nokogiri 1.0.0 → 1.6.8.1

data/lib/nokogiri/xml/node_set.rb

@@ -1,24 +1,37 @@
 module Nokogiri
   module XML
+    ####
+    # A NodeSet contains a list of Nokogiri::XML::Node objects.  Typically
+    # a NodeSet is return as a result of searching a Document via
+    # Nokogiri::XML::Searchable#css or Nokogiri::XML::Searchable#xpath
     class NodeSet
+      include Nokogiri::XML::Searchable
       include Enumerable
 
+      # The Document this NodeSet is associated with
       attr_accessor :document
 
-      def initialize
+      # Create a NodeSet with +document+ defaulting to +list+
+      def initialize document, list = []
+        @document = document
+        document.decorate(self)
+        list.each { |x| self << x }
         yield self if block_given?
       end
 
       ###
       # Get the first element of the NodeSet.
-      def first
-        self[0]
+      def first n = nil
+        return self[0] unless n
+        list = []
+        n.times { |i| list << self[i] }
+        list
       end
 
       ###
       # Get the last element of the NodeSet.
       def last
-        self[length - 1]
+        self[-1]
       end
 
       ###
@@ -27,6 +40,13 @@ module Nokogiri
         length == 0
       end
 
+      ###
+      # Returns the index of the first node in self that is == to +node+. Returns nil if no match is found.
+      def index(node)
+        each_with_index { |member, j| return j if member == node }
+        nil
+      end
+
       ###
       # Insert +datum+ before the first Node in this NodeSet
       def before datum
@@ -39,68 +59,98 @@ module Nokogiri
         last.after datum
       end
 
+      alias :<< :push
+      alias :remove :unlink
+
       ###
-      # Append +node+ to the NodeSet.
-      def << node
-        push(node)
+      # call-seq: css *rules, [namespace-bindings, custom-pseudo-class]
+      #
+      # Search this node set for CSS +rules+. +rules+ must be one or more CSS
+      # selectors. For example:
+      #
+      # For more information see Nokogiri::XML::Searchable#css
+      def css *args
+        rules, handler, ns, _ = extract_params(args)
+
+        inject(NodeSet.new(document)) do |set, node|
+          set += css_internal node, rules, handler, ns
+        end
       end
 
       ###
-      # Unlink this NodeSet and all Node objects it contains from their
-      # current context.
-      def unlink
-        each { |node| node.unlink }
-        self.document = nil
-        self
+      # call-seq: xpath *paths, [namespace-bindings, variable-bindings, custom-handler-class]
+      #
+      # Search this node set for XPath +paths+. +paths+ must be one or more XPath
+      # queries.
+      #
+      # For more information see Nokogiri::XML::Searchable#xpath
+      def xpath *args
+        paths, handler, ns, binds = extract_params(args)
+
+        inject(NodeSet.new(document)) do |set, node|
+          set += node.xpath(*(paths + [ns, handler, binds].compact))
+        end
       end
-      alias :remove :unlink
 
       ###
-      # Search this document for +paths+
-      def search *paths
-        sub_set = NodeSet.new
-        document.decorate(sub_set)
-        each do |node|
-          node.search(*paths).each do |sub_node|
-            sub_set << sub_node
-          end
+      # Search this NodeSet's nodes' immediate children using CSS selector +selector+
+      def > selector
+        ns = document.root.namespaces
+        xpath CSS.xpath_for(selector, :prefix => "./", :ns => ns).first
+      end
+
+      ###
+      # call-seq: search *paths, [namespace-bindings, xpath-variable-bindings, custom-handler-class]
+      #
+      # Search this object for +paths+, and return only the first
+      # result. +paths+ must be one or more XPath or CSS queries.
+      #
+      # See Searchable#search for more information.
+      #
+      # Or, if passed an integer, index into the NodeSet:
+      #
+      #   node_set.at(3) # same as node_set[3]
+      #
+      def at *args
+        if args.length == 1 && args.first.is_a?(Numeric)
+          return self[args.first]
         end
-        sub_set.document = document
-        sub_set
+
+        super(*args)
       end
-      alias :/ :search
-      alias :xpath :search
-      alias :css :search
+      alias :% :at
 
       ###
-      # If path is a string, search this document for +path+ returning the
-      # first Node.  Otherwise, index in to the array with +path+.
-      def at path, ns = {}
-        return self[path] if path.is_a?(Numeric)
-        search(path, ns).first
+      # Filter this list for nodes that match +expr+
+      def filter expr
+        find_all { |node| node.matches?(expr) }
       end
 
       ###
       # Append the class attribute +name+ to all Node objects in the NodeSet.
       def add_class name
         each do |el|
-          next unless el.respond_to? :get_attribute
-          classes = el.get_attribute('class').to_s.split(" ")
-          el.set_attribute('class', classes.push(name).uniq.join(" "))
+          classes = el['class'].to_s.split(/\s+/)
+          el['class'] = classes.push(name).uniq.join " "
         end
         self
       end
 
       ###
       # Remove the class attribute +name+ from all Node objects in the NodeSet.
+      # If +name+ is nil, remove the class attribute from all Nodes in the
+      # NodeSet.
       def remove_class name = nil
         each do |el|
-          next unless el.respond_to? :get_attribute
           if name
-            classes = el.get_attribute('class').to_s.split(" ")
-            el.set_attribute('class', (classes - [name]).uniq.join(" "))
+            classes = el['class'].to_s.split(/\s+/)
+            if classes.empty?
+              el.delete 'class'
+            else
+              el['class'] = (classes - [name]).uniq.join " "
+            end
           else
-            el.remove_attribute("class")
+            el.delete "class"
           end
         end
         self
@@ -110,73 +160,151 @@ module Nokogiri
       # Set the attribute +key+ to +value+ or the return value of +blk+
       # on all Node objects in the NodeSet.
       def attr key, value = nil, &blk
-        if value or blk
-          each do |el|
-            el.set_attribute(key, value || blk[el])
-          end
-          return self      
-        end    
-        if key.is_a? Hash
-          key.each { |k,v| self.attr(k,v) }
-          return self
-        else
-          return self[0].get_attribute(key)
+        unless Hash === key || key && (value || blk)
+          return first.attribute(key)
         end
+
+        hash = key.is_a?(Hash) ? key : { key => value }
+
+        hash.each { |k,v| each { |el| el[k] = v || blk[el] } }
+
+        self
       end
-      alias_method :set, :attr
+      alias :set :attr
+      alias :attribute :attr
 
       ###
       # Remove the attributed named +name+ from all Node objects in the NodeSet
       def remove_attr name
-        each do |el|
-          next unless el.respond_to? :remove_attribute
-          el.remove_attribute(name)
-        end
-        self      
+        each { |el| el.delete name }
+        self
       end
 
       ###
       # Iterate over each node, yielding  to +block+
       def each(&block)
-        x = 0
-        while x < length
+        0.upto(length - 1) do |x|
           yield self[x]
-          x += 1
         end
       end
 
       ###
       # Get the inner text of all contained Node objects
       def inner_text
-        collect{|j| j.inner_text}.join('')
+        collect(&:inner_text).join('')
       end
       alias :text :inner_text
 
+      ###
+      # Get the inner html of all contained Node objects
+      def inner_html *args
+        collect{|j| j.inner_html(*args) }.join('')
+      end
+
       ###
       # Wrap this NodeSet with +html+ or the results of the builder in +blk+
       def wrap(html, &blk)
         each do |j|
-          new_parent = Nokogiri.make(html, &blk)
-          j.replace(new_parent)
-          nest = new_parent
-          if nest.child
-            nest = nest.child until nest.child.nil?
-          end
-          j.parent = nest
+          new_parent = document.parse(html).first
+          j.add_next_sibling(new_parent)
+          new_parent.add_child(j)
         end
         self
       end
 
+      ###
+      # Convert this NodeSet to a string.
       def to_s
-        map { |x| x.to_s }.join
+        map(&:to_s).join
       end
 
-      def to_html
-        map { |x| x.to_html }.join('')
+      ###
+      # Convert this NodeSet to HTML
+      def to_html *args
+        if Nokogiri.jruby?
+          options = args.first.is_a?(Hash) ? args.shift : {}
+          if !options[:save_with]
+            options[:save_with] = Node::SaveOptions::NO_DECLARATION | Node::SaveOptions::NO_EMPTY_TAGS | Node::SaveOptions::AS_HTML
+          end
+          args.insert(0, options)
+        end
+        map { |x| x.to_html(*args) }.join
+      end
+
+      ###
+      # Convert this NodeSet to XHTML
+      def to_xhtml *args
+        map { |x| x.to_xhtml(*args) }.join
+      end
+
+      ###
+      # Convert this NodeSet to XML
+      def to_xml *args
+        map { |x| x.to_xml(*args) }.join
+      end
+
+      alias :size :length
+      alias :to_ary :to_a
+
+      ###
+      # Removes the last element from set and returns it, or +nil+ if
+      # the set is empty
+      def pop
+        return nil if length == 0
+        delete last
+      end
+
+      ###
+      # Returns the first element of the NodeSet and removes it.  Returns
+      # +nil+ if the set is empty.
+      def shift
+        return nil if length == 0
+        delete first
+      end
+
+      ###
+      # Equality -- Two NodeSets are equal if the contain the same number
+      # of elements and if each element is equal to the corresponding
+      # element in the other NodeSet
+      def == other
+        return false unless other.is_a?(Nokogiri::XML::NodeSet)
+        return false unless length == other.length
+        each_with_index do |node, i|
+          return false unless node == other[i]
+        end
+        true
+      end
+
+      ###
+      # Returns a new NodeSet containing all the children of all the nodes in
+      # the NodeSet
+      def children
+        inject(NodeSet.new(document)) { |set, node| set += node.children }
       end
 
-      def size
-        length
+      ###
+      # Returns a new NodeSet containing all the nodes in the NodeSet
+      # in reverse order
+      def reverse
+        node_set = NodeSet.new(document)
+        (length - 1).downto(0) do |x|
+          node_set.push self[x]
+        end
+        node_set
+      end
+
+      ###
+      # Return a nicely formated string representation
+      def inspect
+        "[#{map(&:inspect).join ', '}]"
+      end
+
+      alias :+ :|
+
+      private
+
+      def implied_xpath_contexts # :nodoc:
+        [".//", "self::"]
       end
     end
   end

data/lib/nokogiri/xml/parse_options.rb

@@ -0,0 +1,120 @@
+module Nokogiri
+  module XML
+    ###
+    # Parse options for passing to Nokogiri.XML or Nokogiri.HTML
+    #
+    # == Building combinations of parse options
+    # You can build your own combinations of these parse options by using any of the following methods:
+    # *Note*: All examples attempt to set the +RECOVER+ & +NOENT+ options. All examples use Ruby 2 optional parameter syntax.
+    # [Ruby's bitwise operators] You can use the Ruby bitwise operators to set various combinations.
+    #   <code>Nokogiri.XML('<content>Chapter 1</content', options: Nokogiri::XML::ParseOptions.new((1 << 0) | (1 << 1)))</code>
+    # [Method chaining] Every option has an equivalent method in lowercase. You can chain these methods together to set various combinations.
+    #   <code>Nokogiri.XML('<content>Chapter 1</content', options: Nokogiri::XML::ParseOptions.new.recover.noent)</code>
+    # [Using Ruby Blocks] You can also setup parse combinations in the block passed to Nokogiri.XML or Nokogiri.HTML
+    #   <code>Nokogiri.XML('<content>Chapter 1</content') {|config| config.recover.noent}</code>
+    #
+    # == Removing particular parse options
+    # You can also remove options from an instance of +ParseOptions+ dynamically.
+    # Every option has an equivalent <code>no{option}</code> method in lowercase. You can call these methods on an instance of +ParseOptions+ to remove the option.
+    # Note that this is not available for +STRICT+.
+    #
+    #   # Setting the RECOVER & NOENT options...
+    #   options = Nokogiri::XML::ParseOptions.new.recover.noent
+    #   # later...
+    #   options.norecover # Removes the Nokogiri::XML::ParseOptions::RECOVER option
+    #   options.nonoent # Removes the Nokogiri::XML::ParseOptions::NOENT option
+    #
+    class ParseOptions
+      # Strict parsing
+      STRICT      = 0
+      # Recover from errors
+      RECOVER     = 1 << 0
+      # Substitute entities
+      NOENT       = 1 << 1
+      # Load external subsets
+      DTDLOAD     = 1 << 2
+      # Default DTD attributes
+      DTDATTR     = 1 << 3
+      # validate with the DTD
+      DTDVALID    = 1 << 4
+      # suppress error reports
+      NOERROR     = 1 << 5
+      # suppress warning reports
+      NOWARNING   = 1 << 6
+      # pedantic error reporting
+      PEDANTIC    = 1 << 7
+      # remove blank nodes
+      NOBLANKS    = 1 << 8
+      # use the SAX1 interface internally
+      SAX1        = 1 << 9
+      # Implement XInclude substitution
+      XINCLUDE    = 1 << 10
+      # Forbid network access. Recommended for dealing with untrusted documents.
+      NONET       = 1 << 11
+      # Do not reuse the context dictionary
+      NODICT      = 1 << 12
+      # remove redundant namespaces declarations
+      NSCLEAN     = 1 << 13
+      # merge CDATA as text nodes
+      NOCDATA     = 1 << 14
+      # do not generate XINCLUDE START/END nodes
+      NOXINCNODE  = 1 << 15
+      # compact small text nodes; no modification of the tree allowed afterwards (will possibly crash if you try to modify the tree)
+      COMPACT     = 1 << 16
+      # parse using XML-1.0 before update 5
+      OLD10       = 1 << 17
+      # do not fixup XINCLUDE xml:base uris
+      NOBASEFIX   = 1 << 18
+      # relax any hardcoded limit from the parser
+      HUGE        = 1 << 19
+
+      # the default options used for parsing XML documents
+      DEFAULT_XML  = RECOVER | NONET
+      # the default options used for parsing HTML documents
+      DEFAULT_HTML = RECOVER | NOERROR | NOWARNING | NONET
+
+      attr_accessor :options
+      def initialize options = STRICT
+        @options = options
+      end
+
+      constants.each do |constant|
+        next if constant.to_sym == :STRICT
+        class_eval %{
+          def #{constant.downcase}
+            @options |= #{constant}
+            self
+          end
+
+          def no#{constant.downcase}
+            @options &= ~#{constant}
+            self
+          end
+
+          def #{constant.downcase}?
+            #{constant} & @options == #{constant}
+          end
+        }
+      end
+
+      def strict
+        @options &= ~RECOVER
+        self
+      end
+
+      def strict?
+        @options & RECOVER == STRICT
+      end
+
+      alias :to_i :options
+
+      def inspect
+        options = []
+        self.class.constants.each do |k|
+          options << k.downcase if send(:"#{k.downcase}?")
+        end
+        super.sub(/>$/, " " + options.join(', ') + ">")
+      end
+    end
+  end
+end

data/lib/nokogiri/xml/pp/character_data.rb

@@ -0,0 +1,18 @@
+module Nokogiri
+  module XML
+    module PP
+      module CharacterData
+        def pretty_print pp # :nodoc:
+          nice_name = self.class.name.split('::').last
+          pp.group(2, "#(#{nice_name} ", ')') do
+            pp.pp text
+          end
+        end
+
+        def inspect # :nodoc:
+          "#<#{self.class.name}:#{sprintf("0x%x",object_id)} #{text.inspect}>"
+        end
+      end
+    end
+  end
+end

data/lib/nokogiri/xml/pp/node.rb

@@ -0,0 +1,56 @@
+module Nokogiri
+  module XML
+    module PP
+      module Node
+        def inspect # :nodoc:
+          attributes = inspect_attributes.reject { |x|
+            begin
+              attribute = send x
+              !attribute || (attribute.respond_to?(:empty?) && attribute.empty?)
+            rescue NoMethodError
+              true
+            end
+          }.map { |attribute|
+            "#{attribute.to_s.sub(/_\w+/, 's')}=#{send(attribute).inspect}"
+          }.join ' '
+          "#<#{self.class.name}:#{sprintf("0x%x", object_id)} #{attributes}>"
+        end
+
+        def pretty_print pp # :nodoc:
+          nice_name = self.class.name.split('::').last
+          pp.group(2, "#(#{nice_name}:#{sprintf("0x%x", object_id)} {", '})') do
+
+            pp.breakable
+            attrs = inspect_attributes.map { |t|
+              [t, send(t)] if respond_to?(t)
+            }.compact.find_all { |x|
+              if x.last
+                if [:attribute_nodes, :children].include? x.first
+                  !x.last.empty?
+                else
+                  true
+                end
+              end
+            }
+
+            pp.seplist(attrs) do |v|
+              if [:attribute_nodes, :children].include? v.first
+                pp.group(2, "#{v.first.to_s.sub(/_\w+$/, 's')} = [", "]") do
+                  pp.breakable
+                  pp.seplist(v.last) do |item|
+                    pp.pp item
+                  end
+                end
+              else
+                pp.text "#{v.first} = "
+                pp.pp v.last
+              end
+            end
+            pp.breakable
+
+          end
+        end
+      end
+    end
+  end
+end

data/lib/nokogiri/xml/pp.rb

@@ -0,0 +1,2 @@
+require 'nokogiri/xml/pp/node'
+require 'nokogiri/xml/pp/character_data'

data/lib/nokogiri/xml/processing_instruction.rb

@@ -0,0 +1,8 @@
+module Nokogiri
+  module XML
+    class ProcessingInstruction < Node
+      def initialize document, name, content
+      end
+    end
+  end
+end

data/lib/nokogiri/xml/reader.rb

@@ -1,14 +1,112 @@
 module Nokogiri
   module XML
+    ###
+    # Nokogiri::XML::Reader parses an XML document similar to the way a cursor
+    # would move.  The Reader is given an XML document, and yields nodes
+    # to an each block.
+    #
+    # Here is an example of usage:
+    #
+    #   reader = Nokogiri::XML::Reader(<<-eoxml)
+    #     <x xmlns:tenderlove='http://tenderlovemaking.com/'>
+    #       <tenderlove:foo awesome='true'>snuggles!</tenderlove:foo>
+    #     </x>
+    #   eoxml
+    #
+    #   reader.each do |node|
+    #
+    #     # node is an instance of Nokogiri::XML::Reader
+    #     puts node.name
+    #
+    #   end
+    #
+    # Note that Nokogiri::XML::Reader#each can only be called once!!  Once
+    # the cursor moves through the entire document, you must parse the
+    # document again.  So make sure that you capture any information you
+    # need during the first iteration.
+    #
+    # The Reader parser is good for when you need the speed of a SAX parser,
+    # but do not want to write a Document handler.
     class Reader
       include Enumerable
 
-      def each(&block)
-        while node = self.read
-          block.call(node)
-        end
+      TYPE_NONE = 0
+      # Element node type
+      TYPE_ELEMENT = 1
+      # Attribute node type
+      TYPE_ATTRIBUTE = 2
+      # Text node type
+      TYPE_TEXT = 3
+      # CDATA node type
+      TYPE_CDATA = 4
+      # Entity Reference node type
+      TYPE_ENTITY_REFERENCE = 5
+      # Entity node type
+      TYPE_ENTITY = 6
+      # PI node type
+      TYPE_PROCESSING_INSTRUCTION = 7
+      # Comment node type
+      TYPE_COMMENT = 8
+      # Document node type
+      TYPE_DOCUMENT = 9
+      # Document Type node type
+      TYPE_DOCUMENT_TYPE = 10
+      # Document Fragment node type
+      TYPE_DOCUMENT_FRAGMENT = 11
+      # Notation node type
+      TYPE_NOTATION = 12
+      # Whitespace node type
+      TYPE_WHITESPACE = 13
+      # Significant Whitespace node type
+      TYPE_SIGNIFICANT_WHITESPACE = 14
+      # Element end node type
+      TYPE_END_ELEMENT = 15
+      # Entity end node type
+      TYPE_END_ENTITY = 16
+      # XML Declaration node type
+      TYPE_XML_DECLARATION = 17
+
+      # A list of errors encountered while parsing
+      attr_accessor :errors
+
+      # The encoding for the document
+      attr_reader :encoding
+
+      # The XML source
+      attr_reader :source
+
+      alias :self_closing? :empty_element?
+
+      def initialize source, url = nil, encoding = nil # :nodoc:
+        @source   = source
+        @errors   = []
+        @encoding = encoding
       end
       private :initialize
+
+      ###
+      # Get a list of attributes for the current node.
+      def attributes
+        Hash[attribute_nodes.map { |node|
+          [node.name, node.to_s]
+        }].merge(namespaces || {})
+      end
+
+      ###
+      # Get a list of attributes for the current node
+      def attribute_nodes
+        nodes = attr_nodes
+        nodes.each { |v| v.instance_variable_set(:@_r, self) }
+        nodes
+      end
+
+      ###
+      # Move the cursor through the document yielding the cursor to the block
+      def each
+        while cursor = self.read
+          yield cursor
+        end
+      end
     end
   end
 end