nokogiri 1.14.3 → 1.15.0

data/lib/nokogiri/html5/document_fragment.rb

@@ -36,7 +36,7 @@ module Nokogiri
       attr_reader :quirks_mode
 
       # Create a document fragment.
-      def initialize(doc, tags = nil, ctx = nil, options = {})
+      def initialize(doc, tags = nil, ctx = nil, options = {}) # rubocop:disable Lint/MissingSuper
         self.document = doc
         self.errors = []
         return self unless tags

data/lib/nokogiri/html5/node.rb

@@ -17,6 +17,9 @@
 #  limitations under the License.
 #
 
+#
+#  TODO: this whole file should go away. maybe make it a decorator?
+#
 require_relative "../xml/node"
 
 module Nokogiri
@@ -50,6 +53,8 @@ module Nokogiri
         config = XML::Node::SaveOptions.new(save_options.to_i)
         yield config if block_given?
 
+        encoding = encoding.is_a?(Encoding) ? encoding.name : encoding
+
         config_options = config.options
         if config_options & (XML::Node::SaveOptions::AS_XML | XML::Node::SaveOptions::AS_XHTML) != 0
           # Use Nokogiri's serializing code.

data/lib/nokogiri/html5.rb

@@ -248,8 +248,11 @@ module Nokogiri
       #  * :basic_auth => [username, password]
       def get(uri, options = {})
         # TODO: deprecate
-        warn("Nokogiri::HTML5.get is deprecated and will be removed in a future version of Nokogiri.",
-          uplevel: 1, category: :deprecated)
+        warn(
+          "Nokogiri::HTML5.get is deprecated and will be removed in a future version of Nokogiri.",
+          uplevel: 1,
+          category: :deprecated,
+        )
         get_impl(uri, options)
       end
 

data/lib/nokogiri/jruby/nokogiri_jars.rb

@@ -2,26 +2,26 @@
 begin
   require 'jar_dependencies'
 rescue LoadError
-  require 'xalan/xalan/2.7.2/xalan-2.7.2.jar'
+  require 'xalan/serializer/2.7.3/serializer-2.7.3.jar'
   require 'net/sourceforge/htmlunit/neko-htmlunit/2.63.0/neko-htmlunit-2.63.0.jar'
   require 'nu/validator/jing/20200702VNU/jing-20200702VNU.jar'
   require 'xerces/xercesImpl/2.12.2/xercesImpl-2.12.2.jar'
   require 'org/nokogiri/nekodtd/0.1.11.noko1/nekodtd-0.1.11.noko1.jar'
   require 'net/sf/saxon/Saxon-HE/9.6.0-4/Saxon-HE-9.6.0-4.jar'
+  require 'xalan/xalan/2.7.3/xalan-2.7.3.jar'
   require 'xml-apis/xml-apis/1.4.01/xml-apis-1.4.01.jar'
-  require 'xalan/serializer/2.7.2/serializer-2.7.2.jar'
   require 'isorelax/isorelax/20030108/isorelax-20030108.jar'
 end
 
 if defined? Jars
-  require_jar 'xalan', 'xalan', '2.7.2'
+  require_jar 'xalan', 'serializer', '2.7.3'
   require_jar 'net.sourceforge.htmlunit', 'neko-htmlunit', '2.63.0'
   require_jar 'nu.validator', 'jing', '20200702VNU'
   require_jar 'xerces', 'xercesImpl', '2.12.2'
   require_jar 'org.nokogiri', 'nekodtd', '0.1.11.noko1'
   require_jar 'net.sf.saxon', 'Saxon-HE', '9.6.0-4'
+  require_jar 'xalan', 'xalan', '2.7.3'
   require_jar 'xml-apis', 'xml-apis', '1.4.01'
-  require_jar 'xalan', 'serializer', '2.7.2'
   require_jar 'isorelax', 'isorelax', '20030108'
 end
 
@@ -33,8 +33,8 @@ module Nokogiri
     "net.sourceforge.htmlunit:neko-htmlunit" => "2.63.0",
     "nu.validator:jing" => "20200702VNU",
     "org.nokogiri:nekodtd" => "0.1.11.noko1",
-    "xalan:serializer" => "2.7.2",
-    "xalan:xalan" => "2.7.2",
+    "xalan:serializer" => "2.7.3",
+    "xalan:xalan" => "2.7.3",
     "xerces:xercesImpl" => "2.12.2",
     "xml-apis:xml-apis" => "1.4.01",
   }.freeze

data/lib/nokogiri/version/constant.rb

@@ -2,5 +2,5 @@
 
 module Nokogiri
   # The version of Nokogiri you are using
-  VERSION = "1.14.3"
+  VERSION = "1.15.0"
 end

data/lib/nokogiri/xml/attribute_decl.rb

@@ -12,8 +12,10 @@ module Nokogiri
       undef_method :namespace_definitions
       undef_method :line if method_defined?(:line)
 
-      def inspect
-        "#<#{self.class.name}:#{format("0x%x", object_id)} #{to_s.inspect}>"
+      private
+
+      def inspect_attributes
+        [:to_s]
       end
     end
   end

data/lib/nokogiri/xml/document_fragment.rb

@@ -16,7 +16,7 @@ module Nokogiri
       #  If +ctx+ is present, it is used as a context node for the
       #  subtree created, e.g., namespaces will be resolved relative
       #  to +ctx+.
-      def initialize(document, tags = nil, ctx = nil, options = ParseOptions::DEFAULT_XML)
+      def initialize(document, tags = nil, ctx = nil, options = ParseOptions::DEFAULT_XML) # rubocop:disable Lint/MissingSuper
         return self unless tags
 
         options = Nokogiri::XML::ParseOptions.new(options) if Integer === options

data/lib/nokogiri/xml/element_content.rb

@@ -11,9 +11,11 @@ module Nokogiri
     #   ]>
     #   </root>
     #
-    # ElementContent represents the tree inside the <!ELEMENT> tag shown above
-    # that lists the possible content for the div1 tag.
+    # ElementContent represents the binary tree inside the <!ELEMENT> tag shown above that lists the
+    # possible content for the div1 tag.
     class ElementContent
+      include Nokogiri::XML::PP::Node
+
       # Possible definitions of type
       PCDATA  = 1
       ELEMENT = 2
@@ -33,6 +35,12 @@ module Nokogiri
       def children
         [c1, c2].compact
       end
+
+      private
+
+      def inspect_attributes
+        [:prefix, :name, :type, :occur, :children]
+      end
     end
   end
 end

data/lib/nokogiri/xml/element_decl.rb

@@ -7,8 +7,10 @@ module Nokogiri
       undef_method :namespace_definitions
       undef_method :line if method_defined?(:line)
 
-      def inspect
-        "#<#{self.class.name}:#{format("0x%x", object_id)} #{to_s.inspect}>"
+      private
+
+      def inspect_attributes
+        [:to_s]
       end
     end
   end

data/lib/nokogiri/xml/entity_decl.rb

@@ -13,8 +13,10 @@ module Nokogiri
         doc.create_entity(name, *args)
       end
 
-      def inspect
-        "#<#{self.class.name}:#{format("0x%x", object_id)} #{to_s.inspect}>"
+      private
+
+      def inspect_attributes
+        [:to_s]
       end
     end
   end

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

@@ -62,6 +62,14 @@ module Nokogiri
         end
 
         alias_method :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/node.rb

@@ -1269,11 +1269,11 @@ module Nokogiri
       #
       # These two statements are equivalent:
       #
-      #  node.serialize(:encoding => 'UTF-8', :save_with => FORMAT | AS_XML)
+      #   node.serialize(encoding: 'UTF-8', save_with: FORMAT | AS_XML)
       #
       # or
       #
-      #   node.serialize(:encoding => 'UTF-8') do |config|
+      #   node.serialize(encoding: 'UTF-8') do |config|
       #     config.format.as_xml
       #   end
       #
@@ -1310,7 +1310,7 @@ module Nokogiri
       ###
       # Serialize this Node to XML using +options+
       #
-      #   doc.to_xml(:indent => 5, :encoding => 'UTF-8')
+      #   doc.to_xml(indent: 5, encoding: 'UTF-8')
       #
       # See Node#write_to for a list of +options+
       def to_xml(options = {})
@@ -1321,7 +1321,7 @@ module Nokogiri
       ###
       # Serialize this Node to XHTML using +options+
       #
-      #   doc.to_xhtml(:indent => 5, :encoding => 'UTF-8')
+      #   doc.to_xhtml(indent: 5, encoding: 'UTF-8')
       #
       # See Node#write_to for a list of +options+
       def to_xhtml(options = {})
@@ -1329,25 +1329,32 @@ module Nokogiri
       end
 
       ###
-      # Write Node to +io+ with +options+. +options+ modify the output of
-      # this method.  Valid options are:
+      # :call-seq:
+      #   write_to(io, *options)
+      #
+      # Serialize this node or document to +io+.
       #
-      # * +:encoding+ for changing the encoding
-      # * +:indent_text+ the indentation text, defaults to one space
-      # * +:indent+ the number of +:indent_text+ to use, defaults to 2
-      # * +:save_with+ a combination of SaveOptions constants.
+      # [Parameters]
+      # - +io+ (IO) An IO-like object to which the serialized content will be written.
+      # - +options+ (Hash) See below
+      #
+      # [Options]
+      # * +:encoding+ (String or Encoding) specify the encoding of the output (defaults to document encoding)
+      # * +:indent_text+ (String) the indentation text (defaults to <code>" "</code>)
+      # * +:indent+ (Integer) the number of +:indent_text+ to use (defaults to +2+)
+      # * +:save_with+ (Integer) a combination of SaveOptions constants
       #
       # To save with UTF-8 indented twice:
       #
-      #   node.write_to(io, :encoding => 'UTF-8', :indent => 2)
+      #   node.write_to(io, encoding: 'UTF-8', indent: 2)
       #
       # To save indented with two dashes:
       #
-      #   node.write_to(io, :indent_text => '-', :indent => 2)
+      #   node.write_to(io, indent_text: '-', indent: 2)
       #
       def write_to(io, *options)
         options = options.first.is_a?(Hash) ? options.shift : {}
-        encoding = options[:encoding] || options[0]
+        encoding = options[:encoding] || options[0] || document.encoding
         if Nokogiri.jruby?
           save_options = options[:save_with] || options[1]
           indent_times = options[:indent] || 0
@@ -1365,6 +1372,8 @@ module Nokogiri
         config = SaveOptions.new(save_options.to_i)
         yield config if block_given?
 
+        encoding = encoding.is_a?(Encoding) ? encoding.name : encoding
+
         native_write_to(io, encoding, indentation, config.options)
       end
 

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

@@ -13,9 +13,14 @@ module Nokogiri
             !attribute || (attribute.respond_to?(:empty?) && attribute.empty?)
           rescue NoMethodError
             true
-          end.map do |attribute|
-            "#{attribute.to_s.sub(/_\w+/, "s")}=#{send(attribute).inspect}"
-          end.join(" ")
+          end
+          attributes = if inspect_attributes.length == 1
+            send(attributes.first).inspect
+          else
+            attributes.map do |attribute|
+              "#{attribute}=#{send(attribute).inspect}"
+            end.join(" ")
+          end
           "#<#{self.class.name}:#{format("0x%x", object_id)} #{attributes}>"
         end
 
@@ -23,6 +28,7 @@ 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.filter_map do |t|
               [t, send(t)] if respond_to?(t)
             end.find_all do |x|
@@ -35,19 +41,24 @@ module Nokogiri
               end
             end
 
-            pp.seplist(attrs) do |v|
-              if COLLECTIONS.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)
+            if inspect_attributes.length == 1
+              pp.pp(attrs.first.last)
+            else
+              pp.seplist(attrs) do |v|
+                if COLLECTIONS.include?(v.first)
+                  pp.group(2, "#{v.first} = [", "]") do
+                    pp.breakable
+                    pp.seplist(v.last) do |item|
+                      pp.pp(item)
+                    end
                   end
+                else
+                  pp.text("#{v.first} = ")
+                  pp.pp(v.last)
                 end
-              else
-                pp.text("#{v.first} = ")
-                pp.pp(v.last)
               end
             end
+
             pp.breakable
           end
         end

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

@@ -100,7 +100,7 @@ module Nokogiri
         # +prefix+ is the namespace prefix for the element
         # +uri+ is the associated namespace URI
         # +ns+ is a hash of namespace prefix:urls associated with the element
-        def start_element_namespace(name, attrs = [], prefix = nil, uri = nil, ns = [])
+        def start_element_namespace(name, attrs = [], prefix = nil, uri = nil, ns = []) # rubocop:disable Metrics/ParameterLists
           ###
           # Deal with SAX v1 interface
           name = [prefix, name].compact.join(":")

data/lib/nokogiri/xml/searchable.rb

@@ -36,16 +36,19 @@ module Nokogiri
       #   node.search('.//address[@domestic=$value]', nil, {:value => 'Yes'})
       #
       # 💡 Custom XPath functions and CSS pseudo-selectors may also be defined. To define custom
-      # functions create a class and implement the function you want to define. The first argument
-      # to the method will be the current matching NodeSet. Any other arguments are ones that you
-      # pass in. Note that this class may appear anywhere in the argument list. For example:
+      # functions create a class and implement the function you want to define, which will be in the
+      # `nokogiri` namespace in XPath queries.
+      #
+      # The first argument to the method will be the current matching NodeSet. Any other arguments
+      # are ones that you pass in. Note that this class may appear anywhere in the argument
+      # list. For example:
       #
       #   handler = Class.new {
       #     def regex node_set, regex
       #       node_set.find_all { |node| node['some_attribute'] =~ /#{regex}/ }
       #     end
       #   }.new
-      #   node.search('.//title[regex(., "\w+")]', 'div.employee:regex("[0-9]+")', handler)
+      #   node.search('.//title[nokogiri:regex(., "\w+")]', 'div.employee:regex("[0-9]+")', handler)
       #
       # See Searchable#xpath and Searchable#css for further usage help.
       def search(*args)
@@ -160,16 +163,18 @@ module Nokogiri
       #   node.xpath('.//address[@domestic=$value]', nil, {:value => 'Yes'})
       #
       # 💡 Custom XPath functions may also be defined. To define custom functions create a class and
-      # implement the function you want to define. The first argument to the method will be the
-      # current matching NodeSet. Any other arguments are ones that you pass in. Note that this
-      # class may appear anywhere in the argument list. For example:
+      # implement the function you want to define, which will be in the `nokogiri` namespace.
+      #
+      # The first argument to the method will be the current matching NodeSet. Any other arguments
+      # are ones that you pass in. Note that this class may appear anywhere in the argument
+      # list. For example:
       #
       #   handler = Class.new {
       #     def regex(node_set, regex)
       #       node_set.find_all { |node| node['some_attribute'] =~ /#{regex}/ }
       #     end
       #   }.new
-      #   node.xpath('.//title[regex(., "\w+")]', handler)
+      #   node.xpath('.//title[nokogiri:regex(., "\w+")]', handler)
       #
       def xpath(*args)
         paths, handler, ns, binds = extract_params(args)
@@ -243,8 +248,11 @@ module Nokogiri
           doctype: document.xpath_doctype,
         )
         self.class::IMPLIED_XPATH_CONTEXTS.map do |implied_xpath_context|
-          CSS.xpath_for(rule.to_s, { prefix: implied_xpath_context, ns: ns,
-                                     visitor: visitor, })
+          CSS.xpath_for(rule.to_s, {
+            prefix: implied_xpath_context,
+            ns: ns,
+            visitor: visitor,
+          })
         end.join(" | ")
       end
 

data/lib/nokogiri/xslt.rb

@@ -20,8 +20,59 @@ module Nokogiri
   # Stylesheet object.
   module XSLT
     class << self
-      ###
-      # Parse the stylesheet in +string+, register any +modules+
+      # :call-seq:
+      #   parse(xsl) → Nokogiri::XSLT::Stylesheet
+      #   parse(xsl, modules) → Nokogiri::XSLT::Stylesheet
+      #
+      # Parse the stylesheet in +xsl+, registering optional +modules+ as custom class handlers.
+      #
+      # [Parameters]
+      # - +xsl+ (String) XSL content to be parsed into a stylesheet
+      # - +modules+ (Hash<String ⇒ Class>) A hash of URI-to-handler relations for linking a
+      #   namespace to a custom function handler.
+      #
+      # ⚠ The XSLT handler classes are registered *globally*.
+      #
+      # Also see Nokogiri::XSLT.register
+      #
+      # *Example*
+      #
+      #   xml = Nokogiri.XML(<<~XML)
+      #     <nodes>
+      #       <node>Foo</node>
+      #       <node>Bar</node>
+      #     </nodes>
+      #   XML
+      #
+      #   handler = Class.new do
+      #     def reverse(node)
+      #       node.text.reverse
+      #     end
+      #   end
+      #
+      #   xsl = <<~XSL
+      #     <xsl:stylesheet version="1.0"
+      #       xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+      #       xmlns:myfuncs="http://nokogiri.org/xslt/myfuncs"
+      #       extension-element-prefixes="myfuncs">
+      #       <xsl:template match="/">
+      #         <reversed>
+      #           <xsl:for-each select="nodes/node">
+      #             <reverse><xsl:copy-of select="myfuncs:reverse(.)"/></reverse>
+      #           </xsl:for-each>
+      #         </reversed>
+      #       </xsl:template>
+      #     </xsl:stylesheet>
+      #   XSL
+      #
+      #   xsl = Nokogiri.XSLT(xsl, "http://nokogiri.org/xslt/myfuncs" => handler)
+      #   xsl.transform(xml).to_xml
+      #   # => "<?xml version=\"1.0\"?>\n" +
+      #   #    "<reversed>\n" +
+      #   #    "  <reverse>ooF</reverse>\n" +
+      #   #    "  <reverse>raB</reverse>\n" +
+      #   #    "</reversed>\n"
+      #
       def parse(string, modules = {})
         modules.each do |url, klass|
           XSLT.register(url, klass)
@@ -49,7 +100,7 @@ module Nokogiri
       def quote_params(params)
         params.flatten.each_slice(2).with_object([]) do |kv, quoted_params|
           key, value = kv.map(&:to_s)
-          value = if /'/.match?(value)
+          value = if value.include?("'")
             "concat('#{value.gsub(/'/, %q{', "'", '})}')"
           else
             "'#{value}'"
@@ -58,6 +109,25 @@ module Nokogiri
           quoted_params << value
         end
       end
+
+      #  call-seq:
+      #    register(uri, custom_handler_class)
+      #
+      #  Register a class that implements custom XSLT transformation functions.
+      #
+      #  ⚠ The XSLT handler classes are registered *globally*.
+      #
+      #  [Parameters}
+      #  - +uri+ (String) The namespace for the custom handlers
+      #  - +custom_handler_class+ (Class) A class with ruby methods that can be called during
+      #    transformation
+      #
+      #  See Nokogiri::XSLT.parse for usage.
+      #
+      def register(uri, custom_handler_class)
+        # NOTE: this is implemented in the C extension, see ext/nokogiri/xslt_stylesheet.c
+        raise NotImplementedError, "Nokogiri::XSLT.register is not implemented on JRuby"
+      end if Nokogiri.jruby?
     end
   end
 end

data/lib/nokogiri.rb

@@ -45,11 +45,19 @@ module Nokogiri
         # Expect an HTML indicator to appear within the first 512
         # characters of a document. (<?xml ?> + <?xml-stylesheet ?>
         # shouldn't be that long)
-        Nokogiri.HTML4(string, url, encoding,
-          options || XML::ParseOptions::DEFAULT_HTML)
+        Nokogiri.HTML4(
+          string,
+          url,
+          encoding,
+          options || XML::ParseOptions::DEFAULT_HTML,
+        )
       else
-        Nokogiri.XML(string, url, encoding,
-          options || XML::ParseOptions::DEFAULT_XML)
+        Nokogiri.XML(
+          string,
+          url,
+          encoding,
+          options || XML::ParseOptions::DEFAULT_XML,
+        )
       end.tap do |doc|
         yield doc if block_given?
       end

data/lib/xsd/xmlparser/nokogiri.rb

@@ -71,7 +71,7 @@ module XSD
       # +prefix+ is the namespace prefix for the element
       # +uri+ is the associated namespace URI
       # +ns+ is a hash of namespace prefix:urls associated with the element
-      def start_element_namespace(name, attrs = [], prefix = nil, uri = nil, ns = [])
+      def start_element_namespace(name, attrs = [], prefix = nil, uri = nil, ns = []) # rubocop:disable Metrics/ParameterLists
         ###
         # Deal with SAX v1 interface
         name = [prefix, name].compact.join(":")