RubyGems - feedtools - Versions diffs - 0.2.1 → 0.2.2 - Mend

feedtools 0.2.1 → 0.2.2

Files changed (9) hide show

data/CHANGELOG CHANGED Viewed

@@ -1,3 +1,11 @@
+== FeedTools 0.2.2
+ * fixed http redirection bug
+ * fixed several documentation typos
+ * still more unit tests
+ * improved support for atom
+ * minor improvements to the database caching mechanism
+ * more complete support for rss elements
+ * major improvements to the handling of tags containing html content
 == FeedTools 0.2.1
  * fixed incorrect dependancy on ActiveRecord 1.10.1
  * more unit tests

data/lib/feed_tools.rb CHANGED Viewed

@@ -25,7 +25,7 @@ FEED_TOOLS_ENV = ENV['FEED_TOOLS_ENV'] ||
                  ENV['RAILS_ENV'] ||
                  'production' # :nodoc:
-FEED_TOOLS_VERSION = "0.2.1"
+FEED_TOOLS_VERSION = "0.2.2"
 $:.unshift(File.dirname(__FILE__))
 $:.unshift(File.dirname(__FILE__) + "/../../activerecord/lib")
@@ -131,6 +131,19 @@ module FeedTools
       end
       return nil
     end
+    # Returns true if a connection to the database has been established and the
+    # required table structure is in place.
+    def DatabaseFeedCache.connected?
+      begin
+        ActiveRecord::Base.connection
+        return false if ActiveRecord::Base.configurations.nil?
+        return false unless DatabaseFeedCache.table_exists?
+      rescue => error
+        return false
+      end
+      return true
+    end
     # True if the appropriate database table already exists
     def DatabaseFeedCache.table_exists?
@@ -258,12 +271,25 @@ module FeedTools
   #  find_by_id
   #  find_by_url
   #  initialize_cache
+  #  connected?
   def FeedTools.feed_cache=(new_feed_cache)
     # TODO: ensure that the feed cache class actually does those things.
     # ==================================================================
     @feed_cache = new_feed_cache
   end
+  # Returns true if FeedTools.feed_cache is not nil and a connection with
+  # the cache has been successfully established.  Also returns false if an
+  # error is raised while trying to determine the status of the cache.
+  def FeedTools.feed_cache_connected?
+    begin
+      return false if FeedTools.feed_cache.nil?
+      return FeedTools.feed_cache.connected?
+    rescue
+      return false
+    end
+  end
   # Returns the currently used user agent string.
   def FeedTools.user_agent
     return @user_agent
@@ -455,6 +481,25 @@ module FeedTools
     return true
   end
+  # Escapes all html entities
+  def FeedTools.escape_entities(html)
+    escaped_html = CGI.escapeHTML(html)
+    unescaped_html.gsub!(/'/, "&apos;")
+    unescaped_html.gsub!(/"/, "&quot;")
+    return escaped_html
+  end
+  # Unescapes all html entities
+  def FeedTools.unescape_entities(html)
+    unescaped_html = html
+    unescaped_html.gsub!(/&#x26;/, "&amp;")
+    unescaped_html.gsub!(/&#38;/, "&amp;")
+    unescaped_html = CGI.unescapeHTML(unescaped_html)
+    unescaped_html.gsub!(/&apos;/, "'")
+    unescaped_html.gsub!(/&quot;/, "\"")
+    return unescaped_html
+  end
   # Removes all html tags from the html formatted text.
   def FeedTools.strip_html(html)
     # TODO: do this properly
@@ -467,6 +512,7 @@ module FeedTools
   def FeedTools.tidy_html(html)
     if FeedTools.tidy_enabled?
       is_fragment = true
+      html.gsub!(/&lt;!'/, "&amp;lt;!'")
       if (html.strip =~ /<html>(.|\n)*<body>/) != nil ||
           (html.strip =~ /<\/body>(.|\n)*<\/html>$/) != nil
         is_fragment = false
@@ -484,13 +530,15 @@ module FeedTools
         xml
       end
       if is_fragment
-        # Tidy puts <html>...<body>[our html]</body>...</html> in.
+        # Tidy sticks <html>...<body>[our html]</body>...</html> in.
         # We don't want this.
         tidy_html.strip!
         tidy_html.gsub!(/^<html>(.|\n)*<body>/, "")
         tidy_html.gsub!(/<\/body>(.|\n)*<\/html>$/, "")
         tidy_html.strip!
       end
+      tidy_html.gsub!(/&#x26;/, "&amp;")
+      tidy_html.gsub!(/&#38;/, "&amp;")
     else
       tidy_html = html
     end
@@ -502,7 +550,7 @@ module FeedTools
   # be escaped.  If mode is set to :strip, dangerous and unknown
   # elements and all children will be removed entirely.
   # Dangerous or unknown attributes are always removed.
-  def FeedTools.sanitize_html(html, mode=:escape)
+  def FeedTools.sanitize_html(html, mode=:strip)
     # Lists borrowed from Mark Pilgrim's feedparser
     acceptable_elements = ['a', 'abbr', 'acronym', 'address', 'area', 'b',
@@ -527,18 +575,14 @@ module FeedTools
       'span', 'src', 'start', 'summary', 'tabindex', 'target', 'title',
       'type', 'usemap', 'valign', 'value', 'vspace', 'width']
-    # Stupid hack to pass this unit test:
-    # http://feedparser.org/tests/wellformed/rss/
-    #   item_description_not_a_doctype.xml
-    html.gsub!(/<!'/, "&lt;!'")
-    # The closer we are to proper xhtml, the more accurate the
-    # sanitization will be.
-    html = FeedTools.tidy_html(html)
+    # Replace with appropriate named entities
+    html.gsub!(/&#x26;/, "&amp;")
+    html.gsub!(/&#38;/, "&amp;")
+    html.gsub!(/&lt;!'/, "&amp;lt;!'")
     # Hackity hack.  But it works, and it seems plenty fast enough.
     html_doc = HTree.parse_xml("<root>" + html + "</root>").to_rexml
     sanitize_node = lambda do |html_node|
       if html_node.respond_to? :children
         for child in html_node.children
@@ -564,13 +608,83 @@ module FeedTools
       html_node
     end
     sanitize_node.call(html_doc.root)
-    return html_doc.root.inner_xml
+    html = html_doc.root.inner_xml
+    return html
   end
   class Feed
     include REXML
     include AttributeDictionary
+    # Represents a feed/feed item's category
+    class Category
+      # The category term value
+      attr_accessor :term
+      # The categorization scheme
+      attr_accessor :scheme
+      # A human-readable description of the category
+      attr_accessor :label
+      # Relays any unknown methods to the term so that you can treat the
+      # category object as a string
+      def method_missing(msg, *params)
+        self.term.send(msg, params)
+      end
+      # Relays the to_s method to the term field
+      def to_s
+        self.term.to_s
+      end
+      # Relays the inspect method to the term field
+      def inspect
+        self.term.inspect
+      end
+    end
+    # Represents a feed/feed item's author
+    class Author
+      # The author's real name
+      attr_accessor :name
+      # The author's email address
+      attr_accessor :email
+      # The url of the author's homepage
+      attr_accessor :url
+      # The raw value of the author tag if present
+      attr_accessor :raw
+      # Relays any unknown methods to the name so that you can treat the
+      # author object as a string
+      def method_missing(msg, *params)
+        self.name.send(msg, params)
+      end
+      # Relays the to_s method to the name field
+      def to_s
+        self.name.to_s
+      end
+      # Relays the inspect method to the name field
+      def inspect
+        self.name.inspect
+      end
+    end
+    # Represents a feed's cloud.
+    class Cloud
+      # The domain of the cloud.
+      attr_accessor :domain
+      # The path for the cloud.
+      attr_accessor :path
+      # The port the cloud is listening on.
+      attr_accessor :port
+      # The web services protocol the cloud uses.
+      # Possible values are either "xml-rpc" or "soap".
+      attr_accessor :protocol
+      # The procedure to use to request notification.
+      attr_accessor :register_procedure
+    end
     # Loads the feed specified by the url, pulling the data from the cache if it hasn't expired.
     def Feed.open(url)
       # clean up the url
@@ -671,14 +785,14 @@ module FeedTools
           Net::HTTP.start(feed_uri.host, (feed_uri.port or 80)) do |http|
             response = http.request_get(feed_uri.path, http_headers)
             case response
             when Net::HTTPSuccess
               # We've reached the final destination, process all previous
               # redirections, and see if we need to update the url.
               for redirected_response in response_chain
                 if redirected_response.last.code.to_i == 301
-                  self.url = redirected_response.first
+                  self.url = redirected_response.last['location']
                 else
                   # Jump out as soon as we hit anything that isn't a
                   # permanently moved redirection.
@@ -690,7 +804,7 @@ module FeedTools
               if response.code.to_i == 304
                 response.error!
               else
-                if response['Location'].nil?
+                if response['location'].nil?
                   raise FeedAccessError,
                     "No location to redirect to supplied: " + response.code
                 end
@@ -913,6 +1027,30 @@ module FeedTools
       @cache_object = new_cache_object
     end
+    # Returns the type of feed
+    # Possible values:
+    # "rss", "atom", "cdf", "!okay/news"
+    def feed_type
+      if @feed_type.nil?
+        case self.root_node.name.downcase
+        when "feed"
+          @feed_type = "atom"
+        when "rdf:rdf"
+          @feed_type = "rss"
+        when "rss"
+          @feed_type = "rss"
+        when "channel"
+          @feed_type = "cdf"
+        end
+      end
+      return @feed_type
+    end
+    # Sets the default feed type
+    def feed_type=(new_feed_type)
+      @feed_type = new_feed_type
+    end
     # Returns the feed's unique id
     def id
       if @id.nil?
@@ -948,24 +1086,30 @@ module FeedTools
     # Returns the feed title
     def title
       if @title.nil?
-        if XPath.first(channel_node, "title/@type").to_s == "xhtml" ||
-            XPath.first(channel_node, "title/@mode").to_s == "xhtml"
+        repair_entities = false
+        if XPath.first(channel_node, "title/@type").to_s == "xhtml" ||
+            XPath.first(channel_node, "title/@mode").to_s == "xhtml" ||
+            XPath.first(channel_node, "title/@type").to_s == "xml" ||
+            XPath.first(channel_node, "title/@mode").to_s == "xml" ||
+            XPath.first(channel_node, "title/@type").to_s ==
+              "application/xhtml+xml"
           @title = XPath.first(channel_node, "title").inner_xml
         elsif XPath.first(channel_node, "title/@type").to_s == "escaped" ||
             XPath.first(channel_node, "title/@mode").to_s == "escaped"
-          @title = CGI.unescapeHTML(
+          @title = FeedTools.unescape_entities(
             XPath.first(channel_node, "title/text()").to_s)
         else
-          @title = CGI.unescapeHTML(
-            XPath.first(channel_node, "title/text()").to_s)
+          title_node = XPath.first(channel_node, "title")
+          @title = title_node.inner_xml
+          repair_entities = true
         end
         unless @title.nil?
-          @title = CGI.unescapeHTML(FeedTools.sanitize_html(@title, :strip))
-        end
-        if @title != "" && !(@title.nil?)
-          @title = FeedTools.strip_html(@title).strip
+          @title = FeedTools.sanitize_html(@title, :strip)
+          @title = FeedTools.unescape_entities(@title) if repair_entities
+          @title = FeedTools.tidy_html(@title)
         end
         @title.gsub!(/\n/, " ")
+        @title.strip!
         @title = nil if @title == ""
         self.cache_object.title = @title unless self.cache_object.nil?
       end
@@ -981,57 +1125,57 @@ module FeedTools
     # Returns the feed description
     def description
       if @description.nil?
-        # get the feed description from the xml document
-        @description = XPath.first(channel_node, "description/text()").to_s
-        if @description != ""
-          if XPath.first(channel_node, "description/@encoding").to_s != ""
-            @description = "[Embedded data objects are not supported.]"
+        repair_entities = false
+        description_node = XPath.first(channel_node, "description")
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "tagline")
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "subtitle")
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "summary")
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "abstract")
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "info")
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "content:encoded")
+          @bozo = true unless description_node.nil?
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "content")
+          @bozo = true unless description_node.nil?
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "xhtml:body")
+          @bozo = true unless description_node.nil?
+        end
+        if description_node.nil?
+          description_node = XPath.first(channel_node, "body")
+          @bozo = true unless description_node.nil?
+        end
+        unless description_node.nil?
+          if XPath.first(description_node, "@encoding").to_s != ""
+            @description =
+              "[Embedded data objects are not currently supported.]"
+          elsif XPath.first(description_node, "@type").to_s == "xhtml" ||
+              XPath.first(description_node, "@mode").to_s == "xhtml" ||
+              XPath.first(description_node, "@type").to_s == "xml" ||
+              XPath.first(description_node, "@mode").to_s == "xml" ||
+              XPath.first(description_node, "@type").to_s ==
+                "application/xhtml+xml"
+            @description = description_node.inner_xml
+          elsif XPath.first(description_node, "@type").to_s == "escaped" ||
+              XPath.first(description_node, "@mode").to_s == "escaped"
+            @description = FeedTools.unescape_entities(
+              description_node.inner_xml)
           else
-            @description = CGI.unescapeHTML(description)
-          end
-        end
-        if @description == ""
-          @description = XPath.first(channel_node, "subtitle/text()").to_s
-          if @description != "" &&
-              XPath.first(channel_node, "subtitle/@mode").to_s == "escaped"
-            @description = CGI.unescapeHTML(description)
-          end
-        end
-        if @description == ""
-          @description = XPath.first(channel_node, "tagline/text()").to_s
-          if @description != "" &&
-              XPath.first(channel_node, "tagline/@mode").to_s == "escaped"
-            @description = CGI.unescapeHTML(description)
-          end
-        end
-        if @description == "" && XPath.first(channel_node, "tagline") == nil
-          @description = XPath.first(channel_node, "info/text()").to_s
-          if @description != "" &&
-              XPath.first(channel_node, "info/@mode").to_s == "escaped"
-            @description = CGI.unescapeHTML(description)
-          end
-        end
-        if @description == ""
-          @description = CGI.unescapeHTML(
-            XPath.first(channel_node, "abstract/text()").to_s)
-        end
-        if @description == ""
-          @description = CGI.unescapeHTML(
-            XPath.first(channel_node, "summary/text()").to_s)
-        end
-        if @description == ""
-          # I don't think this is valid for anyone to do, but this is probably
-          # what they meant if they do it.
-          @description = CGI.unescapeHTML(
-            XPath.first(channel_node, "content:encoded/text()").to_s)
-          if @description != ""
-            @bozo = true
-          end
-        end
-        if @description == ""
-          begin
-            @description = XPath.first(channel_node, "description").inner_xml
-          rescue
+            @description = description_node.inner_xml
+            repair_entities = true
           end
         end
         if @description == ""
@@ -1043,13 +1187,12 @@ module FeedTools
           @description = "" if @description.nil?
         end
-        @description =
-          FeedTools.sanitize_html(@description) unless @description.nil?
-        # If it started with a bunch of divs, hack them right off.  We can put
-        # them back later if they're needed.
-        @description.gsub!(/^(<div[^>]*>)*/, "")
-        @description.gsub!(/(<\/div>)*$/, "")
+        unless @description.nil?
+          @description = FeedTools.sanitize_html(@description, :strip)
+          @description = FeedTools.unescape_entities(@description) if repair_entities
+          @description = FeedTools.tidy_html(@description)
+        end
         @description.gsub!(/\n/, " ") if @description.size < 80
         @description = @description.strip unless @description.nil?
         @description = nil if @description == ""
@@ -1065,7 +1208,7 @@ module FeedTools
     # Returns the contents of the itunes:summary element
     def itunes_summary
       if @itunes_summary.nil?
-        @itunes_summary = CGI.unescapeHTML(XPath.first(root_node,
+        @itunes_summary = FeedTools.unescape_entities(XPath.first(root_node,
           "itunes:summary/text()").to_s)
         if @itunes_summary == ""
           @itunes_summary = nil
@@ -1084,7 +1227,7 @@ module FeedTools
     # Returns the contents of the itunes:subtitle element
     def itunes_subtitle
       if @itunes_subtitle.nil?
-        @itunes_subtitle = CGI.unescapeHTML(XPath.first(root_node,
+        @itunes_subtitle = FeedTools.unescape_entities(XPath.first(root_node,
           "itunes:subtitle/text()").to_s)
         if @itunes_subtitle == ""
           @itunes_subtitle = nil
@@ -1211,6 +1354,197 @@ module FeedTools
       return @icon_link
     end
+    # Returns the feed author
+    def author
+      if @author.nil?
+        @author = FeedTools::Feed::Author.new
+        # Set the author name
+        @author.name = FeedTools.unescape_entities(
+          XPath.first(channel_node, "author/name/text()").to_s)
+        @author.raw = FeedTools.unescape_entities(
+          XPath.first(channel_node, "author/text()").to_s)
+        if @author.raw == ""
+          @author.raw = FeedTools.unescape_entities(
+            XPath.first(channel_node, "dc:creator/text()").to_s)
+        end
+        if @author.raw == ""
+          @author.raw = FeedTools.unescape_entities(
+            XPath.first(channel_node, "dc:author/text()").to_s)
+        end
+        unless @author.raw == ""
+          raw_scan = @author.raw.scan(
+            /(.*)\((\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b)\)/i)
+          if raw_scan.nil? || raw_scan.size == 0
+            raw_scan = @author.raw.scan(
+              /(\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b)\s*\((.*)\)/i)
+            author_raw_pair = raw_scan.first.reverse unless raw_scan.size == 0
+          else
+            author_raw_pair = raw_scan.first
+          end
+          if raw_scan.nil? || raw_scan.size == 0
+            email_scan = @author.raw.scan(
+              /\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b/i)
+            if email_scan != nil && email_scan.size > 0
+              @author.email = email_scan.first.strip
+            end
+          end
+          unless author_raw_pair.nil? || author_raw_pair.size == 0
+            @author.name = author_raw_pair.first.strip
+            @author.email = author_raw_pair.last.strip
+          else
+            unless @author.raw.include?("@")
+              # We can be reasonably sure we are looking at something
+              # that the creator didn't intend to contain an email address if
+              # it got through the preceeding regexes and it doesn't
+              # contain the tell-tale '@' symbol.
+              @author.name = @author.raw
+            end
+          end
+        end
+        @author.name = nil if @author.name == ""
+        @author.raw = nil if @author.raw == ""
+        # Set the author email
+        if @author.email == ""
+          @author.email = FeedTools.unescape_entities(
+            XPath.first(channel_node, "author/email/text()").to_s)
+        end
+        @author.email = nil if @author.email == ""
+        # Set the author url
+        @author.url = FeedTools.unescape_entities(
+          XPath.first(channel_node, "author/url/text()").to_s)
+        @author.url = nil if @author.url == ""
+        # Fallback on the itunes module if we didn't find an author name
+        begin
+          @author.name = self.itunes_author if @author.name.nil?
+        rescue
+          @author.name = nil
+        end
+      end
+      return @author
+    end
+    # Sets the feed author
+    def author=(new_author)
+      if new_author.respond_to?(:name) &&
+          new_author.respond_to?(:email) &&
+          new_author.respond_to?(:url)
+        # It's a complete author object, just set it.
+        @author = new_author
+      else
+        # We're not looking at an author object, this is probably a string,
+        # default to setting the author's name.
+        if @author.nil?
+          @author = FeedTools::Feed::Author.new
+        end
+        @author.name = new_author
+      end
+    end
+    # Returns the feed publisher
+    def publisher
+      if @publisher.nil?
+        @publisher = FeedTools::Feed::Author.new
+        # Set the author name
+        @publisher.raw = FeedTools.unescape_entities(
+          XPath.first(channel_node, "dc:publisher/text()").to_s)
+        unless @publisher.raw == ""
+          raw_scan = @publisher.raw.scan(
+            /(.*)\((\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b)\)/i)
+          if raw_scan.nil? || raw_scan.size == 0
+            raw_scan = @publisher.raw.scan(
+              /(\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b)\s*\((.*)\)/i)
+            unless raw_scan.size == 0
+              publisher_raw_pair = raw_scan.first.reverse
+            end
+          else
+            publisher_raw_pair = raw_scan.first
+          end
+          if raw_scan.nil? || raw_scan.size == 0
+            email_scan = @publisher.raw.scan(
+              /\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b/i)
+            if email_scan != nil && email_scan.size > 0
+              @publisher.email = email_scan.first.strip
+            end
+          end
+          unless publisher_raw_pair.nil? || publisher_raw_pair.size == 0
+            @publisher.name = publisher_raw_pair.first.strip
+            @publisher.email = publisher_raw_pair.last.strip
+          else
+            unless @publisher.raw.include?("@")
+              # We can be reasonably sure we are looking at something
+              # that the creator didn't intend to contain an email address if
+              # it got through the preceeding regexes and it doesn't
+              # contain the tell-tale '@' symbol.
+              @publisher.name = @publisher.raw
+            end
+          end
+        end
+        @publisher.name = nil if @publisher.name == ""
+        @publisher.raw = nil if @publisher.raw == ""
+        @publisher.email = nil if @publisher.email == ""
+        @publisher.url = nil if @publisher.url == ""
+      end
+      return @publisher
+    end
+    # Sets the feed publisher
+    def publisher=(new_publisher)
+      if new_publisher.respond_to?(:name) &&
+          new_publisher.respond_to?(:email) &&
+          new_publisher.respond_to?(:url)
+        # It's a complete Author object, just set it.
+        @publisher = new_publisher
+      else
+        # We're not looking at an Author object, this is probably a string,
+        # default to setting the publisher's name.
+        if @publisher.nil?
+          @publisher = FeedTools::Feed::Author.new
+        end
+        @publisher.name = new_publisher
+      end
+    end
+    # Returns the contents of the itunes:author element
+    #
+    # Returns any incorrectly placed channel-level itunes:author
+    # elements.  They're actually amazingly common.  People don't read specs.
+    # There is no setter for this, since this is a "bozo" attribute.
+    def itunes_author
+      if @itunes_author.nil?
+        @itunes_author = FeedTools.unescape_entities(XPath.first(channel_node,
+          "itunes:author/text()").to_s)
+        @itunes_author = nil if @itunes_author == ""
+        @bozo = true unless @itunes_author.nil?
+      end
+      return @itunes_author
+    end
+    # Returns the feed's copyright information
+    def copyright
+      if @copyright.nil?
+        @copyright = XPath.first(channel_node, "copyright/text()").to_s
+        if @copyright == ""
+          @copyright = XPath.first(channel_node, "dc:rights/text()").to_s
+        end
+        @copyright = FeedTools.sanitize_html(@copyright, :strip)
+        @copyright = nil if @copyright == ""
+      end
+      return @copyright
+    end
+    # Sets the feed's copyright information
+    def copyright=(new_copyright)
+      @copyright = new_copyright
+    end
     # Returns the number of seconds before the feed should expire
     def time_to_live
       if @time_to_live.nil?
@@ -1275,6 +1609,33 @@ module FeedTools
       @time_to_live = 1 if @time_to_live < 1
     end
+    # Returns the feed's cloud
+    def cloud
+      if @cloud.nil?
+        @cloud = FeedTools::Feed::Cloud.new
+        @cloud.domain = XPath.first(channel_node, "cloud/@domain").to_s
+        @cloud.port = XPath.first(channel_node, "cloud/@port").to_s
+        @cloud.path = XPath.first(channel_node, "cloud/@path").to_s
+        @cloud.register_procedure =
+          XPath.first(channel_node, "cloud/@registerProcedure").to_s
+        @cloud.protocol =
+          XPath.first(channel_node, "cloud/@protocol").to_s.downcase
+        @cloud.domain = nil if @cloud.domain == ""
+        @cloud.port = nil if @cloud.port == ""
+        @cloud.port = @cloud.port.to_i unless @cloud.port.nil?
+        @cloud.port = nil if @cloud.port == 0
+        @cloud.path = nil if @cloud.path == ""
+        @cloud.register_procedure = nil if @cloud.register_procedure == ""
+        @cloud.protocol = nil if @cloud.protocol == ""
+      end
+      return @cloud
+    end
+    # Sets the feed's cloud
+    def cloud=(new_cloud)
+      @cloud = new_cloud
+    end
     # Returns the feed generator
     def generator
       if @generator.nil?
@@ -1454,7 +1815,8 @@ module FeedTools
     end
     # Generates xml based on the content of the feed
-    def build_xml(feed_type="rss", version=0.0, xml_builder=Builder::XmlMarkup.new(:indent => 2))
+    def build_xml(feed_type=(self.feed_type or "rss"), version=0.0,
+        xml_builder=Builder::XmlMarkup.new(:indent => 2))
       if feed_type == "rss" && version == 0.0
         version = 1.0
       elsif feed_type == "atom" && version == 0.0
@@ -1801,22 +2163,47 @@ module FeedTools
       return @root_node
     end
+    # Returns the feed items's unique id
+    def id
+      if @id.nil?
+        @id = XPath.first(root_node, "id/text()").to_s
+        if @id == ""
+          @id = XPath.first(root_node, "guid/text()").to_s
+        end
+        @id = nil if @id == ""
+      end
+      return @id
+    end
+    # Sets the feed item's unique id
+    def id=(new_id)
+      @id = new_id
+    end
     # Returns the feed item title
     def title
       if @title.nil?
+        repair_entities = false
         if XPath.first(root_node, "title/@type").to_s == "xhtml" ||
-            XPath.first(root_node, "title/@mode").to_s == "xhtml"
+            XPath.first(root_node, "title/@mode").to_s == "xhtml" ||
+            XPath.first(root_node, "title/@type").to_s == "xml" ||
+            XPath.first(root_node, "title/@mode").to_s == "xml" ||
+            XPath.first(root_node, "title/@type").to_s ==
+              "application/xhtml+xml"
           @title = XPath.first(root_node, "title").inner_xml
         elsif XPath.first(root_node, "title/@type").to_s == "escaped" ||
             XPath.first(root_node, "title/@mode").to_s == "escaped"
-          @title = CGI.unescapeHTML(
+          @title = FeedTools.unescape_entities(
             XPath.first(root_node, "title/text()").to_s)
         else
-          @title = CGI.unescapeHTML(
-            XPath.first(root_node, "title/text()").to_s)
+          title_node = XPath.first(root_node, "title")
+          @title = title_node.inner_xml
+          repair_entities = true
         end
         unless @title.nil?
-          @title = CGI.unescapeHTML(FeedTools.sanitize_html(@title, :strip))
+          @title = FeedTools.sanitize_html(@title, :strip)
+          @title = FeedTools.unescape_entities(@title) if repair_entities
+          @title = FeedTools.tidy_html(@title)
         end
         if @title != ""
           # Some blogging tools include the number of comments in a post
@@ -1826,10 +2213,10 @@ module FeedTools
           #
           # If for some incredibly wierd reason you need the actual
           # unstripped title, just use find_node("title/text()").to_s
-          @title = FeedTools.strip_html(
-            @title.strip.gsub(/\[\d*\]$/, "")).strip
-          @title.gsub!(/\n/, " ")
+          @title = @title.strip.gsub(/\[\d*\]$/, "").strip
         end
+        @title.gsub!(/\n/, " ")
+        @title.strip!
         @title = nil if @title == ""
       end
       return @title
@@ -1843,53 +2230,54 @@ module FeedTools
     # Returns the feed item description
     def description
       if @description.nil?
-        # get the item content
-        @description = ""
-        body_node = XPath.first(root_node, "xhtml:body")
-        if body_node == nil
-          body_node = XPath.first(root_node, "body")
-        end
-        if body_node != nil
-          @description = body_node.inner_xml
-        end
-        if @description == ""
-          @description =
-            CGI.unescapeHTML(XPath.first(root_node, "content:encoded/text()").to_s)
-        end
-        if @description == ""
-          begin
-            @description = XPath.first(root_node, "description").cdatas.first.to_s
-          rescue
-            @description = ""
-          end
-          if @description == ""
-            @description = XPath.first(root_node, "description/text()").to_s
-          end
-          if @description != ""
-            if XPath.first(root_node, "description/@encoding").to_s != ""
-              # Not supported... yet.
-              @description = "[Embedded data objects are not supported.]"
-            else
-              @description = CGI.unescapeHTML(@description)
-            end
-          end
-        end
-        if @description == ""
-          @description = XPath.first(root_node, "content/text()").to_s
-          if @description != "" &&
-              (XPath.first(root_node, "content/@mode").to_s == "escaped" ||
-              XPath.first(root_node, "content/@type").to_s == "escaped")
-            @description = CGI.unescapeHTML(@description)
-          end
-          if XPath.first(root_node, "content/@mode").to_s == "xhtml" ||
-              XPath.first(root_node, "content/@type").to_s == "xhtml"
-            @description = XPath.first(root_node, "content").inner_xml
-          end
-        end
-        if @description == ""
-          begin
-            @description = XPath.first(root_node, "description").inner_xml
-          rescue
+        repair_entities = false
+        description_node = XPath.first(root_node, "description")
+        if description_node.nil?
+          description_node = XPath.first(root_node, "xhtml:body")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "body")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "tagline")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "subtitle")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "summary")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "abstract")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "content:encoded")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "content")
+        end
+        if description_node.nil?
+          description_node = XPath.first(root_node, "info")
+          @bozo = true unless description_node.nil?
+        end
+        unless description_node.nil?
+          if XPath.first(description_node, "@encoding").to_s != ""
+            @description =
+              "[Embedded data objects are not currently supported.]"
+          elsif XPath.first(description_node, "@type").to_s == "xhtml" ||
+              XPath.first(description_node, "@mode").to_s == "xhtml" ||
+              XPath.first(description_node, "@type").to_s == "xml" ||
+              XPath.first(description_node, "@mode").to_s == "xml" ||
+              XPath.first(description_node, "@type").to_s ==
+                "application/xhtml+xml"
+            @description = description_node.inner_xml
+          elsif XPath.first(description_node, "@type").to_s == "escaped" ||
+              XPath.first(description_node, "@mode").to_s == "escaped"
+            @description = FeedTools.unescape_entities(
+              description_node.inner_xml)
+          else
+            @description = description_node.inner_xml
+            repair_entities = true
           end
         end
         if @description == ""
@@ -1900,20 +2288,13 @@ module FeedTools
           @description = self.itunes_subtitle
           @description = "" if @description.nil?
         end
-        if @description == ""
-          @description = self.media_text
-          @description = "" if @description.nil?
-        end
         unless @description.nil?
-          @description = FeedTools.sanitize_html(@description)
+          @description = FeedTools.sanitize_html(@description, :strip)
+          @description = FeedTools.unescape_entities(@description) if repair_entities
+          @description = FeedTools.tidy_html(@description)
         end
-        # If it started with a bunch of divs, hack them right off.  We can put
-        # them back later if they're needed.
-        @description.gsub!(/^(<div[^>]*>)*/, "")
-        @description.gsub!(/(<\/div>)*$/, "")
         @description.gsub!(/\n/, " ") if @description.size < 80
         @description = @description.strip unless @description.nil?
         @description = nil if @description == ""
@@ -1925,6 +2306,66 @@ module FeedTools
     def description=(new_description)
       @description = new_description
     end
+    # Returns the contents of the itunes:summary element
+    def itunes_summary
+      if @itunes_summary.nil?
+        @itunes_summary = FeedTools.unescape_entities(XPath.first(root_node,
+          "itunes:summary/text()").to_s)
+        if @itunes_summary == ""
+          @itunes_summary = nil
+        end
+        unless @itunes_summary.nil?
+          @itunes_summary = FeedTools.sanitize_html(@itunes_summary)
+        end
+      end
+      return @itunes_summary
+    end
+    # Sets the contents of the itunes:summary element
+    def itunes_summary=(new_itunes_summary)
+      @itunes_summary = new_itunes_summary
+    end
+    # Returns the contents of the itunes:subtitle element
+    def itunes_subtitle
+      if @itunes_subtitle.nil?
+        @itunes_subtitle = FeedTools.unescape_entities(XPath.first(root_node,
+          "itunes:subtitle/text()").to_s)
+        if @itunes_subtitle == ""
+          @itunes_subtitle = nil
+        end
+        unless @itunes_subtitle.nil?
+          @itunes_subtitle = FeedTools.sanitize_html(@itunes_subtitle)
+        end
+      end
+      return @itunes_subtitle
+    end
+    # Sets the contents of the itunes:subtitle element
+    def itunes_subtitle=(new_itunes_subtitle)
+      @itunes_subtitle = new_itunes_subtitle
+    end
+    # Returns the contents of the media:text element
+    def media_text
+      if @media_text.nil?
+        @media_text = FeedTools.unescape_entities(XPath.first(root_node,
+          "itunes:subtitle/text()").to_s)
+        if @media_text == ""
+          @media_text = nil
+        end
+        unless @media_text.nil?
+          @media_text = FeedTools.sanitize_html(@media_text)
+        end
+      end
+      return @media_text
+    end
+    # Sets the contents of the media:text element
+    def media_text=(new_media_text)
+      @media_text = new_media_text
+    end
     # Returns the feed item link
     def link
@@ -1948,7 +2389,7 @@ module FeedTools
           end
         end
         if @link != ""
-          @link = CGI.unescapeHTML(@link)
+          @link = FeedTools.unescape_entities(@link)
         end
         if @link != "" && (@link =~ /http:\/\//) != 0 && (@link =~ /https:\/\//) != 0
           if (feed.base[-1..-1] == "/" && @link[0..0] == "/")
@@ -2084,23 +2525,6 @@ module FeedTools
       @media_thumbnail_link = new_media_thumbnail_link
     end
-    # Returns the feed items's unique id
-    def id
-      if @id.nil?
-        @id = XPath.first(root_node, "id/text()").to_s
-        if @id == ""
-          @id = XPath.first(root_node, "guid/text()").to_s
-        end
-        @id = nil if @id == ""
-      end
-      return @id
-    end
-    # Sets the feed item's unique id
-    def id=(new_id)
-      @id = new_id
-    end
     # Returns all feed item enclosures
     def enclosures
       if @enclosures.nil?
@@ -2116,7 +2540,7 @@ module FeedTools
         # sometimes these also manage to show up in atom files.
         for enclosure_node in rss_enclosures
           enclosure = Enclosure.new
-          enclosure.url = CGI.unescapeHTML(enclosure_node.attributes["url"].to_s)
+          enclosure.url = FeedTools.unescape_entities(enclosure_node.attributes["url"].to_s)
           enclosure.type = enclosure_node.attributes["type"].to_s
           enclosure.file_size = enclosure_node.attributes["length"].to_i
           enclosure.credits = []
@@ -2127,7 +2551,7 @@ module FeedTools
         # Parse atom-type enclosures.  If there are repeats of the same enclosure object,
         # we merge the two together.
         for enclosure_node in atom_enclosures
-          enclosure_url = CGI.unescapeHTML(enclosure_node.attributes["href"].to_s)
+          enclosure_url = FeedTools.unescape_entities(enclosure_node.attributes["href"].to_s)
           enclosure = nil
           new_enclosure = false
           for existing_enclosure in @enclosures
@@ -2156,7 +2580,7 @@ module FeedTools
         parse_media_content = lambda do |media_content_nodes|
           affected_enclosures = []
           for enclosure_node in media_content_nodes
-            enclosure_url = CGI.unescapeHTML(enclosure_node.attributes["url"].to_s)
+            enclosure_url = FeedTools.unescape_entities(enclosure_node.attributes["url"].to_s)
             enclosure = nil
             new_enclosure = false
             for existing_enclosure in @enclosures
@@ -2182,9 +2606,9 @@ module FeedTools
               (enclosure_node.attributes["isDefault"].to_s.downcase == "true")
             if XPath.first(enclosure_node, "media:thumbnail/@url").to_s != ""
               enclosure.thumbnail = EnclosureThumbnail.new(
-                CGI.unescapeHTML(XPath.first(enclosure_node, "media:thumbnail/@url").to_s),
-                CGI.unescapeHTML(XPath.first(enclosure_node, "media:thumbnail/@height").to_s),
-                CGI.unescapeHTML(XPath.first(enclosure_node, "media:thumbnail/@width").to_s)
+                FeedTools.unescape_entities(XPath.first(enclosure_node, "media:thumbnail/@url").to_s),
+                FeedTools.unescape_entities(XPath.first(enclosure_node, "media:thumbnail/@height").to_s),
+                FeedTools.unescape_entities(XPath.first(enclosure_node, "media:thumbnail/@width").to_s)
               )
               if enclosure.thumbnail.height == ""
                 enclosure.thumbnail.height = nil
@@ -2196,9 +2620,9 @@ module FeedTools
             enclosure.categories = []
             for category in XPath.match(enclosure_node, "media:category")
               enclosure.categories << EnclosureCategory.new(
-                CGI.unescapeHTML(category.text),
-                CGI.unescapeHTML(category.attributes["scheme"].to_s),
-                CGI.unescapeHTML(category.attributes["label"].to_s)
+                FeedTools.unescape_entities(category.text),
+                FeedTools.unescape_entities(category.attributes["scheme"].to_s),
+                FeedTools.unescape_entities(category.attributes["label"].to_s)
               )
               if enclosure.categories.last.scheme == ""
                 enclosure.categories.last.scheme = nil
@@ -2209,16 +2633,16 @@ module FeedTools
             end
             if XPath.first(enclosure_node, "media:hash/text()").to_s != ""
               enclosure.hash = EnclosureHash.new(
-                FeedTools.sanitize_html(CGI.unescapeHTML(XPath.first(
+                FeedTools.sanitize_html(FeedTools.unescape_entities(XPath.first(
                   enclosure_node, "media:hash/text()").to_s), :strip),
                 "md5"
               )
             end
             if XPath.first(enclosure_node, "media:player/@url").to_s != ""
               enclosure.player = EnclosurePlayer.new(
-                CGI.unescapeHTML(XPath.first(enclosure_node, "media:player/@url").to_s),
-                CGI.unescapeHTML(XPath.first(enclosure_node, "media:player/@height").to_s),
-                CGI.unescapeHTML(XPath.first(enclosure_node, "media:player/@width").to_s)
+                FeedTools.unescape_entities(XPath.first(enclosure_node, "media:player/@url").to_s),
+                FeedTools.unescape_entities(XPath.first(enclosure_node, "media:player/@height").to_s),
+                FeedTools.unescape_entities(XPath.first(enclosure_node, "media:player/@width").to_s)
               )
               if enclosure.player.height == ""
                 enclosure.player.height = nil
@@ -2230,8 +2654,8 @@ module FeedTools
             enclosure.credits = []
             for credit in XPath.match(enclosure_node, "media:credit")
               enclosure.credits << EnclosureCredit.new(
-                CGI.unescapeHTML(CGI.unescapeHTML(credit.text)),
-                CGI.unescapeHTML(credit.attributes["role"].to_s.downcase)
+                FeedTools.unescape_entities(credit.text),
+                FeedTools.unescape_entities(credit.attributes["role"].to_s.downcase)
               )
               if enclosure.credits.last.role == ""
                 enclosure.credits.last.role = nil
@@ -2240,7 +2664,7 @@ module FeedTools
             enclosure.explicit = (XPath.first(enclosure_node,
               "media:adult/text()").to_s.downcase == "true")
             if XPath.first(enclosure_node, "media:text/text()").to_s != ""
-              enclosure.text = CGI.unescapeHTML(XPath.first(enclosure_node,
+              enclosure.text = FeedTools.unescape_entities(XPath.first(enclosure_node,
                 "media:text/text()").to_s)
             end
             affected_enclosures << enclosure
@@ -2271,11 +2695,11 @@ module FeedTools
             if enclosure.thumbnail.nil? &&
                 XPath.first(media_group, "media:thumbnail/@url").to_s != ""
               enclosure.thumbnail = EnclosureThumbnail.new(
-                CGI.unescapeHTML(
+                FeedTools.unescape_entities(
                   XPath.first(media_group, "media:thumbnail/@url").to_s),
-                CGI.unescapeHTML(
+                FeedTools.unescape_entities(
                   XPath.first(media_group, "media:thumbnail/@height").to_s),
-                CGI.unescapeHTML(
+                FeedTools.unescape_entities(
                   XPath.first(media_group, "media:thumbnail/@width").to_s)
               )
               if enclosure.thumbnail.height == ""
@@ -2289,9 +2713,9 @@ module FeedTools
               enclosure.categories = []
               for category in XPath.match(media_group, "media:category")
                 enclosure.categories << EnclosureCategory.new(
-                  CGI.unescapeHTML(category.text),
-                  CGI.unescapeHTML(category.attributes["scheme"].to_s),
-                  CGI.unescapeHTML(category.attributes["label"].to_s)
+                  FeedTools.unescape_entities(category.text),
+                  FeedTools.unescape_entities(category.attributes["scheme"].to_s),
+                  FeedTools.unescape_entities(category.attributes["label"].to_s)
                 )
                 if enclosure.categories.last.scheme == ""
                   enclosure.categories.last.scheme = nil
@@ -2304,16 +2728,16 @@ module FeedTools
             if enclosure.hash.nil? &&
                 XPath.first(media_group, "media:hash/text()").to_s != ""
               enclosure.hash = EnclosureHash.new(
-                CGI.unescapeHTML(XPath.first(media_group, "media:hash/text()").to_s),
+                FeedTools.unescape_entities(XPath.first(media_group, "media:hash/text()").to_s),
                 "md5"
               )
             end
             if enclosure.player.nil? &&
                 XPath.first(media_group, "media:player/@url").to_s != ""
               enclosure.player = EnclosurePlayer.new(
-                CGI.unescapeHTML(XPath.first(media_group, "media:player/@url").to_s),
-                CGI.unescapeHTML(XPath.first(media_group, "media:player/@height").to_s),
-                CGI.unescapeHTML(XPath.first(media_group, "media:player/@width").to_s)
+                FeedTools.unescape_entities(XPath.first(media_group, "media:player/@url").to_s),
+                FeedTools.unescape_entities(XPath.first(media_group, "media:player/@height").to_s),
+                FeedTools.unescape_entities(XPath.first(media_group, "media:player/@width").to_s)
               )
               if enclosure.player.height == ""
                 enclosure.player.height = nil
@@ -2326,8 +2750,8 @@ module FeedTools
               enclosure.credits = []
               for credit in XPath.match(media_group, "media:credit")
                 enclosure.credits << EnclosureCredit.new(
-                  CGI.unescapeHTML(CGI.unescapeHTML(credit.text)),
-                  CGI.unescapeHTML(credit.attributes["role"].to_s.downcase)
+                  FeedTools.unescape_entities(credit.text),
+                  FeedTools.unescape_entities(credit.attributes["role"].to_s.downcase)
                 )
                 if enclosure.credits.last.role == ""
                   enclosure.credits.last.role = nil
@@ -2340,7 +2764,7 @@ module FeedTools
             end
             if enclosure.text.nil? &&
                 XPath.first(media_group, "media:text/text()").to_s != ""
-              enclosure.text = FeedTools.sanitize_html(CGI.unescapeHTML(
+              enclosure.text = FeedTools.sanitize_html(FeedTools.unescape_entities(
                 XPath.first(media_group, "media:text/text()").to_s), :strip)
             end
           end
@@ -2373,9 +2797,9 @@ module FeedTools
               enclosure.categories = []
             end
             enclosure.categories << EnclosureCategory.new(
-              CGI.unescapeHTML(category_path),
-              CGI.unescapeHTML("http://www.apple.com/itunes/store/"),
-              CGI.unescapeHTML("iTunes Music Store Categories")
+              FeedTools.unescape_entities(category_path),
+              FeedTools.unescape_entities("http://www.apple.com/itunes/store/"),
+              FeedTools.unescape_entities("iTunes Music Store Categories")
             )
           end
         end
@@ -2464,136 +2888,140 @@ module FeedTools
     def enclosures=(new_enclosures)
       @enclosures = new_enclosures
     end
-    # Returns the feed item author
-    def author_name
-      # TODO: make this not suck, actually ensure we're looking at a name
-      # and not an email address.
-      # Also, factor in itunes module.
-      # =================================================================
-      if @author_name.nil?
-        @author_name = CGI.unescapeHTML(XPath.first(root_node, "author/name/text()").to_s)
-        if @author_name == ""
-          @author_name = CGI.unescapeHTML(XPath.first(root_node, "dc:creator/text()").to_s)
-        end
-        if @author_name == ""
-          @author_name = CGI.unescapeHTML(XPath.first(root_node, "author/text()").to_s)
-        end
-      end
-      return @author_name
-    end
-    # Sets the feed item author
-    def author_name=(new_author_name)
-      @author_name = new_author_name
-    end
-    # Returns the contents of the itunes:summary element
-    def itunes_summary
-      if @itunes_summary.nil?
-        @itunes_summary = CGI.unescapeHTML(XPath.first(root_node,
-          "itunes:summary/text()").to_s)
-        if @itunes_summary == ""
-          @itunes_summary = nil
-        end
-        unless @itunes_summary.nil?
-          @itunes_summary = FeedTools.sanitize_html(@itunes_summary)
+    # Returns the feed item author
+    def author
+      if @author.nil?
+        @author = FeedTools::Feed::Author.new
+        # Set the author name
+        @author.name = FeedTools.unescape_entities(
+          XPath.first(root_node, "author/name/text()").to_s)
+        @author.raw = FeedTools.unescape_entities(
+          XPath.first(root_node, "author/text()").to_s)
+        if @author.raw == ""
+          @author.raw = FeedTools.unescape_entities(
+            XPath.first(root_node, "dc:creator/text()").to_s)
+        end
+        if @author.raw == ""
+          @author.raw = FeedTools.unescape_entities(
+            XPath.first(root_node, "dc:author/text()").to_s)
+        end
+        unless @author.raw == ""
+          raw_scan = @author.raw.scan(
+            /(.*)\((\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b)\)/i)
+          if raw_scan.nil? || raw_scan.size == 0
+            raw_scan = @author.raw.scan(
+              /(\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b)\s*\((.*)\)/i)
+            author_raw_pair = raw_scan.first.reverse unless raw_scan.size == 0
+          else
+            author_raw_pair = raw_scan.first
+          end
+          if raw_scan.nil? || raw_scan.size == 0
+            email_scan = @author.raw.scan(
+              /\b[A-Z0-9._%-\+]+@[A-Z0-9._%-]+\.[A-Z]{2,4}\b/i)
+            if email_scan != nil && email_scan.size > 0
+              @author.email = email_scan.first.strip
+            end
+          end
+          unless author_raw_pair.nil? || author_raw_pair.size == 0
+            @author.name = author_raw_pair.first.strip
+            @author.email = author_raw_pair.last.strip
+          else
+            unless @author.raw.include?("@")
+              # We can be reasonably sure we are looking at something
+              # that the creator didn't intend to contain an email address if
+              # it got through the preceeding regexes and it doesn't
+              # contain the tell-tale '@' symbol.
+              @author.name = @author.raw
+            end
+          end
         end
-      end
-      return @itunes_summary
-    end
-    # Sets the contents of the itunes:summary element
-    def itunes_summary=(new_itunes_summary)
-      @itunes_summary = new_itunes_summary
-    end
+        @author.name = nil if @author.name == ""
+        @author.raw = nil if @author.raw == ""
-    # Returns the contents of the itunes:subtitle element
-    def itunes_subtitle
-      if @itunes_subtitle.nil?
-        @itunes_subtitle = CGI.unescapeHTML(XPath.first(root_node,
-          "itunes:subtitle/text()").to_s)
-        if @itunes_subtitle == ""
-          @itunes_subtitle = nil
+        # Set the author email
+        if @author.email == ""
+          @author.email = FeedTools.unescape_entities(
+            XPath.first(root_node, "author/email/text()").to_s)
         end
-        unless @itunes_subtitle.nil?
-          @itunes_subtitle = FeedTools.sanitize_html(@itunes_subtitle)
-        end
-      end
-      return @itunes_subtitle
-    end
+        @author.email = nil if @author.email == ""
-    # Sets the contents of the itunes:subtitle element
-    def itunes_subtitle=(new_itunes_subtitle)
-      @itunes_subtitle = new_itunes_subtitle
-    end
+        # Set the author url
+        @author.url = FeedTools.unescape_entities(
+          XPath.first(root_node, "author/url/text()").to_s)
+        @author.url = nil if @author.url == ""
-    # Returns the contents of the media:text element
-    def media_text
-      if @media_text.nil?
-        @media_text = CGI.unescapeHTML(XPath.first(root_node,
-          "itunes:subtitle/text()").to_s)
-        if @media_text == ""
-          @media_text = nil
-        end
-        unless @media_text.nil?
-          @media_text = FeedTools.sanitize_html(@media_text)
+        # Fallback on the itunes module if we didn't find an author name
+        begin
+          @author.name = self.itunes_author if @author.name.nil?
+        rescue
+          @author.name = nil
         end
       end
-      return @media_text
+      return @author
     end
-    # Sets the contents of the media:text element
-    def media_text=(new_media_text)
-      @media_text = new_media_text
+    # Sets the feed item author
+    def author=(new_author)
+      if new_author.respond_to?(:name) &&
+          new_author.respond_to?(:email) &&
+          new_author.respond_to?(:url)
+        # It's a complete author object, just set it.
+        @author = new_author
+      else
+        # We're not looking at an author object, this is probably a string,
+        # default to setting the author's name.
+        if @author.nil?
+          @author = FeedTools::Feed::Author.new
+        end
+        @author.name = new_author
+      end
     end
     # Returns the contents of the itunes:author element
     #
     # This inherits from any incorrectly placed channel-level itunes:author
-    # elements.  They're actually amazingly commong.  People don't read specs.
+    # elements.  They're actually amazingly common.  People don't read specs.
     def itunes_author
       if @itunes_author.nil?
-        @itunes_author = CGI.unescapeHTML(XPath.first(root_node,
+        @itunes_author = FeedTools.unescape_entities(XPath.first(root_node,
           "itunes:author/text()").to_s)
-        if @itunes_author == ""
-          @itunes_author = CGI.unescapeHTML(XPath.first(feed.channel_node,
-            "itunes:author/text()").to_s)
-        end
-        if @itunes_author == ""
-          @itunes_author = nil
-        end
+        @itunes_author = feed.itunes_author if @itunes_author == ""
+        @itunes_author = nil if @itunes_author == ""
       end
       return @itunes_author
     end
     # Sets the contents of the itunes:author element
     def itunes_author=(new_itunes_author)
       @itunes_author = new_itunes_author
-    end
+    end
     # Returns the number of seconds that the associated media runs for
-    def duration
-      if @duration.nil?
-        itunes_duration = CGI.unescapeHTML(XPath.first(root_node,
+    def itunes_duration
+      if @itunes_duration.nil?
+        raw_duration = FeedTools.unescape_entities(XPath.first(root_node,
           "itunes:duration/text()").to_s)
-        if itunes_duration != ""
-          hms = itunes_duration.split(":").map { |x| x.to_i }
+        if raw_duration != ""
+          hms = raw_duration.split(":").map { |x| x.to_i }
           if hms.size == 3
-            @duration = hms[0].hour + hms[1].minute + hms[2]
+            @itunes_duration = hms[0].hour + hms[1].minute + hms[2]
           elsif hms.size == 2
-            @duration = hms[0].minute + hms[1]
+            @itunes_duration = hms[0].minute + hms[1]
           elsif hms.size == 1
-            @duration = hms[0]
+            @itunes_duration = hms[0]
           end
         end
       end
-      return @duration
+      return @itunes_duration
     end
     # Sets the number of seconds that the associate media runs for
-    def duration=(new_duration)
-      @duration = new_duration
+    def itunes_duration=(new_itunes_duration)
+      @itunes_duration = new_itunes_duration
     end
     # Sets the itunes:summary
@@ -2722,7 +3150,8 @@ module FeedTools
     end
     # Generates xml based on the content of the feed item
-    def build_xml(feed_type="rss", version=0.0, xml_builder=Builder::XmlMarkup.new(:indent => 2))
+    def build_xml(feed_type=(self.feed.feed_type or "rss"), version=0.0,
+        xml_builder=Builder::XmlMarkup.new(:indent => 2))
       if feed_type == "rss" && (version == 0.9 || version == 1.0 || version == 1.1)
         # RDF-based rss format
         if link.nil?
@@ -2831,9 +3260,9 @@ module FeedTools
   end
 end
-module REXML #:nodoc:
-  class Element #:nodoc:
-    def inner_xml #:nodoc:
+module REXML # :nodoc:
+  class Element # :nodoc:
+    def inner_xml # :nodoc:
       result = ""
       self.each_child do |child|
         result << child.to_s
@@ -2848,4 +3277,4 @@ begin
     FeedTools.feed_cache.initialize_cache
   end
 rescue
-end
+end