nokogiri 1.10.10 → 1.12.3

data/lib/nokogiri/xml/node_set.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     ####

data/lib/nokogiri/xml/notation.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     class Notation < Struct.new(:name, :public_id, :system_id)

data/lib/nokogiri/xml/parse_options.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     ###
@@ -5,11 +6,11 @@ module Nokogiri
     #
     # == 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.
+    # *Note*: All examples attempt to set the +RECOVER+ & +NOENT+ options.
     # [Ruby's bitwise operators] You can use the Ruby bitwise operators to set various combinations.
-    #   Nokogiri.XML('<content>Chapter 1</content', options: Nokogiri::XML::ParseOptions.new((1 << 0) | (1 << 1)))
+    #   Nokogiri.XML('<content>Chapter 1</content', nil, nil, Nokogiri::XML::ParseOptions.new((1 << 0) | (1 << 1)))
     # [Method chaining] Every option has an equivalent method in lowercase. You can chain these methods together to set various combinations.
-    #   Nokogiri.XML('<content>Chapter 1</content', options: Nokogiri::XML::ParseOptions.new.recover.noent)
+    #   Nokogiri.XML('<content>Chapter 1</content', nil, nil, Nokogiri::XML::ParseOptions.new.recover.noent)
     # [Using Ruby Blocks] You can also setup parse combinations in the block passed to Nokogiri.XML or Nokogiri.HTML
     #   Nokogiri.XML('<content>Chapter 1</content') {|config| config.recover.noent}
     #
@@ -70,8 +71,12 @@ module Nokogiri
 
       # the default options used for parsing XML documents
       DEFAULT_XML  = RECOVER | NONET
+      # the default options used for parsing XSLT stylesheets
+      DEFAULT_XSLT = RECOVER | NONET | NOENT | DTDLOAD | DTDATTR | NOCDATA
       # the default options used for parsing HTML documents
       DEFAULT_HTML = RECOVER | NOERROR | NOWARNING | NONET
+      # the default options used for parsing XML schemas
+      DEFAULT_SCHEMA = NONET
 
       attr_accessor :options
       def initialize options = STRICT
@@ -106,6 +111,10 @@ module Nokogiri
         @options & RECOVER == STRICT
       end
 
+      def ==(other)
+        other.to_i == to_i
+      end
+
       alias :to_i :options
 
       def inspect

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

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     module PP

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

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     module PP

data/lib/nokogiri/xml/pp.rb

@@ -1,2 +1,3 @@
-require 'nokogiri/xml/pp/node'
-require 'nokogiri/xml/pp/character_data'
+# frozen_string_literal: true
+require_relative "pp/node"
+require_relative "pp/character_data"

data/lib/nokogiri/xml/processing_instruction.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     class ProcessingInstruction < Node

data/lib/nokogiri/xml/reader.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     ###
@@ -85,19 +86,15 @@ module Nokogiri
       private :initialize
 
       ###
-      # Get a list of attributes for the current node.
+      # Get the attributes of the current node as a Hash
+      # @return [Hash<String, String>] Attribute names and values
       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
+        attrs_hash = attribute_nodes.each_with_object({}) do |node, hash|
+                       hash[node.name] = node.to_s
+                     end
+        ns = namespaces
+        attrs_hash.merge!(ns) if ns
+        attrs_hash
       end
 
       ###

data/lib/nokogiri/xml/relax_ng.rb

@@ -1,11 +1,12 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     class << self
       ###
       # Create a new Nokogiri::XML::RelaxNG document from +string_or_io+.
       # See Nokogiri::XML::RelaxNG for an example.
-      def RelaxNG string_or_io
-        RelaxNG.new(string_or_io)
+      def RelaxNG(string_or_io, options = ParseOptions::DEFAULT_SCHEMA)
+        RelaxNG.new(string_or_io, options)
       end
     end
 
@@ -26,6 +27,10 @@ module Nokogiri
     #   end
     #
     # The list of errors are Nokogiri::XML::SyntaxError objects.
+    #
+    # NOTE: RelaxNG input is always treated as TRUSTED documents, meaning that they will cause the
+    # underlying parsing libraries to access network resources. This is counter to Nokogiri's
+    # "untrusted by default" security policy, but is a limitation of the underlying libraries.
     class RelaxNG < Nokogiri::XML::Schema
     end
   end

data/lib/nokogiri/xml/sax/document.rb

@@ -1,20 +1,20 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     ###
-    # SAX Parsers are event driven parsers.  Nokogiri provides two different
-    # event based parsers when dealing with XML.  If you want to do SAX style
-    # parsing using HTML, check out Nokogiri::HTML::SAX.
+    # SAX Parsers are event driven parsers. Nokogiri provides two different event based parsers when
+    # dealing with XML. If you want to do SAX style parsing using HTML, check out
+    # Nokogiri::HTML4::SAX.
     #
-    # The basic way a SAX style parser works is by creating a parser,
-    # telling the parser about the events we're interested in, then giving
-    # the parser some XML to process.  The parser will notify you when
-    # it encounters events you said you would like to know about.
+    # The basic way a SAX style parser works is by creating a parser, telling the parser about the
+    # events we're interested in, then giving the parser some XML to process. The parser will notify
+    # you when it encounters events you said you would like to know about.
     #
-    # To register for events, you simply subclass Nokogiri::XML::SAX::Document,
-    # and implement the methods for which you would like notification.
+    # To register for events, you simply subclass Nokogiri::XML::SAX::Document, and implement the
+    # methods for which you would like notification.
     #
-    # For example, if I want to be notified when a document ends, and when an
-    # element starts, I would write a class like this:
+    # For example, if I want to be notified when a document ends, and when an element starts, I
+    # would write a class like this:
     #
     #   class MyDocument < Nokogiri::XML::SAX::Document
     #     def end_document
@@ -26,8 +26,7 @@ module Nokogiri
     #     end
     #   end
     #
-    # Then I would instantiate a SAX parser with this document, and feed the
-    # parser some XML
+    # Then I would instantiate a SAX parser with this document, and feed the parser some XML
     #
     #   # Create a new parser
     #   parser = Nokogiri::XML::SAX::Parser.new(MyDocument.new)
@@ -35,25 +34,21 @@ module Nokogiri
     #   # Feed the parser some XML
     #   parser.parse(File.open(ARGV[0]))
     #
-    # Now my document handler will be called when each node starts, and when
-    # then document ends.  To see what kinds of events are available, take
-    # a look at Nokogiri::XML::SAX::Document.
+    # Now my document handler will be called when each node starts, and when then document ends. To
+    # see what kinds of events are available, take a look at Nokogiri::XML::SAX::Document.
     #
-    # Two SAX parsers for XML are available, a parser that reads from a string
-    # or IO object as it feels necessary, and a parser that lets you spoon
-    # feed it XML.  If you want to let Nokogiri deal with reading your XML,
-    # use the Nokogiri::XML::SAX::Parser.  If you want to have fine grain
+    # Two SAX parsers for XML are available, a parser that reads from a string or IO object as it
+    # feels necessary, and a parser that lets you spoon feed it XML. If you want to let Nokogiri
+    # deal with reading your XML, use the Nokogiri::XML::SAX::Parser. If you want to have fine grain
     # control over the XML input, use the Nokogiri::XML::SAX::PushParser.
     module SAX
       ###
-      # This class is used for registering types of events you are interested
-      # in handling.  All of the methods on this class are available as
-      # possible events while parsing an XML document.  To register for any
-      # particular event, just subclass this class and implement the methods
-      # you are interested in knowing about.
+      # This class is used for registering types of events you are interested in handling. All of
+      # the methods on this class are available as possible events while parsing an XML document. To
+      # register for any particular event, just subclass this class and implement the methods you
+      # are interested in knowing about.
       #
-      # To only be notified about start and end element events, write a class
-      # like this:
+      # To only be notified about start and end element events, write a class like this:
       #
       #   class MyDocument < Nokogiri::XML::SAX::Document
       #     def start_element name, attrs = []
@@ -65,8 +60,8 @@ module Nokogiri
       #     end
       #   end
       #
-      # You can use this event handler for any SAX style parser included with
-      # Nokogiri.  See Nokogiri::XML::SAX, and Nokogiri::HTML::SAX.
+      # You can use this event handler for any SAX style parser included with Nokogiri. See
+      # Nokogiri::XML::SAX, and Nokogiri::HTML4::SAX.
       class Document
         ###
         # Called when an XML declaration is parsed
@@ -128,7 +123,7 @@ module Nokogiri
         end
 
         ###
-        # Characters read between a tag.  This method might be called multiple
+        # Characters read between a tag. This method might be called multiple
         # times given one contiguous string of characters.
         #
         # +string+ contains the character data

data/lib/nokogiri/xml/sax/parser.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     module SAX

data/lib/nokogiri/xml/sax/parser_context.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     module SAX

data/lib/nokogiri/xml/sax/push_parser.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     module SAX

data/lib/nokogiri/xml/sax.rb

@@ -1,4 +1,5 @@
-require 'nokogiri/xml/sax/document'
-require 'nokogiri/xml/sax/parser_context'
-require 'nokogiri/xml/sax/parser'
-require 'nokogiri/xml/sax/push_parser'
+# frozen_string_literal: true
+require_relative "sax/document"
+require_relative "sax/parser_context"
+require_relative "sax/parser"
+require_relative "sax/push_parser"

data/lib/nokogiri/xml/schema.rb

@@ -1,11 +1,12 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     class << self
       ###
       # Create a new Nokogiri::XML::Schema object using a +string_or_io+
       # object.
-      def Schema string_or_io
-        Schema.new(string_or_io)
+      def Schema(string_or_io, options = ParseOptions::DEFAULT_SCHEMA)
+        Schema.new(string_or_io, options)
       end
     end
 
@@ -26,15 +27,23 @@ module Nokogiri
     #   end
     #
     # The list of errors are Nokogiri::XML::SyntaxError objects.
+    #
+    # NOTE: As of v1.11.0, Schema treats inputs as UNTRUSTED by default, and so external entities
+    # are not resolved from the network (`http://` or `ftp://`). Previously, parsing treated
+    # documents as "trusted" by default which was counter to Nokogiri's "untrusted by default"
+    # security policy. If a document is trusted, then the caller may turn off the NONET option via
+    # the ParseOptions to re-enable external entity resolution over a network connection.
     class Schema
       # Errors while parsing the schema file
       attr_accessor :errors
+      # The Nokogiri::XML::ParseOptions used to parse the schema
+      attr_accessor :parse_options
 
       ###
       # Create a new Nokogiri::XML::Schema object using a +string_or_io+
       # object.
-      def self.new string_or_io
-        from_document Nokogiri::XML(string_or_io)
+      def self.new string_or_io, options = ParseOptions::DEFAULT_SCHEMA
+        from_document(Nokogiri::XML(string_or_io), options)
       end
 
       ###

data/lib/nokogiri/xml/searchable.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     #
@@ -11,7 +12,9 @@ module Nokogiri
       # Regular expression used by Searchable#search to determine if a query
       # string is CSS or XPath
       LOOKS_LIKE_XPATH = /^(\.\/|\/|\.\.|\.$)/
-      
+
+      # @!group Searching via XPath or CSS Queries
+
       ###
       # call-seq: search *paths, [namespace-bindings, xpath-variable-bindings, custom-handler-class]
       #
@@ -45,7 +48,7 @@ module Nokogiri
       #   )
       #
       # See Searchable#xpath and Searchable#css for further usage help.
-      def search *args
+      def search(*args)
         paths, handler, ns, binds = extract_params(args)
 
         xpaths = paths.map(&:to_s).map do |path|
@@ -54,6 +57,7 @@ module Nokogiri
 
         xpath(*(xpaths + [ns, handler, binds].compact))
       end
+
       alias :/ :search
 
       ###
@@ -63,9 +67,10 @@ module Nokogiri
       # result. +paths+ must be one or more XPath or CSS queries.
       #
       # See Searchable#search for more information.
-      def at *args
+      def at(*args)
         search(*args).first
       end
+
       alias :% :at
 
       ###
@@ -101,7 +106,7 @@ module Nokogiri
       # found in an XML document, where tags names are case-sensitive
       # (e.g., "H1" is distinct from "h1").
       #
-      def css *args
+      def css(*args)
         rules, handler, ns, _ = extract_params(args)
 
         css_internal self, rules, handler, ns
@@ -114,7 +119,7 @@ module Nokogiri
       # match. +rules+ must be one or more CSS selectors.
       #
       # See Searchable#css for more information.
-      def at_css *args
+      def at_css(*args)
         css(*args).first
       end
 
@@ -148,7 +153,7 @@ module Nokogiri
       #     end
       #   }.new)
       #
-      def xpath *args
+      def xpath(*args)
         paths, handler, ns, binds = extract_params(args)
 
         xpath_internal self, paths, handler, ns, binds
@@ -161,17 +166,19 @@ module Nokogiri
       # match. +paths+ must be one or more XPath queries.
       #
       # See Searchable#xpath for more information.
-      def at_xpath *args
+      def at_xpath(*args)
         xpath(*args).first
       end
 
+      # @!endgroup
+
       private
 
-      def css_internal node, rules, handler, ns
+      def css_internal(node, rules, handler, ns)
         xpath_internal node, css_rules_to_xpath(rules, ns), handler, ns, nil
       end
 
-      def xpath_internal node, paths, handler, ns, binds
+      def xpath_internal(node, paths, handler, ns, binds)
         document = node.document
         return NodeSet.new(document) unless document
 
@@ -186,12 +193,12 @@ module Nokogiri
         end
       end
 
-      def xpath_impl node, path, handler, ns, binds
+      def xpath_impl(node, path, handler, ns, binds)
         ctx = XPathContext.new(node)
         ctx.register_namespaces(ns)
-        path = path.gsub(/xmlns:/, ' :') unless Nokogiri.uses_libxml?
+        path = path.gsub(/xmlns:/, " :") unless Nokogiri.uses_libxml?
 
-        binds.each do |key,value|
+        binds.each do |key, value|
           ctx.register_variable key.to_s, value
         end if binds
 
@@ -202,13 +209,15 @@ module Nokogiri
         rules.map { |rule| xpath_query_from_css_rule(rule, ns) }
       end
 
-      def xpath_query_from_css_rule rule, ns
+      def xpath_query_from_css_rule(rule, ns)
+        visitor = Nokogiri::CSS::XPathVisitorOptimallyUseBuiltins.new
         self.class::IMPLIED_XPATH_CONTEXTS.map do |implied_xpath_context|
-          CSS.xpath_for(rule.to_s, :prefix => implied_xpath_context, :ns => ns)
-        end.join(' | ')
+          CSS.xpath_for(rule.to_s, {:prefix => implied_xpath_context, :ns => ns,
+                                    :visitor => visitor})
+        end.join(" | ")
       end
 
-      def extract_params params # :nodoc:
+      def extract_params(params) # :nodoc:
         handler = params.find do |param|
           ![Hash, String, Symbol].include?(param.class)
         end

data/lib/nokogiri/xml/syntax_error.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     ###

data/lib/nokogiri/xml/text.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module XML
     class Text < Nokogiri::XML::CharacterData