nokogiri 1.10.10-java → 1.11.0.rc1-java

data/ext/nokogiri/extconf.rb

@@ -486,50 +486,60 @@ else
           url: "http://zlib.net/fossils/#{recipe.name}-#{recipe.version}.tar.gz",
           sha256: dependencies["zlib"]["sha256"]
         }]
-      class << recipe
-        attr_accessor :cross_build_p
-
-        def configure
-          Dir.chdir work_path do
-            mk = File.read 'win32/Makefile.gcc'
-            File.open 'win32/Makefile.gcc', 'wb' do |f|
-              f.puts "BINARY_PATH = #{path}/bin"
-              f.puts "LIBRARY_PATH = #{path}/lib"
-              f.puts "INCLUDE_PATH = #{path}/include"
-              mk.sub!(/^PREFIX\s*=\s*$/, "PREFIX = #{host}-") if cross_build_p
-              f.puts mk
+      if windows?
+        class << recipe
+          attr_accessor :cross_build_p
+
+          def configure
+            Dir.chdir work_path do
+              mk = File.read 'win32/Makefile.gcc'
+              File.open 'win32/Makefile.gcc', 'wb' do |f|
+                f.puts "BINARY_PATH = #{path}/bin"
+                f.puts "LIBRARY_PATH = #{path}/lib"
+                f.puts "INCLUDE_PATH = #{path}/include"
+                mk.sub!(/^PREFIX\s*=\s*$/, "PREFIX = #{host}-") if cross_build_p
+                f.puts mk
+              end
             end
           end
-        end
 
-        def configured?
-          Dir.chdir work_path do
-            !! (File.read('win32/Makefile.gcc') =~ /^BINARY_PATH/)
+          def configured?
+            Dir.chdir work_path do
+              !! (File.read('win32/Makefile.gcc') =~ /^BINARY_PATH/)
+            end
           end
-        end
 
-        def compile
-          execute "compile", "make -f win32/Makefile.gcc"
-        end
+          def compile
+            execute "compile", "make -f win32/Makefile.gcc"
+          end
 
-        def install
-          execute "install", "make -f win32/Makefile.gcc install"
+          def install
+            execute "install", "make -f win32/Makefile.gcc install"
+          end
+        end
+        recipe.cross_build_p = cross_build_p
+      else
+        class << recipe
+          def configure
+            execute "configure", ["env", "CHOST=#{host}", "CFLAGS=-fPIC #{ENV['CFLAGS']}", "./configure", "--static", configure_prefix]
+          end
         end
       end
-      recipe.cross_build_p = cross_build_p
     end
 
-    libiconv_recipe = process_recipe("libiconv", dependencies["libiconv"]["version"], static_p, cross_build_p) do |recipe|
-      recipe.files = [{
-          url: "http://ftp.gnu.org/pub/gnu/libiconv/#{recipe.name}-#{recipe.version}.tar.gz",
-          sha256: dependencies["libiconv"]["sha256"]
-        }]
-      recipe.configure_options += [
-        "CPPFLAGS=-Wall",
-        "CFLAGS=-O2 -g",
-        "CXXFLAGS=-O2 -g",
-        "LDFLAGS="
-      ]
+    unless nix?
+      libiconv_recipe = process_recipe("libiconv", dependencies["libiconv"]["version"], static_p, cross_build_p) do |recipe|
+        recipe.files = [{
+            url: "http://ftp.gnu.org/pub/gnu/libiconv/#{recipe.name}-#{recipe.version}.tar.gz",
+            sha256: dependencies["libiconv"]["sha256"]
+          }]
+        recipe.configure_options += [
+          "CPPFLAGS=-Wall",
+          "CFLAGS=-O2 -g",
+          "CXXFLAGS=-O2 -g",
+          "LDFLAGS="
+        ]
+      end
     end
   else
     if darwin? && !have_header('iconv.h')
@@ -617,7 +627,6 @@ EOM
       end
 
       # Defining a macro that expands to a C string; double quotes are significant.
-      $CPPFLAGS << ' ' << "-DNOKOGIRI_#{recipe.name.upcase}_PATH=\"#{recipe.path}\"".inspect
       $CPPFLAGS << ' ' << "-DNOKOGIRI_#{recipe.name.upcase}_PATCHES=\"#{recipe.patch_files.map { |path| File.basename(path) }.join(' ')}\"".inspect
 
       case libname

data/ext/nokogiri/nokogiri.c

@@ -76,24 +76,30 @@ void Init_nokogiri()
   mNokogiriHtmlSax  = rb_define_module_under(mNokogiriHtml, "SAX");
 
   rb_const_set( mNokogiri,
-                rb_intern("LIBXML_VERSION"),
+                rb_intern("LIBXML_COMPILED_VERSION"),
                 NOKOGIRI_STR_NEW2(LIBXML_DOTTED_VERSION)
               );
   rb_const_set( mNokogiri,
-                rb_intern("LIBXML_PARSER_VERSION"),
+                rb_intern("LIBXML_LOADED_VERSION"),
                 NOKOGIRI_STR_NEW2(xmlParserVersion)
               );
 
+
+  rb_const_set( mNokogiri,
+                rb_intern("LIBXSLT_COMPILED_VERSION"),
+                NOKOGIRI_STR_NEW2(LIBXSLT_DOTTED_VERSION)
+              );
+  rb_const_set( mNokogiri,
+                rb_intern("LIBXSLT_LOADED_VERSION"),
+                NOKOGIRI_STR_NEW2(xsltEngineVersion)
+              );
+
 #ifdef NOKOGIRI_USE_PACKAGED_LIBRARIES
   rb_const_set(mNokogiri, rb_intern("NOKOGIRI_USE_PACKAGED_LIBRARIES"), Qtrue);
-  rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXML2_PATH"), NOKOGIRI_STR_NEW2(NOKOGIRI_LIBXML2_PATH));
-  rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXSLT_PATH"), NOKOGIRI_STR_NEW2(NOKOGIRI_LIBXSLT_PATH));
   rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXML2_PATCHES"), rb_str_split(NOKOGIRI_STR_NEW2(NOKOGIRI_LIBXML2_PATCHES), " "));
   rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXSLT_PATCHES"), rb_str_split(NOKOGIRI_STR_NEW2(NOKOGIRI_LIBXSLT_PATCHES), " "));
 #else
   rb_const_set(mNokogiri, rb_intern("NOKOGIRI_USE_PACKAGED_LIBRARIES"), Qfalse);
-  rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXML2_PATH"), Qnil);
-  rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXSLT_PATH"), Qnil);
   rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXML2_PATCHES"), Qnil);
   rb_const_set(mNokogiri, rb_intern("NOKOGIRI_LIBXSLT_PATCHES"), Qnil);
 #endif

data/ext/nokogiri/nokogiri.h

@@ -29,6 +29,7 @@ int vasprintf (char **strp, const char *fmt, va_list ap);
 #include <libxml/relaxng.h>
 #include <libxml/xinclude.h>
 #include <libxslt/extensions.h>
+#include <libxslt/xsltconfig.h>
 #include <libxml/c14n.h>
 #include <ruby.h>
 #include <ruby/st.h>

data/ext/nokogiri/xml_document.c

@@ -38,6 +38,15 @@ static void remove_private(xmlNodePtr node)
   node->_private = NULL;
 }
 
+static void mark(xmlDocPtr doc)
+{
+  nokogiriTuplePtr tuple = (nokogiriTuplePtr)doc->_private;
+  if(tuple) {
+      rb_gc_mark(tuple->doc);
+      rb_gc_mark(tuple->node_cache);
+  }
+}
+
 static void dealloc(xmlDocPtr doc)
 {
   st_table *node_hash;
@@ -588,7 +597,7 @@ VALUE Nokogiri_wrap_xml_document(VALUE klass, xmlDocPtr doc)
 
   VALUE rb_doc = Data_Wrap_Struct(
       klass ? klass : cNokogiriXmlDocument,
-      0,
+      mark,
       dealloc,
       doc
   );

data/ext/nokogiri/xml_node.c

@@ -1332,6 +1332,25 @@ static VALUE line(VALUE self)
   return INT2NUM(xmlGetLineNo(node));
 }
 
+/*
+ * call-seq:
+ *  line=(num)
+ *
+ * Sets the line for this Node. num must be less than 65535.
+ */
+static VALUE set_line(VALUE self, VALUE num)
+{
+  xmlNodePtr node;
+  Data_Get_Struct(self, xmlNode, node);
+
+  int value = NUM2INT(num);
+  if (value < 65535) {
+    node->line = value;
+  }
+
+  return num;
+}
+
 /*
  * call-seq:
  *  add_namespace_definition(prefix, href)
@@ -1728,6 +1747,7 @@ void init_xml_node()
   rb_define_method(klass, "create_external_subset", create_external_subset, 3);
   rb_define_method(klass, "pointer_id", pointer_id, 0);
   rb_define_method(klass, "line", line, 0);
+  rb_define_method(klass, "line=", set_line, 1);
   rb_define_method(klass, "content", get_native_content, 0);
   rb_define_method(klass, "native_content=", set_native_content, 1);
   rb_define_method(klass, "lang", get_lang, 0);

data/ext/nokogiri/xml_schema.c

@@ -133,31 +133,6 @@ static VALUE read_memory(VALUE klass, VALUE content)
   return rb_schema;
 }
 
-/* Schema creation will remove and deallocate "blank" nodes.
- * If those blank nodes have been exposed to Ruby, they could get freed
- * out from under the VALUE pointer.  This function checks to see if any of
- * those nodes have been exposed to Ruby, and if so we should raise an exception.
- */
-static int has_blank_nodes_p(VALUE cache)
-{
-    long i;
-
-    if (NIL_P(cache)) {
-        return 0;
-    }
-
-    for (i = 0; i < RARRAY_LEN(cache); i++) {
-        xmlNodePtr node;
-        VALUE element = rb_ary_entry(cache, i);
-        Data_Get_Struct(element, xmlNode, node);
-        if (xmlIsBlankNode(node)) {
-            return 1;
-        }
-    }
-
-    return 0;
-}
-
 /*
  * call-seq:
  *  from_document(doc)
@@ -177,10 +152,6 @@ static VALUE from_document(VALUE klass, VALUE document)
   /* In case someone passes us a node. ugh. */
   doc = doc->doc;
 
-  if (has_blank_nodes_p(DOC_NODE_CACHE(doc))) {
-    rb_raise(rb_eArgError, "Creating a schema from a document that has blank nodes exposed to Ruby is dangerous");
-  }
-
   ctx = xmlSchemaNewDocParserCtxt(doc);
 
   errors = rb_ary_new();

data/ext/nokogiri/xslt_stylesheet.c

@@ -109,10 +109,6 @@ static VALUE serialize(VALUE self, VALUE xmlobj)
     return rval ;
 }
 
-static void swallow_superfluous_xml_errors(void * userdata, xmlErrorPtr error, ...)
-{
-}
-
 /*
  *  call-seq:
  *    transform(document, params = [])

data/lib/nokogiri.rb

@@ -1,28 +1,11 @@
 # -*- coding: utf-8 -*-
+# frozen_string_literal: true
 # Modify the PATH on windows so that the external DLLs will get loaded.
 
 require 'rbconfig'
 
 if defined?(RUBY_ENGINE) && RUBY_ENGINE == "jruby"
-  # The line below caused a problem on non-GAE rack environment.
-  # unless defined?(JRuby::Rack::VERSION) || defined?(AppEngine::ApiProxy)
-  #
-  # However, simply cutting defined?(JRuby::Rack::VERSION) off resulted in
-  # an unable-to-load-nokogiri problem. Thus, now, Nokogiri checks the presense
-  # of appengine-rack.jar in $LOAD_PATH. If Nokogiri is on GAE, Nokogiri
-  # should skip loading xml jars. This is because those are in WEB-INF/lib and
-  # already set in the classpath.
-  unless $LOAD_PATH.to_s.include?("appengine-rack")
-    require 'stringio'
-    require 'isorelax.jar'
-    require 'jing.jar'
-    require 'nekohtml.jar'
-    require 'nekodtd.jar'
-    require 'xercesImpl.jar'
-    require 'serializer.jar'
-    require 'xalan.jar'
-    require 'xml-apis.jar'
-  end
+  require 'nokogiri/jruby/dependencies'
 end
 
 begin

data/lib/nokogiri/css.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 require 'nokogiri/css/node'
 require 'nokogiri/css/xpath_visitor'
 x = $-w

data/lib/nokogiri/css/node.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module CSS
     class Node

data/lib/nokogiri/css/parser.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 #
 # DO NOT MODIFY!!!!
 # This file is automatically generated by Racc 1.4.14

data/lib/nokogiri/css/parser_extras.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 require 'thread'
 
 module Nokogiri

data/lib/nokogiri/css/syntax_error.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 require 'nokogiri/syntax_error'
 module Nokogiri
   module CSS

data/lib/nokogiri/css/tokenizer.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 #--
 # DO NOT MODIFY!!!!
 # This file is automatically generated by rex 1.0.7

data/lib/nokogiri/css/xpath_visitor.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module CSS
     class XPathVisitor # :nodoc:
@@ -51,7 +52,8 @@ module Nokogiri
         when /^comment\(/
           "comment()"
         when /^has\(/
-          ".//#{node.value[1].accept(self)}"
+          is_direct = node.value[1].value[0].nil? # e.g. "has(> a)", "has(~ a)", "has(+ a)"
+          ".#{"//" if !is_direct}#{node.value[1].accept(self)}"
         else
           args = ['.'] + node.value[1..-1]
           "#{node.value.first}#{args.join(', ')})"

data/lib/nokogiri/decorators/slop.rb

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

data/lib/nokogiri/html.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 require 'nokogiri/html/entity_lookup'
 require 'nokogiri/html/document'
 require 'nokogiri/html/document_fragment'

data/lib/nokogiri/html/builder.rb

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

data/lib/nokogiri/html/document.rb

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

data/lib/nokogiri/html/document_fragment.rb

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

data/lib/nokogiri/html/element_description.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module HTML
     class ElementDescription

data/lib/nokogiri/html/element_description_defaults.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module HTML
     class ElementDescription

data/lib/nokogiri/html/entity_lookup.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   module HTML
     class EntityDescription < Struct.new(:value, :name, :description); end

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

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

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

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

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

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

data/lib/nokogiri/jruby/dependencies.rb

@@ -0,0 +1,20 @@
+# frozen_string_literal: true
+# The line below caused a problem on non-GAE rack environment.
+# unless defined?(JRuby::Rack::VERSION) || defined?(AppEngine::ApiProxy)
+#
+# However, simply cutting defined?(JRuby::Rack::VERSION) off resulted in
+# an unable-to-load-nokogiri problem. Thus, now, Nokogiri checks the presense
+# of appengine-rack.jar in $LOAD_PATH. If Nokogiri is on GAE, Nokogiri
+# should skip loading xml jars. This is because those are in WEB-INF/lib and
+# already set in the classpath.
+unless $LOAD_PATH.to_s.include?("appengine-rack")
+  require 'stringio'
+  require 'isorelax.jar'
+  require 'jing.jar'
+  require 'nekohtml.jar'
+  require 'nekodtd.jar'
+  require 'xercesImpl.jar'
+  require 'serializer.jar'
+  require 'xalan.jar'
+  require 'xml-apis.jar'
+end

data/lib/nokogiri/syntax_error.rb

@@ -1,3 +1,4 @@
+# frozen_string_literal: true
 module Nokogiri
   class SyntaxError < ::StandardError
   end

data/lib/nokogiri/version.rb

@@ -1,6 +1,7 @@
+# frozen_string_literal: true
 module Nokogiri
   # The version of Nokogiri you are using
-  VERSION = "1.10.10"
+  VERSION = "1.11.0.rc1"
 
   class VersionInfo # :nodoc:
     def jruby?
@@ -11,19 +12,30 @@ module Nokogiri
       defined?(RUBY_ENGINE) ? RUBY_ENGINE : "mri"
     end
 
-    def loaded_parser_version
-      LIBXML_PARSER_VERSION.
+    def loaded_libxml_version
+      Gem::Version.new(LIBXML_LOADED_VERSION.
         scan(/^(\d+)(\d\d)(\d\d)(?!\d)/).first.
         collect(&:to_i).
-        join(".")
+        join("."))
     end
 
-    def compiled_parser_version
-      LIBXML_VERSION
+    def compiled_libxml_version
+      Gem::Version.new LIBXML_COMPILED_VERSION
+    end
+
+    def loaded_libxslt_version
+      Gem::Version.new(LIBXSLT_LOADED_VERSION.
+        scan(/^(\d+)(\d\d)(\d\d)(?!\d)/).first.
+        collect(&:to_i).
+        join("."))
+    end
+
+    def compiled_libxslt_version
+      Gem::Version.new LIBXSLT_COMPILED_VERSION
     end
 
     def libxml2?
-      defined?(LIBXML_VERSION)
+      defined?(LIBXML_COMPILED_VERSION)
     end
 
     def libxml2_using_system?
@@ -35,47 +47,62 @@ module Nokogiri
     end
 
     def warnings
-      return [] unless libxml2?
+      warnings = []
 
-      if compiled_parser_version != loaded_parser_version
-        ["Nokogiri was built against LibXML version #{compiled_parser_version}, but has dynamically loaded #{loaded_parser_version}"]
-      else
-        []
+      if libxml2?
+        if compiled_libxml_version != loaded_libxml_version
+          warnings << "Nokogiri was built against libxml version #{compiled_libxml_version}, but has dynamically loaded #{loaded_libxml_version}"
+        end
+
+        if compiled_libxslt_version != loaded_libxslt_version
+          warnings << "Nokogiri was built against libxslt version #{compiled_libxslt_version}, but has dynamically loaded #{loaded_libxslt_version}"
+        end
       end
+
+      warnings
     end
 
     def to_hash
-      hash_info = {}
-      hash_info["warnings"] = []
-      hash_info["nokogiri"] = Nokogiri::VERSION
-      hash_info["ruby"] = {}
-      hash_info["ruby"]["version"] = ::RUBY_VERSION
-      hash_info["ruby"]["platform"] = ::RUBY_PLATFORM
-      hash_info["ruby"]["description"] = ::RUBY_DESCRIPTION
-      hash_info["ruby"]["engine"] = engine
-      hash_info["ruby"]["jruby"] = jruby? if jruby?
+      {}.tap do |vi|
+        vi["warnings"] = []
+        vi["nokogiri"] = Nokogiri::VERSION
+        vi["ruby"] = {}.tap do |ruby|
+          ruby["version"] = ::RUBY_VERSION
+          ruby["platform"] = ::RUBY_PLATFORM
+          ruby["description"] = ::RUBY_DESCRIPTION
+          ruby["engine"] = engine
+          ruby["jruby"] = jruby? if jruby?
+        end
 
-      if libxml2?
-        hash_info["libxml"] = {}
-        hash_info["libxml"]["binding"] = "extension"
-        if libxml2_using_packaged?
-          hash_info["libxml"]["source"] = "packaged"
-          hash_info["libxml"]["libxml2_path"] = NOKOGIRI_LIBXML2_PATH
-          hash_info["libxml"]["libxslt_path"] = NOKOGIRI_LIBXSLT_PATH
-          hash_info["libxml"]["libxml2_patches"] = NOKOGIRI_LIBXML2_PATCHES
-          hash_info["libxml"]["libxslt_patches"] = NOKOGIRI_LIBXSLT_PATCHES
-        else
-          hash_info["libxml"]["source"] = "system"
+        if libxml2?
+          vi["libxml"] = {}.tap do |libxml|
+            if libxml2_using_packaged?
+              libxml["source"] = "packaged"
+              libxml["patches"] = NOKOGIRI_LIBXML2_PATCHES
+            else
+              libxml["source"] = "system"
+            end
+            libxml["compiled"] = compiled_libxml_version.to_s
+            libxml["loaded"] = loaded_libxml_version.to_s
+          end
+
+          vi["libxslt"] = {}.tap do |libxslt|
+            if libxml2_using_packaged?
+              libxslt["source"] = "packaged"
+              libxslt["patches"] = NOKOGIRI_LIBXSLT_PATCHES
+            else
+              libxslt["source"] = "system"
+            end
+            libxslt["compiled"] = compiled_libxslt_version.to_s
+            libxslt["loaded"] = loaded_libxslt_version.to_s
+          end
+
+          vi["warnings"] = warnings
+        elsif jruby?
+          vi["xerces"] = Nokogiri::XERCES_VERSION
+          vi["nekohtml"] = Nokogiri::NEKO_VERSION
         end
-        hash_info["libxml"]["compiled"] = compiled_parser_version
-        hash_info["libxml"]["loaded"] = loaded_parser_version
-        hash_info["warnings"] = warnings
-      elsif jruby?
-        hash_info["xerces"] = Nokogiri::XERCES_VERSION
-        hash_info["nekohtml"] = Nokogiri::NEKO_VERSION
       end
-
-      hash_info
     end
 
     def to_markdown
@@ -96,14 +123,27 @@ module Nokogiri
     def self.instance; @@instance; end
   end
 
-  # More complete version information about libxml
-  VERSION_INFO = VersionInfo.instance.to_hash
-
-  def self.uses_libxml? # :nodoc:
-    VersionInfo.instance.libxml2?
+  def self.uses_libxml?(requirement = nil) # :nodoc:
+    return false unless VersionInfo.instance.libxml2?
+    return true unless requirement
+    return Gem::Requirement.new(requirement).satisfied_by?(VersionInfo.instance.loaded_libxml_version)
   end
 
   def self.jruby? # :nodoc:
     VersionInfo.instance.jruby?
   end
+
+  # Ensure constants used in this file are loaded
+  if Nokogiri.jruby?
+    require "nokogiri/jruby/dependencies"
+  end
+  begin
+    RUBY_VERSION =~ /(\d+\.\d+)/
+    require "nokogiri/#{$1}/nokogiri"
+  rescue LoadError
+    require "nokogiri/nokogiri"
+  end
+
+  # More complete version information about libxml
+  VERSION_INFO = VersionInfo.instance.to_hash
 end