oga 1.0.2-java → 1.0.3-java

data/lib/oga/xml/lexer.rb

@@ -34,12 +34,20 @@ module Oga
     # However, it is perfectly save to use different instances per thread.
     # There is no _global_ state used by this lexer.
     #
-    # @!attribute [r] html
-    #  @return [TrueClass|FalseClass]
+    # ## Strict Mode
+    #
+    # By default the lexer is rather permissive regarding the input. For
+    # example, missing closing tags are inserted by default. To disable this
+    # behaviour the lexer can be run in "strict mode" by setting `:strict` to
+    # `true`:
+    #
+    #     lexer = Oga::XML::Lexer.new('...', :strict => true)
+    #
+    # Strict mode only applies to XML documents.
+    #
+    # @private
     #
     class Lexer
-      attr_reader :html
-
       # These are all constant/frozen to remove the need for String allocations
       # every time they are referenced in the lexer.
       HTML_SCRIPT = 'script'.freeze
@@ -99,13 +107,17 @@ module Oga
       #
       # @param [Hash] options
       #
-      # @option options [Symbol] :html When set to `true` the lexer will treat
-      #  the input as HTML instead of SGML/XML. This makes it possible to lex
-      #  HTML void elements such as `<link href="">`.
+      # @option options [TrueClass|FalseClass] :html When set to `true` the
+      #  lexer will treat the input as HTML instead of XML. This makes it
+      #  possible to lex HTML void elements such as `<link href="">`.
+      #
+      # @option options [TrueClass|FalseClass] :strict Enables/disables strict
+      #  parsing of XML documents, disabled by default.
       #
       def initialize(data, options = {})
-        @data = data
-        @html = options[:html]
+        @data   = data
+        @html   = options[:html]
+        @strict = options[:strict] || false
 
         reset
       end
@@ -163,7 +175,7 @@ module Oga
 
         reset
 
-        return tokens
+        tokens
       end
 
       ##
@@ -193,7 +205,7 @@ module Oga
         end
 
         # Add any missing closing tags
-        unless @elements.empty?
+        if !strict? and !@elements.empty?
           @elements.length.times { on_element_end }
         end
       ensure
@@ -204,21 +216,28 @@ module Oga
       # @return [TrueClass|FalseClass]
       #
       def html?
-        return !!html
+        @html == true
+      end
+
+      ##
+      # @return [TrueClass|FalseClass]
+      #
+      def strict?
+        @strict
       end
 
       ##
       # @return [TrueClass|FalseClass]
       #
       def html_script?
-        return html? && current_element == HTML_SCRIPT
+        html? && current_element == HTML_SCRIPT
       end
 
       ##
       # @return [TrueClass|FalseClass]
       #
       def html_style?
-        return html? && current_element == HTML_STYLE
+        html? && current_element == HTML_STYLE
       end
 
       private
@@ -250,7 +269,7 @@ module Oga
       # @return [String]
       #
       def current_element
-        return @elements.last
+        @elements.last
       end
 
       ##

data/lib/oga/xml/namespace.rb

@@ -4,14 +4,12 @@ module Oga
     # The Namespace class contains information about XML namespaces such as the
     # name and URI.
     #
-    # @!attribute [r] name
-    #  @return [String]
-    #
-    # @!attribute [r] uri
-    #  @return [String]
-    #
     class Namespace
-      attr_accessor :name, :uri
+      # @return [String]
+      attr_accessor :name
+
+      # @return [String]
+      attr_accessor :uri
 
       ##
       # @param [Hash] options
@@ -28,14 +26,14 @@ module Oga
       # @return [String]
       #
       def to_s
-        return name.to_s
+        name.to_s
       end
 
       ##
       # @return [String]
       #
       def inspect
-        return "Namespace(name: #{name.inspect} uri: #{uri.inspect})"
+        "Namespace(name: #{name.inspect} uri: #{uri.inspect})"
       end
 
       ##
@@ -43,7 +41,7 @@ module Oga
       # @return [TrueClass|FalseClass]
       #
       def ==(other)
-        return other.is_a?(self.class) && name == other.name && uri == other.uri
+        other.is_a?(self.class) && name == other.name && uri == other.uri
       end
     end # Namespace
   end # XML

data/lib/oga/xml/node.rb

@@ -5,12 +5,10 @@ module Oga
     # {Oga::XML::NodeSet} and can be used to query surrounding and parent
     # nodes.
     #
-    # @!attribute [r] node_set
-    #  @return [Oga::XML::NodeSet]
-    #
     class Node
       include Traversal
 
+      # @return [Oga::XML::NodeSet]
       attr_reader :node_set
 
       ##
@@ -42,7 +40,7 @@ module Oga
       # @return [Oga::XML::NodeSet]
       #
       def children
-        return @children ||= NodeSet.new([], self)
+        @children ||= NodeSet.new([], self)
       end
 
       ##
@@ -65,7 +63,7 @@ module Oga
       # @return [Oga::XML::Node]
       #
       def parent
-        return node_set ? node_set.owner : nil
+        node_set ? node_set.owner : nil
       end
 
       ##
@@ -76,7 +74,7 @@ module Oga
       def previous
         index = node_set.index(self) - 1
 
-        return index >= 0 ? node_set[index] : nil
+        index >= 0 ? node_set[index] : nil
       end
 
       ##
@@ -88,7 +86,7 @@ module Oga
         index  = node_set.index(self) + 1
         length = node_set.length
 
-        return index <= length ? node_set[index] : nil
+        index <= length ? node_set[index] : nil
       end
 
       ##
@@ -142,7 +140,7 @@ module Oga
           @root_node = node
         end
 
-        return @root_node
+        @root_node
       end
 
       ##
@@ -186,14 +184,14 @@ module Oga
           @html_p = root.is_a?(Document) && root.html?
         end
 
-        return @html_p
+        @html_p
       end
 
       ##
       # @return [TrueClass|FalseClass]
       #
       def xml?
-        return !html?
+        !html?
       end
     end # Element
   end # XML

data/lib/oga/xml/node_set.rb

@@ -31,12 +31,10 @@ module Oga
     # If ownership was not handled then you'd have to manually set the
     # `element` variable's `node_set` attribute after pushing it into a set.
     #
-    # @!attribute [rw] owner
-    #  @return [Oga::XML::Node]
-    #
     class NodeSet
       include Enumerable
 
+      # @return [Oga::XML::Node]
       attr_accessor :owner
 
       ##
@@ -65,7 +63,7 @@ module Oga
       # @return [Oga::XML::Node]
       #
       def last
-        return @nodes[-1]
+        @nodes[-1]
       end
 
       ##
@@ -74,7 +72,7 @@ module Oga
       # @return [TrueClass|FalseClass]
       #
       def empty?
-        return @nodes.empty?
+        @nodes.empty?
       end
 
       ##
@@ -83,7 +81,7 @@ module Oga
       # @return [Fixnum]
       #
       def length
-        return @nodes.length
+        @nodes.length
       end
 
       alias_method :count, :length
@@ -96,7 +94,7 @@ module Oga
       # @return [Fixnum]
       #
       def index(node)
-        return @nodes.index(node)
+        @nodes.index(node)
       end
 
       ##
@@ -137,7 +135,7 @@ module Oga
 
         remove_ownership(node)
 
-        return node
+        node
       end
 
       ##
@@ -150,7 +148,7 @@ module Oga
 
         remove_ownership(node)
 
-        return node
+        node
       end
 
       ##
@@ -174,7 +172,7 @@ module Oga
       # @return [Oga::XML::Node]
       #
       def [](index)
-        return @nodes[index]
+        @nodes[index]
       end
 
       ##
@@ -183,7 +181,7 @@ module Oga
       # @return [Array]
       #
       def to_a
-        return @nodes
+        @nodes
       end
 
       ##
@@ -194,7 +192,7 @@ module Oga
       # @return [Oga::XML::NodeSet]
       #
       def +(other)
-        return self.class.new(to_a | other.to_a)
+        self.class.new(to_a | other.to_a)
       end
 
       ##
@@ -204,7 +202,7 @@ module Oga
       # @param [Oga::XML::NodeSet] other
       #
       def ==(other)
-        return other.is_a?(NodeSet) && other.equal_nodes?(@nodes)
+        other.is_a?(NodeSet) && other.equal_nodes?(@nodes)
       end
 
       ##
@@ -216,7 +214,7 @@ module Oga
       # @param [Array<Oga::XML::Node>] nodes
       #
       def equal_nodes?(nodes)
-        return @nodes == nodes
+        @nodes == nodes
       end
 
       ##
@@ -261,7 +259,7 @@ module Oga
 
         remove_ownership(removed) if removed
 
-        return removed
+        removed
       end
 
       ##
@@ -279,7 +277,7 @@ module Oga
           end
         end
 
-        return values
+        values
       end
 
       alias_method :attr, :attribute
@@ -298,7 +296,7 @@ module Oga
           end
         end
 
-        return text
+        text
       end
 
       ##
@@ -307,7 +305,7 @@ module Oga
       def inspect
         values = @nodes.map(&:inspect).join(', ')
 
-        return "NodeSet(#{values})"
+        "NodeSet(#{values})"
       end
 
       private

data/lib/oga/xml/parser.rb

@@ -291,7 +291,7 @@ class Parser < LL::Driver
   # @return [Oga::XML::Document]
   #
   def on_document(children = [])
-    document = Document.new(:type => @lexer.html ? :html : :xml)
+    document = Document.new(:type => @lexer.html? ? :html : :xml)
 
     children.each do |child|
       if child.is_a?(Doctype)

data/lib/oga/xml/processing_instruction.rb

@@ -3,10 +3,8 @@ module Oga
     ##
     # Class used for storing information about a single processing instruction.
     #
-    # @!attribute [rw] name
-    #  @return [String]
-    #
     class ProcessingInstruction < CharacterNode
+      # @return [String]
       attr_accessor :name
 
       ##
@@ -25,14 +23,14 @@ module Oga
       # @return [String]
       #
       def to_xml
-        return "<?#{name}#{text}?>"
+        "<?#{name}#{text}?>"
       end
 
       ##
       # @return [String]
       #
       def inspect
-        return "ProcessingInstruction(name: #{name.inspect} text: #{text.inspect})"
+        "ProcessingInstruction(name: #{name.inspect} text: #{text.inspect})"
       end
     end # ProcessingInstruction
   end # XML

data/lib/oga/xml/pull_parser.rb

@@ -19,16 +19,13 @@ module Oga
     # This parses yields proper XML instances such as {Oga::XML::Element}.
     # Doctypes and XML declarations are ignored by this parser.
     #
-    # @!attribute [r] node
-    #  The current node.
-    #  @return [Oga::XML::Node]
-    #
-    # @!attribute [r] nesting
-    #  Array containing the names of the currently nested elements.
-    #  @return [Array]
-    #
     class PullParser < Parser
-      attr_reader :node, :nesting
+      # @return [Oga::XML::Node]
+      attr_reader :node
+
+      # Array containing the names of the currently nested elements.
+      # @return [Array]
+      attr_reader :nesting
 
       ##
       # @return [Array]

data/lib/oga/xml/querying.rb

@@ -13,7 +13,7 @@ module Oga
       # @see [Oga::XPath::Evaluator#initialize]
       #
       def xpath(expression, variables = {})
-        return XPath::Evaluator.new(self, variables).evaluate(expression)
+        XPath::Evaluator.new(self, variables).evaluate(expression)
       end
 
       ##
@@ -25,7 +25,7 @@ module Oga
       def at_xpath(*args)
         result = xpath(*args)
 
-        return result.is_a?(XML::NodeSet) ? result.first : result
+        result.is_a?(XML::NodeSet) ? result.first : result
       end
 
       ##
@@ -37,7 +37,7 @@ module Oga
       def css(expression)
         ast = CSS::Parser.parse_with_cache(expression)
 
-        return XPath::Evaluator.new(self).evaluate_ast(ast)
+        XPath::Evaluator.new(self).evaluate_ast(ast)
       end
 
       ##
@@ -49,7 +49,7 @@ module Oga
       def at_css(*args)
         result = css(*args)
 
-        return result.is_a?(XML::NodeSet) ? result.first : result
+        result.is_a?(XML::NodeSet) ? result.first : result
       end
     end # Querying
   end # XML