nokogiri 1.13.6 → 1.14.2

data/lib/nokogiri/xml/parse_options.rb

@@ -1,84 +1,163 @@
+# coding: utf-8
 # frozen_string_literal: true
 
 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.
-    # [Ruby's bitwise operators] You can use the Ruby bitwise operators to set various combinations.
-    #   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', 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}
-    #
-    # == 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
+    # Options that control the parsing behavior for XML::Document, XML::DocumentFragment,
+    # HTML4::Document, HTML4::DocumentFragment, XSLT::Stylesheet, and XML::Schema.
+    #
+    # These options directly expose libxml2's parse options, which are all boolean in the sense that
+    # an option is "on" or "off".
+    #
+    # 💡 Note that HTML5 parsing has a separate, orthogonal set of options due to the nature of the
+    # HTML5 specification. See Nokogiri::HTML5.
+    #
+    # ⚠ Not all parse options are supported on JRuby. Nokogiri will attempt to invoke the equivalent
+    # behavior in Xerces/NekoHTML on JRuby when it's possible.
+    #
+    # == Setting and unsetting parse options
+    #
+    # You can build your own combinations of parse options by using any of the following methods:
+    #
+    # [ParseOptions method chaining]
+    #
+    #   Every option has an equivalent method in lowercase. You can chain these methods together to
+    #   set various combinations.
+    #
+    #     # Set the HUGE & PEDANTIC options
+    #     po = Nokogiri::XML::ParseOptions.new.huge.pedantic
+    #     doc = Nokogiri::XML::Document.parse(xml, nil, nil, po)
+    #
+    #   Every option has an equivalent <code>no{option}</code> method in lowercase. You can call these
+    #   methods on an instance of ParseOptions to unset the option.
+    #
+    #     # Set the HUGE & PEDANTIC options
+    #     po = Nokogiri::XML::ParseOptions.new.huge.pedantic
+    #
+    #     # later we want to modify the options
+    #     po.nohuge # Unset the HUGE option
+    #     po.nopedantic # Unset the PEDANTIC option
+    #
+    #   💡 Note that some options begin with "no" leading to the logical but perhaps unintuitive
+    #   double negative:
+    #
+    #     po.nocdata # Set the NOCDATA parse option
+    #     po.nonocdata # Unset the NOCDATA parse option
+    #
+    #   💡 Note that negation is not available for STRICT, which is itself a negation of all other
+    #   features.
+    #
+    #
+    # [Using Ruby Blocks]
+    #
+    #   Most parsing methods will accept a block for configuration of parse options, and we
+    #   recommend chaining the setter methods:
+    #
+    #     doc = Nokogiri::XML::Document.parse(xml) { |config| config.huge.pedantic }
+    #
+    #
+    # [ParseOptions constants]
+    #
+    #   You can also use the constants declared under Nokogiri::XML::ParseOptions to set various
+    #   combinations. They are bits in a bitmask, and so can be combined with bitwise operators:
+    #
+    #     po = Nokogiri::XML::ParseOptions.new(Nokogiri::XML::ParseOptions::HUGE | Nokogiri::XML::ParseOptions::PEDANTIC)
+    #     doc = Nokogiri::XML::Document.parse(xml, nil, nil, po)
     #
     class ParseOptions
       # Strict parsing
       STRICT      = 0
-      # Recover from errors
+
+      # Recover from errors. On by default for XML::Document, XML::DocumentFragment,
+      # HTML4::Document, HTML4::DocumentFragment, XSLT::Stylesheet, and XML::Schema.
       RECOVER     = 1 << 0
-      # Substitute entities
+
+      # Substitute entities. Off by default.
+      #
+      # ⚠ This option enables entity substitution, contrary to what the name implies.
+      #
+      # ⚠ <b>It is UNSAFE to set this option</b> when parsing untrusted documents.
       NOENT       = 1 << 1
-      # Load external subsets
+
+      # Load external subsets. On by default for XSLT::Stylesheet.
+      #
+      # ⚠ <b>It is UNSAFE to set this option</b> when parsing untrusted documents.
       DTDLOAD     = 1 << 2
-      # Default DTD attributes
+
+      # Default DTD attributes. On by default for XSLT::Stylesheet.
       DTDATTR     = 1 << 3
-      # validate with the DTD
+
+      # Validate with the DTD. Off by default.
       DTDVALID    = 1 << 4
-      # suppress error reports
+
+      # Suppress error reports. On by default for HTML4::Document and HTML4::DocumentFragment
       NOERROR     = 1 << 5
-      # suppress warning reports
+
+      # Suppress warning reports. On by default for HTML4::Document and HTML4::DocumentFragment
       NOWARNING   = 1 << 6
-      # pedantic error reporting
+
+      # Enable pedantic error reporting. Off by default.
       PEDANTIC    = 1 << 7
-      # remove blank nodes
+
+      # Remove blank nodes. Off by default.
       NOBLANKS    = 1 << 8
-      # use the SAX1 interface internally
+
+      # Use the SAX1 interface internally. Off by default.
       SAX1        = 1 << 9
-      # Implement XInclude substitution
+
+      # Implement XInclude substitution. Off by default.
       XINCLUDE    = 1 << 10
-      # Forbid network access. Recommended for dealing with untrusted documents.
+
+      # Forbid network access. On by default for XML::Document, XML::DocumentFragment,
+      # HTML4::Document, HTML4::DocumentFragment, XSLT::Stylesheet, and XML::Schema.
+      #
+      # ⚠ <b>It is UNSAFE to unset this option</b> when parsing untrusted documents.
       NONET       = 1 << 11
-      # Do not reuse the context dictionary
+
+      # Do not reuse the context dictionary. Off by default.
       NODICT      = 1 << 12
-      # remove redundant namespaces declarations
+
+      # Remove redundant namespaces declarations. Off by default.
       NSCLEAN     = 1 << 13
-      # merge CDATA as text nodes
+
+      # Merge CDATA as text nodes. On by default for XSLT::Stylesheet.
       NOCDATA     = 1 << 14
-      # do not generate XINCLUDE START/END nodes
+
+      # Do not generate XInclude START/END nodes. Off by default.
       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 small text nodes. Off by default.
+      #
+      # ⚠ No modification of the DOM tree is allowed after parsing. libxml2 may crash if you try to
+      # modify the tree.
       COMPACT     = 1 << 16
-      # parse using XML-1.0 before update 5
+
+      # Parse using XML-1.0 before update 5. Off by default
       OLD10       = 1 << 17
-      # do not fixup XINCLUDE xml:base uris
+
+      # Do not fixup XInclude xml:base uris. Off by default
       NOBASEFIX   = 1 << 18
-      # relax any hardcoded limit from the parser
+
+      # Relax any hardcoded limit from the parser. Off by default.
+      #
+      # ⚠ There may be a performance penalty when this option is set.
       HUGE        = 1 << 19
-      # line numbers stored as long int (instead of a short int)
+
+      # Support line numbers up to <code>long int</code> (default is a <code>short int</code>). On
+      # by default for for XML::Document, XML::DocumentFragment, HTML4::Document,
+      # HTML4::DocumentFragment, XSLT::Stylesheet, and XML::Schema.
       BIG_LINES   = 1 << 22
 
-      # the default options used for parsing XML documents
+      # The options mask used by default for parsing XML::Document and XML::DocumentFragment
       DEFAULT_XML  = RECOVER | NONET | BIG_LINES
-      # the default options used for parsing XSLT stylesheets
+
+      # The options mask used by default used for parsing XSLT::Stylesheet
       DEFAULT_XSLT = RECOVER | NONET | NOENT | DTDLOAD | DTDATTR | NOCDATA | BIG_LINES
-      # the default options used for parsing HTML documents
+
+      # The options mask used by default used for parsing HTML4::Document and HTML4::DocumentFragment
       DEFAULT_HTML = RECOVER | NOERROR | NOWARNING | NONET | BIG_LINES
-      # the default options used for parsing XML schemas
+
+      # The options mask used by default used for parsing XML::Schema
       DEFAULT_SCHEMA = NONET | BIG_LINES
 
       attr_accessor :options
@@ -90,7 +169,7 @@ module Nokogiri
       constants.each do |constant|
         next if constant.to_sym == :STRICT
 
-        class_eval %{
+        class_eval <<~RUBY, __FILE__, __LINE__ + 1
           def #{constant.downcase}
             @options |= #{constant}
             self
@@ -104,7 +183,7 @@ module Nokogiri
           def #{constant.downcase}?
             #{constant} & @options == #{constant}
           end
-        }
+        RUBY
       end
 
       def strict

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

@@ -5,6 +5,8 @@ module Nokogiri
     # :nodoc: all
     module PP
       module Node
+        COLLECTIONS = [:attribute_nodes, :children]
+
         def inspect
           attributes = inspect_attributes.reject do |x|
             attribute = send(x)
@@ -21,11 +23,11 @@ module Nokogiri
           nice_name = self.class.name.split("::").last
           pp.group(2, "#(#{nice_name}:#{format("0x%x", object_id)} {", "})") do
             pp.breakable
-            attrs = inspect_attributes.map do |t|
+            attrs = inspect_attributes.filter_map do |t|
               [t, send(t)] if respond_to?(t)
-            end.compact.find_all do |x|
+            end.find_all do |x|
               if x.last
-                if [:attribute_nodes, :children].include?(x.first)
+                if COLLECTIONS.include?(x.first)
                   !x.last.empty?
                 else
                   true
@@ -34,7 +36,7 @@ module Nokogiri
             end
 
             pp.seplist(attrs) do |v|
-              if [:attribute_nodes, :children].include?(v.first)
+              if COLLECTIONS.include?(v.first)
                 pp.group(2, "#{v.first.to_s.sub(/_\w+$/, "s")} = [", "]") do
                   pp.breakable
                   pp.seplist(v.last) do |item|

data/lib/nokogiri/xml/processing_instruction.rb

@@ -3,7 +3,8 @@
 module Nokogiri
   module XML
     class ProcessingInstruction < Node
-      def initialize(document, name, content) # rubocop:disable Style/RedundantInitialize
+      def initialize(document, name, content)
+        super(document, name)
       end
     end
   end

data/lib/nokogiri/xml/reader.rb

@@ -83,16 +83,14 @@ module Nokogiri
       end
       private :initialize
 
-      # Get the attributes of the current node as a Hash
+      # Get the attributes and namespaces of the current node as a Hash.
       #
-      # [Returns] (Hash<String, String>) Attribute names and values
+      # This is the union of Reader#attribute_hash and Reader#namespaces
+      #
+      # [Returns]
+      #   (Hash<String, String>) Attribute names and values, and namespace prefixes and hrefs.
       def attributes
-        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
+        attribute_hash.merge(namespaces)
       end
 
       ###

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

@@ -88,9 +88,8 @@ module Nokogiri
 
         ###
         # Parse given +io+
-        def parse_io(io, encoding = "ASCII")
-          @encoding = check_encoding(encoding)
-          ctx = ParserContext.io(io, ENCODINGS[@encoding])
+        def parse_io(io, encoding = @encoding)
+          ctx = ParserContext.io(io, ENCODINGS[check_encoding(encoding)])
           yield ctx if block_given?
           ctx.parse_with(self)
         end

data/lib/nokogiri/xslt.rb

@@ -47,7 +47,7 @@ module Nokogiri
       # [Returns] Array of string parameters, with quotes correctly escaped for use with XSLT::Stylesheet.transform
       #
       def quote_params(params)
-        params.flatten.each_slice(2).each_with_object([]) do |kv, quoted_params|
+        params.flatten.each_slice(2).with_object([]) do |kv, quoted_params|
           key, value = kv.map(&:to_s)
           value = if /'/.match?(value)
             "concat('#{value.gsub(/'/, %q{', "'", '})}')"

data/lib/nokogiri.rb

@@ -1,8 +1,6 @@
 # coding: utf-8
 # frozen_string_literal: true
 
-require "rbconfig"
-
 if defined?(RUBY_ENGINE) && RUBY_ENGINE == "jruby"
   require_relative "nokogiri/jruby/dependencies"
 end
@@ -88,17 +86,10 @@ module Nokogiri
 
     # :nodoc:
     def install_default_aliases
-      # Make sure to support some popular encoding aliases not known by
-      # all iconv implementations.
-      {
-        "Windows-31J" => "CP932",	# Windows-31J is the IANA registered name of CP932.
-      }.each do |alias_name, name|
-        EncodingHandler.alias(name, alias_name) if EncodingHandler[alias_name].nil?
-      end
+      warn("Nokogiri.install_default_aliases is deprecated and will be removed in a future version of Nokogiri. Please call Nokogiri::EncodingHandler.install_default_aliases instead.")
+      Nokogiri::EncodingHandler.install_default_aliases
     end
   end
-
-  Nokogiri.install_default_aliases
 end
 
 ###
@@ -124,5 +115,6 @@ require_relative "nokogiri/html"
 require_relative "nokogiri/decorators/slop"
 require_relative "nokogiri/css"
 require_relative "nokogiri/html4/builder"
+require_relative "nokogiri/encoding_handler"
 
 require_relative "nokogiri/html5" if Nokogiri.uses_gumbo?

data/lib/xsd/xmlparser/nokogiri.rb

@@ -95,7 +95,9 @@ module XSD
       end
 
       ["xmldecl", "start_document", "end_document", "comment"].each do |name|
-        class_eval %{ def #{name}(*args); end }
+        class_eval <<~RUBY, __FILE__, __LINE__ + 1
+          def #{name}(*args); end
+        RUBY
       end
 
       add_factory(self)