nokogiri 1.11.3 → 1.13.8

data/lib/nokogiri.rb

@@ -1,75 +1,67 @@
-# -*- coding: utf-8 -*-
+# coding: utf-8
 # frozen_string_literal: true
-# Modify the PATH on windows so that the external DLLs will get loaded.
 
-require 'rbconfig'
+require "rbconfig"
 
 if defined?(RUBY_ENGINE) && RUBY_ENGINE == "jruby"
-  require 'nokogiri/jruby/dependencies'
+  require_relative "nokogiri/jruby/dependencies"
 end
 
-require 'nokogiri/extension'
-
-require 'nokogiri/version'
-require 'nokogiri/syntax_error'
-require 'nokogiri/xml'
-require 'nokogiri/xslt'
-require 'nokogiri/html'
-require 'nokogiri/decorators/slop'
-require 'nokogiri/css'
-require 'nokogiri/html/builder'
+require_relative "nokogiri/extension"
 
 # Nokogiri parses and searches XML/HTML very quickly, and also has
 # correctly implemented CSS3 selector support as well as XPath 1.0
 # support.
 #
 # Parsing a document returns either a Nokogiri::XML::Document, or a
-# Nokogiri::HTML::Document depending on the kind of document you parse.
+# Nokogiri::HTML4::Document depending on the kind of document you parse.
 #
 # Here is an example:
 #
-#   require 'nokogiri'
-#   require 'open-uri'
+#     require 'nokogiri'
+#     require 'open-uri'
+#
+#     # Get a Nokogiri::HTML4::Document for the page we’re interested in...
 #
-#   # Get a Nokogiri::HTML:Document for the page we’re interested in...
+#     doc = Nokogiri::HTML4(URI.open('http://www.google.com/search?q=tenderlove'))
 #
-#   doc = Nokogiri::HTML(URI.open('http://www.google.com/search?q=tenderlove'))
+#     # Do funky things with it using Nokogiri::XML::Node methods...
 #
-#   # Do funky things with it using Nokogiri::XML::Node methods...
+#     ####
+#     # Search for nodes by css
+#     doc.css('h3.r a.l').each do |link|
+#       puts link.content
+#     end
 #
-#   ####
-#   # Search for nodes by css
-#   doc.css('h3.r a.l').each do |link|
-#     puts link.content
-#   end
+# See also:
 #
-# See Nokogiri::XML::Searchable#css for more information about CSS searching.
-# See Nokogiri::XML::Searchable#xpath for more information about XPath searching.
+# - Nokogiri::XML::Searchable#css for more information about CSS searching
+# - Nokogiri::XML::Searchable#xpath for more information about XPath searching
 module Nokogiri
   class << self
     ###
     # Parse an HTML or XML document.  +string+ contains the document.
-    def parse string, url = nil, encoding = nil, options = nil
+    def parse(string, url = nil, encoding = nil, options = nil)
       if string.respond_to?(:read) ||
-          /^\s*<(?:!DOCTYPE\s+)?html[\s>]/i === string[0, 512]
+          /^\s*<(?:!DOCTYPE\s+)?html[\s>]/i.match?(string[0, 512])
         # Expect an HTML indicator to appear within the first 512
         # characters of a document. (<?xml ?> + <?xml-stylesheet ?>
         # shouldn't be that long)
-        Nokogiri.HTML(string, url, encoding,
+        Nokogiri.HTML4(string, url, encoding,
           options || XML::ParseOptions::DEFAULT_HTML)
       else
         Nokogiri.XML(string, url, encoding,
           options || XML::ParseOptions::DEFAULT_XML)
-      end.tap { |doc|
+      end.tap do |doc|
         yield doc if block_given?
-      }
+      end
     end
 
     ###
     # Create a new Nokogiri::XML::DocumentFragment
-    def make input = nil, opts = {}, &blk
+    def make(input = nil, opts = {}, &blk)
       if input
-        Nokogiri::HTML.fragment(input).children.first
+        Nokogiri::HTML4.fragment(input).children.first
       else
         Nokogiri(&blk)
       end
@@ -94,14 +86,15 @@ module Nokogiri
       Nokogiri(*args, &block).slop!
     end
 
+    # :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 { |alias_name, name|
+        "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
     end
   end
 
@@ -109,15 +102,27 @@ module Nokogiri
 end
 
 ###
-# Parse a document contained in +args+.  Nokogiri will try to guess what
-# type of document you are attempting to parse.  For more information, see
-# Nokogiri.parse
+# Parse a document contained in +args+.  Nokogiri will try to guess what type of document you are
+# attempting to parse.  For more information, see Nokogiri.parse
 #
-# To specify the type of document, use Nokogiri.XML or Nokogiri.HTML.
+# To specify the type of document, use {Nokogiri.XML}, {Nokogiri.HTML4}, or {Nokogiri.HTML5}.
 def Nokogiri(*args, &block)
-  if block_given?
-    Nokogiri::HTML::Builder.new(&block).doc.root
+  if block
+    Nokogiri::HTML4::Builder.new(&block).doc.root
   else
     Nokogiri.parse(*args)
   end
 end
+
+require_relative "nokogiri/version"
+require_relative "nokogiri/class_resolver"
+require_relative "nokogiri/syntax_error"
+require_relative "nokogiri/xml"
+require_relative "nokogiri/xslt"
+require_relative "nokogiri/html4"
+require_relative "nokogiri/html"
+require_relative "nokogiri/decorators/slop"
+require_relative "nokogiri/css"
+require_relative "nokogiri/html4/builder"
+
+require_relative "nokogiri/html5" if Nokogiri.uses_gumbo?

data/lib/xsd/xmlparser/nokogiri.rb

@@ -1,8 +1,9 @@
 # frozen_string_literal: true
-require 'nokogiri'
 
-module XSD # :nodoc:
-  module XMLParser # :nodoc:
+require "nokogiri"
+
+module XSD
+  module XMLParser
     ###
     # Nokogiri XML parser for soap4r.
     #
@@ -27,40 +28,40 @@ module XSD # :nodoc:
     class Nokogiri < XSD::XMLParser::Parser
       ###
       # Create a new XSD parser with +host+ and +opt+
-      def initialize host, opt = {}
+      def initialize(host, opt = {})
         super
-        @parser = ::Nokogiri::XML::SAX::Parser.new(self, @charset || 'UTF-8')
+        @parser = ::Nokogiri::XML::SAX::Parser.new(self, @charset || "UTF-8")
       end
 
       ###
       # Start parsing +string_or_readable+
-      def do_parse string_or_readable
+      def do_parse(string_or_readable)
         @parser.parse(string_or_readable)
       end
 
       ###
       # Handle the start_element event with +name+ and +attrs+
-      def start_element name, attrs = []
+      def start_element(name, attrs = [])
         super(name, Hash[*attrs.flatten])
       end
 
       ###
       # Handle the end_element event with +name+
-      def end_element name
+      def end_element(name)
         super
       end
 
       ###
       # Handle errors with message +msg+
-      def error msg
-        raise ParseError.new(msg)
+      def error(msg)
+        raise ParseError, msg
       end
-      alias :warning :error
+      alias_method :warning, :error
 
       ###
       # Handle cdata_blocks containing +string+
-      def cdata_block string
-        characters string
+      def cdata_block(string)
+        characters(string)
       end
 
       ###
@@ -70,16 +71,16 @@ module XSD # :nodoc:
       # +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 = [])
         ###
         # Deal with SAX v1 interface
-        name = [prefix, name].compact.join(':')
-        attributes = ns.map { |ns_prefix,ns_uri|
-          [['xmlns', ns_prefix].compact.join(':'), ns_uri]
-        } + attrs.map { |attr|
-          [[attr.prefix, attr.localname].compact.join(':'), attr.value]
-        }.flatten
-        start_element name, attributes
+        name = [prefix, name].compact.join(":")
+        attributes = ns.map do |ns_prefix, ns_uri|
+          [["xmlns", ns_prefix].compact.join(":"), ns_uri]
+        end + attrs.map do |attr|
+          [[attr.prefix, attr.localname].compact.join(":"), attr.value]
+        end.flatten
+        start_element(name, attributes)
       end
 
       ###
@@ -87,13 +88,13 @@ module XSD # :nodoc:
       # +name+ is the element's name
       # +prefix+ is the namespace prefix associated with the element
       # +uri+ is the associated namespace URI
-      def end_element_namespace name, prefix = nil, uri = nil
+      def end_element_namespace(name, prefix = nil, uri = nil)
         ###
         # Deal with SAX v1 interface
-        end_element [prefix, name].compact.join(':')
+        end_element([prefix, name].compact.join(":"))
       end
 
-      %w{ xmldecl start_document end_document comment }.each do |name|
+      ["xmldecl", "start_document", "end_document", "comment"].each do |name|
         class_eval %{ def #{name}(*args); end }
       end
 

data/patches/libxml2/{0004-libxml2.la-is-in-top_builddir.patch → 0003-libxml2.la-is-in-top_builddir.patch}

@@ -16,7 +16,7 @@ index cf96d41..1372d8b 100644
  	}
  
 -libxml2.la: $(libxml2_la_OBJECTS) $(libxml2_la_DEPENDENCIES) $(EXTRA_libxml2_la_DEPENDENCIES) 
-+$(top_builddir)/libxml2.la: $(libxml2_la_OBJECTS) $(libxml2_la_DEPENDENCIES) $(EXTRA_libxml2_la_DEPENDENCIES) 
++$(top_builddir)/libxml2.la: $(libxml2_la_OBJECTS) $(libxml2_la_DEPENDENCIES) $(EXTRA_libxml2_la_DEPENDENCIES)
  	$(AM_V_CCLD)$(libxml2_la_LINK) -rpath $(libdir) $(libxml2_la_OBJECTS) $(libxml2_la_LIBADD) $(LIBS)
  
  testdso.la: $(testdso_la_OBJECTS) $(testdso_la_DEPENDENCIES) $(EXTRA_testdso_la_DEPENDENCIES) 

data/patches/libxml2/{0008-use-glibc-strlen.patch → 0004-use-glibc-strlen.patch}

@@ -31,18 +31,18 @@ diff --git a/xmlstring.c b/xmlstring.c
 index e8a1e45d..df247dff 100644
 --- a/xmlstring.c
 +++ b/xmlstring.c
-@@ -423,14 +423,9 @@ xmlStrsub(const xmlChar *str, int start, int len) {
+@@ -423,12 +423,7 @@ xmlStrsub(const xmlChar *str, int start, int len) {
 
  int
  xmlStrlen(const xmlChar *str) {
--    int len = 0;
+-    size_t len = 0;
 -
      if (str == NULL) return(0);
 -    while (*str != 0) { /* non input consuming */
 -        str++;
 -        len++;
 -    }
--    return(len);
+-    return(len > INT_MAX ? 0 : len);
 +
 +    return strlen((const char*)str);
  }

data/patches/libxml2/{0009-avoid-isnan-isinf.patch → 0005-avoid-isnan-isinf.patch}

@@ -51,7 +51,7 @@ diff --git a/xpath.c b/xpath.c
 index 9f64ab9..5b6d999 100644
 --- a/xpath.c
 +++ b/xpath.c
-@@ -509,11 +509,7 @@ xmlXPathInit(void) {
+@@ -515,11 +515,7 @@ xmlXPathInit(void) {
   */
  int
  xmlXPathIsNaN(double val) {
@@ -63,16 +63,16 @@ index 9f64ab9..5b6d999 100644
  }
  
  /**
-@@ -524,15 +520,11 @@ xmlXPathIsNaN(double val) {
+@@ -530,15 +530,11 @@ xmlXPathIsNaN(double val) {
   */
  int
  xmlXPathIsInf(double val) {
 -#ifdef isinf
 -    return isinf(val) ? (val > 0 ? 1 : -1) : 0;
 -#else
-     if (val >= INFINITY)
+     if (val >= xmlXPathPINF)
          return 1;
-     if (val <= -INFINITY)
+     if (val <= -xmlXPathPINF)
          return -1;
      return 0;
 -#endif