down, :bottom_up].include?(direction)
         raise ArgumentError, "direction #{direction} must be one of :top_down or :bottom_up"
       end
-      @direction, @block = direction, block
+
+      @direction = direction
+      @block = block
     end
 
     #
@@ -84,7 +86,7 @@ module Loofah
     #  +scrub+, which will be called for each document node.
     #
     def scrub(node)
-      raise ScrubberNotFound, "No scrub method has been defined on #{self.class.to_s}"
+      raise ScrubberNotFound, "No scrub method has been defined on #{self.class}"
     end
 
     #
@@ -103,8 +105,8 @@ module Loofah
     def html5lib_sanitize(node)
       case node.type
       when Nokogiri::XML::Node::ELEMENT_NODE
-        if HTML5::Scrub.allowed_element? node.name
-          HTML5::Scrub.scrub_attributes node
+        if HTML5::Scrub.allowed_element?(node.name)
+          HTML5::Scrub.scrub_attributes(node)
           return Scrubber::CONTINUE
         end
       when Nokogiri::XML::Node::TEXT_NODE, Nokogiri::XML::Node::CDATA_SECTION_NODE
@@ -120,8 +122,8 @@ module Loofah
     def traverse_conditionally_top_down(node)
       if block
         return if block.call(node) == STOP
-      else
-        return if scrub(node) == STOP
+      elsif scrub(node) == STOP
+        return
       end
       node.children.each { |j| traverse_conditionally_top_down(j) }
     end

data/lib/loofah/scrubbers.rb

@@ -1,4 +1,5 @@
 # frozen_string_literal: true
+
 module Loofah
   #
   #  Loofah provides some built-in scrubbers for sanitizing with
@@ -11,7 +12,7 @@ module Loofah
   #  +:strip+ removes unknown/unsafe tags, but leaves behind the pristine contents:
   #
   #     unsafe_html = "ohai! <div>div is safe</div> <foo>but foo is <b>not</b></foo>"
-  #     Loofah.fragment(unsafe_html).scrub!(:strip)
+  #     Loofah.html5_fragment(unsafe_html).scrub!(:strip)
   #     => "ohai! <div>div is safe</div> but foo is <b>not</b>"
   #
   #
@@ -20,7 +21,7 @@ module Loofah
   #  +:prune+ removes unknown/unsafe tags and their contents (including their subtrees):
   #
   #     unsafe_html = "ohai! <div>div is safe</div> <foo>but foo is <b>not</b></foo>"
-  #     Loofah.fragment(unsafe_html).scrub!(:prune)
+  #     Loofah.html5_fragment(unsafe_html).scrub!(:prune)
   #     => "ohai! <div>div is safe</div> "
   #
   #
@@ -29,7 +30,7 @@ module Loofah
   #  +:escape+ performs HTML entity escaping on the unknown/unsafe tags:
   #
   #     unsafe_html = "ohai! <div>div is safe</div> <foo>but foo is <b>not</b></foo>"
-  #     Loofah.fragment(unsafe_html).scrub!(:escape)
+  #     Loofah.html5_fragment(unsafe_html).scrub!(:escape)
   #     => "ohai! <div>div is safe</div> &lt;foo&gt;but foo is &lt;b&gt;not&lt;/b&gt;&lt;/foo&gt;"
   #
   #
@@ -41,7 +42,7 @@ module Loofah
   #  layer of paint on top of the HTML input to make it look nice.
   #
   #     messy_markup = "ohai! <div id='foo' class='bar' style='margin: 10px'>div with attributes</div>"
-  #     Loofah.fragment(messy_markup).scrub!(:whitewash)
+  #     Loofah.html5_fragment(messy_markup).scrub!(:whitewash)
   #     => "ohai! <div>div with attributes</div>"
   #
   #  One use case for this scrubber is to clean up HTML that was
@@ -56,7 +57,7 @@ module Loofah
   #  +:nofollow+ adds a rel="nofollow" attribute to all links
   #
   #     link_farmers_markup = "ohai! <a href='http://www.myswarmysite.com/'>I like your blog post</a>"
-  #     Loofah.fragment(link_farmers_markup).scrub!(:nofollow)
+  #     Loofah.html5_fragment(link_farmers_markup).scrub!(:nofollow)
   #     => "ohai! <a href='http://www.myswarmysite.com/' rel="nofollow">I like your blog post</a>"
   #
   #
@@ -65,7 +66,7 @@ module Loofah
   #  +:noopener+ adds a rel="noopener" attribute to all links
   #
   #     link_farmers_markup = "ohai! <a href='http://www.myswarmysite.com/'>I like your blog post</a>"
-  #     Loofah.fragment(link_farmers_markup).scrub!(:noopener)
+  #     Loofah.html5_fragment(link_farmers_markup).scrub!(:noopener)
   #     => "ohai! <a href='http://www.myswarmysite.com/' rel="noopener">I like your blog post</a>"
   #
   #
@@ -74,7 +75,7 @@ module Loofah
   #  +:unprintable+ removes unprintable Unicode characters.
   #
   #     markup = "<p>Some text with an unprintable character at the end\u2028</p>"
-  #     Loofah.fragment(markup).scrub!(:unprintable)
+  #     Loofah.html5_fragment(markup).scrub!(:unprintable)
   #     => "<p>Some text with an unprintable character at the end</p>"
   #
   #  You may not be able to see the unprintable character in the above example, but there is a
@@ -90,19 +91,20 @@ module Loofah
     #  +:strip+ removes unknown/unsafe tags, but leaves behind the pristine contents:
     #
     #     unsafe_html = "ohai! <div>div is safe</div> <foo>but foo is <b>not</b></foo>"
-    #     Loofah.fragment(unsafe_html).scrub!(:strip)
+    #     Loofah.html5_fragment(unsafe_html).scrub!(:strip)
     #     => "ohai! <div>div is safe</div> but foo is <b>not</b>"
     #
     class Strip < Scrubber
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :bottom_up
       end
 
       def scrub(node)
         return CONTINUE if html5lib_sanitize(node) == CONTINUE
+
         node.before(node.children)
         node.remove
-        return STOP
+        STOP
       end
     end
 
@@ -112,18 +114,19 @@ module Loofah
     #  +:prune+ removes unknown/unsafe tags and their contents (including their subtrees):
     #
     #     unsafe_html = "ohai! <div>div is safe</div> <foo>but foo is <b>not</b></foo>"
-    #     Loofah.fragment(unsafe_html).scrub!(:prune)
+    #     Loofah.html5_fragment(unsafe_html).scrub!(:prune)
     #     => "ohai! <div>div is safe</div> "
     #
     class Prune < Scrubber
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :top_down
       end
 
       def scrub(node)
         return CONTINUE if html5lib_sanitize(node) == CONTINUE
+
         node.remove
-        return STOP
+        STOP
       end
     end
 
@@ -133,19 +136,20 @@ module Loofah
     #  +:escape+ performs HTML entity escaping on the unknown/unsafe tags:
     #
     #     unsafe_html = "ohai! <div>div is safe</div> <foo>but foo is <b>not</b></foo>"
-    #     Loofah.fragment(unsafe_html).scrub!(:escape)
+    #     Loofah.html5_fragment(unsafe_html).scrub!(:escape)
     #     => "ohai! <div>div is safe</div> &lt;foo&gt;but foo is &lt;b&gt;not&lt;/b&gt;&lt;/foo&gt;"
     #
     class Escape < Scrubber
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :top_down
       end
 
       def scrub(node)
         return CONTINUE if html5lib_sanitize(node) == CONTINUE
-        node.add_next_sibling Nokogiri::XML::Text.new(node.to_s, node.document)
+
+        node.add_next_sibling(Nokogiri::XML::Text.new(node.to_s, node.document))
         node.remove
-        return STOP
+        STOP
       end
     end
 
@@ -158,7 +162,7 @@ module Loofah
     #  layer of paint on top of the HTML input to make it look nice.
     #
     #     messy_markup = "ohai! <div id='foo' class='bar' style='margin: 10px'>div with attributes</div>"
-    #     Loofah.fragment(messy_markup).scrub!(:whitewash)
+    #     Loofah.html5_fragment(messy_markup).scrub!(:whitewash)
     #     => "ohai! <div>div with attributes</div>"
     #
     #  One use case for this scrubber is to clean up HTML that was
@@ -168,14 +172,14 @@ module Loofah
     #  Certainly not me.
     #
     class Whitewash < Scrubber
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :top_down
       end
 
       def scrub(node)
         case node.type
         when Nokogiri::XML::Node::ELEMENT_NODE
-          if HTML5::Scrub.allowed_element? node.name
+          if HTML5::Scrub.allowed_element?(node.name)
             node.attributes.each { |attr| node.remove_attribute(attr.first) }
             return CONTINUE if node.namespaces.empty?
           end
@@ -193,18 +197,19 @@ module Loofah
     #  +:nofollow+ adds a rel="nofollow" attribute to all links
     #
     #     link_farmers_markup = "ohai! <a href='http://www.myswarmysite.com/'>I like your blog post</a>"
-    #     Loofah.fragment(link_farmers_markup).scrub!(:nofollow)
+    #     Loofah.html5_fragment(link_farmers_markup).scrub!(:nofollow)
     #     => "ohai! <a href='http://www.myswarmysite.com/' rel="nofollow">I like your blog post</a>"
     #
     class NoFollow < Scrubber
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :top_down
       end
 
       def scrub(node)
         return CONTINUE unless (node.type == Nokogiri::XML::Node::ELEMENT_NODE) && (node.name == "a")
+
         append_attribute(node, "rel", "nofollow")
-        return STOP
+        STOP
       end
     end
 
@@ -214,35 +219,37 @@ module Loofah
     #  +:noopener+ adds a rel="noopener" attribute to all links
     #
     #     link_farmers_markup = "ohai! <a href='http://www.myswarmysite.com/'>I like your blog post</a>"
-    #     Loofah.fragment(link_farmers_markup).scrub!(:noopener)
+    #     Loofah.html5_fragment(link_farmers_markup).scrub!(:noopener)
     #     => "ohai! <a href='http://www.myswarmysite.com/' rel="noopener">I like your blog post</a>"
     #
     class NoOpener < Scrubber
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :top_down
       end
 
       def scrub(node)
         return CONTINUE unless (node.type == Nokogiri::XML::Node::ELEMENT_NODE) && (node.name == "a")
+
         append_attribute(node, "rel", "noopener")
-        return STOP
+        STOP
       end
     end
 
     # This class probably isn't useful publicly, but is used for #to_text's current implemention
     class NewlineBlockElements < Scrubber # :nodoc:
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :bottom_up
       end
 
       def scrub(node)
         return CONTINUE unless Loofah::Elements::LINEBREAKERS.include?(node.name)
+
         replacement = if Loofah::Elements::INLINE_LINE_BREAK.include?(node.name)
           "\n"
         else
           "\n#{node.content}\n"
         end
-        node.add_next_sibling Nokogiri::XML::Text.new(replacement, node.document)
+        node.add_next_sibling(Nokogiri::XML::Text.new(replacement, node.document))
         node.remove
       end
     end
@@ -253,7 +260,7 @@ module Loofah
     #  +:unprintable+ removes unprintable Unicode characters.
     #
     #     markup = "<p>Some text with an unprintable character at the end\u2028</p>"
-    #     Loofah.fragment(markup).scrub!(:unprintable)
+    #     Loofah.html5_fragment(markup).scrub!(:unprintable)
     #     => "<p>Some text with an unprintable character at the end</p>"
     #
     #  You may not be able to see the unprintable character in the above example, but there is a
@@ -263,7 +270,7 @@ module Loofah
     #     http://timelessrepo.com/json-isnt-a-javascript-subset
     #
     class Unprintable < Scrubber
-      def initialize
+      def initialize # rubocop:disable Lint/MissingSuper
         @direction = :top_down
       end
 
@@ -279,21 +286,23 @@ module Loofah
     #  A hash that maps a symbol (like +:prune+) to the appropriate Scrubber (Loofah::Scrubbers::Prune).
     #
     MAP = {
-      :escape => Escape,
-      :prune => Prune,
-      :whitewash => Whitewash,
-      :strip => Strip,
-      :nofollow => NoFollow,
-      :noopener => NoOpener,
-      :newline_block_elements => NewlineBlockElements,
-      :unprintable => Unprintable,
+      escape: Escape,
+      prune: Prune,
+      whitewash: Whitewash,
+      strip: Strip,
+      nofollow: NoFollow,
+      noopener: NoOpener,
+      newline_block_elements: NewlineBlockElements,
+      unprintable: Unprintable,
     }
 
-    #
-    #  Returns an array of symbols representing the built-in scrubbers
-    #
-    def self.scrubber_symbols
-      MAP.keys
+    class << self
+      #
+      #  Returns an array of symbols representing the built-in scrubbers
+      #
+      def scrubber_symbols
+        MAP.keys
+      end
     end
   end
 end

data/lib/loofah/version.rb

@@ -1,5 +1,6 @@
 # frozen_string_literal: true
+
 module Loofah
   # The version of Loofah you are using
-  VERSION = "2.20.0"
+  VERSION = "2.21.0.rc1"
 end

data/lib/loofah/xml/document.rb

@@ -1,4 +1,5 @@
 # frozen_string_literal: true
+
 module Loofah
   module XML # :nodoc:
     #

data/lib/loofah/xml/document_fragment.rb

@@ -1,4 +1,5 @@
 # frozen_string_literal: true
+
 module Loofah
   module XML # :nodoc:
     #
@@ -8,15 +9,10 @@ module Loofah
     #
     class DocumentFragment < Nokogiri::XML::DocumentFragment
       class << self
-        #
-        #  Overridden Nokogiri::XML::DocumentFragment
-        #  constructor. Applications should use Loofah.fragment to
-        #  parse a fragment.
-        #
         def parse(tags)
           doc = Loofah::XML::Document.new
           doc.encoding = tags.encoding.name if tags.respond_to?(:encoding)
-          self.new(doc, tags)
+          new(doc, tags)
         end
       end
     end