nokogiri 1.7.2-java → 1.8.0-java

data/ext/nokogiri/extconf.rb

@@ -223,8 +223,8 @@ def iconv_configure_flags
 
       return [
         '--with-iconv=yes',
-        *("CPPFLAGS=#{idirs.map { |dir| '-I' << dir }.join(' ')}" if idirs),
-        *("LDFLAGS=#{ldirs.map { |dir| '-L' << dir }.join(' ')}" if ldirs),
+        *("CPPFLAGS=#{idirs.map { |dir| '-I' + dir }.join(' ')}" if idirs),
+        *("LDFLAGS=#{ldirs.map { |dir| '-L' + dir }.join(' ')}" if ldirs),
       ]
     end
   end
@@ -260,11 +260,12 @@ end
 
 def process_recipe(name, version, static_p, cross_p)
   MiniPortile.new(name, version).tap do |recipe|
-    recipe.target = portsdir = File.join(ROOT, "ports")
+    recipe.target = File.join(ROOT, "ports")
     # Prefer host_alias over host in order to use i586-mingw32msvc as
     # correct compiler prefix for cross build, but use host if not set.
     recipe.host = RbConfig::CONFIG["host_alias"].empty? ? RbConfig::CONFIG["host"] : RbConfig::CONFIG["host_alias"]
     recipe.patch_files = Dir[File.join(ROOT, "patches", name, "*.patch")].sort
+    recipe.configure_options << "--libdir=#{File.join(recipe.path, "lib")}"
 
     yield recipe
 
@@ -307,7 +308,7 @@ def process_recipe(name, version, static_p, cross_p)
     if RbConfig::CONFIG['target_cpu'] == 'universal'
       %w[CFLAGS LDFLAGS].each do |key|
         unless env[key].include?('-arch')
-          env[key] << ' ' << RbConfig::CONFIG['ARCH_FLAG']
+          env[key] += ' ' + RbConfig::CONFIG['ARCH_FLAG']
         end
       end
     end
@@ -391,7 +392,7 @@ when arg_config('--clean')
   do_clean
 end
 
-if openbsd?
+if openbsd? && !using_system_libraries?
   ENV['CC'] ||= find_executable('egcc') or
     abort "Please install gcc 4.9+ from ports using `pkg_add -v gcc`"
 end
@@ -456,7 +457,7 @@ else
   # The gem version constraint in the Rakefile is not respected at install time.
   # Keep this version in sync with the one in the Rakefile !
   require 'rubygems'
-  gem 'mini_portile2', '~> 2.1.0'
+  gem 'mini_portile2', '~> 2.2.0'
   require 'mini_portile2'
   message "Using mini_portile version #{MiniPortile::VERSION}\n"
 
@@ -473,8 +474,8 @@ else
   if cross_build_p || windows?
     zlib_recipe = process_recipe("zlib", dependencies["zlib"]["version"], static_p, cross_build_p) do |recipe|
       recipe.files = [{
-          url: "http://zlib.net/#{recipe.name}-#{recipe.version}.tar.gz",
-          md5: dependencies["zlib"]["md5"]
+          url: "http://zlib.net/fossils/#{recipe.name}-#{recipe.version}.tar.gz",
+          sha256: dependencies["zlib"]["sha256"]
         }]
       class << recipe
         attr_accessor :cross_build_p
@@ -512,7 +513,7 @@ else
     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",
-          md5: dependencies["libiconv"]["md5"]
+          sha256: dependencies["libiconv"]["sha256"]
         }]
       recipe.configure_options += [
         "CPPFLAGS=-Wall",
@@ -547,7 +548,7 @@ EOM
   libxml2_recipe = process_recipe("libxml2", dependencies["libxml2"]["version"], static_p, cross_build_p) do |recipe|
     recipe.files = [{
         url: "http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz",
-        md5: dependencies["libxml2"]["md5"]
+        sha256: dependencies["libxml2"]["sha256"]
       }]
     recipe.configure_options += [
       "--without-python",
@@ -563,7 +564,7 @@ EOM
   libxslt_recipe = process_recipe("libxslt", dependencies["libxslt"]["version"], static_p, cross_build_p) do |recipe|
     recipe.files = [{
         url: "http://xmlsoft.org/sources/#{recipe.name}-#{recipe.version}.tar.gz",
-        md5: dependencies["libxslt"]["md5"]
+        sha256: dependencies["libxslt"]["sha256"]
       }]
     recipe.configure_options += [
       "--without-python",
@@ -657,12 +658,6 @@ have_func('xmlRelaxNGSetValidStructuredErrors')
 have_func('xmlSchemaSetValidStructuredErrors')
 have_func('xmlSchemaSetParserStructuredErrors')
 
-if ENV['CPUPROFILE']
-  unless find_library('profiler', 'ProfilerEnable', *LIB_DIRS)
-    abort "google performance tools are not installed"
-  end
-end
-
 create_makefile('nokogiri/nokogiri')
 
 if enable_config('clean', true)

data/ext/nokogiri/nokogiri.h

@@ -34,16 +34,6 @@ int vasprintf (char **strp, const char *fmt, va_list ap);
 #include <ruby/st.h>
 #include <ruby/encoding.h>
 
-#ifndef UNUSED
-# if defined(__GNUC__)
-#  define MAYBE_UNUSED(name) name __attribute__((unused))
-#  define UNUSED(name) MAYBE_UNUSED(UNUSED_ ## name)
-# else
-#  define MAYBE_UNUSED(name) name
-#  define UNUSED(name) name
-# endif
-#endif
-
 #ifndef NORETURN
 # if defined(__GNUC__)
 #  define NORETURN(name) __attribute__((noreturn)) name

data/ext/nokogiri/xml_attr.c

@@ -11,11 +11,11 @@ static VALUE set_value(VALUE self, VALUE content)
   xmlAttrPtr attr;
   Data_Get_Struct(self, xmlAttr, attr);
 
-  if(attr->children) xmlFreeNodeList(attr->children);
+  if (attr->children) { xmlFreeNodeList(attr->children); }
 
   attr->children = attr->last = NULL;
 
-  if(content) {
+  if (content) {
     xmlChar *buffer;
     xmlNode *tmp;
 
@@ -30,7 +30,7 @@ static VALUE set_value(VALUE self, VALUE content)
     for(tmp = attr->children; tmp; tmp = tmp->next) {
       tmp->parent = (xmlNode *)attr;
       tmp->doc = attr->doc;
-      if(tmp->next == NULL) attr->last = tmp;
+      if (tmp->next == NULL) { attr->last = tmp; }
     }
 
     /* Free up memory */
@@ -57,20 +57,24 @@ static VALUE new(int argc, VALUE *argv, VALUE klass)
 
   rb_scan_args(argc, argv, "2*", &document, &name, &rest);
 
+  if (! rb_obj_is_kind_of(document, cNokogiriXmlDocument)) {
+    rb_raise(rb_eArgError, "parameter must be a Nokogiri::XML::Document");
+  }
+
   Data_Get_Struct(document, xmlDoc, xml_doc);
 
   node = xmlNewDocProp(
-      xml_doc,
-      (const xmlChar *)StringValueCStr(name),
-      NULL
-  );
+           xml_doc,
+           (const xmlChar *)StringValueCStr(name),
+           NULL
+         );
 
   nokogiri_root_node((xmlNodePtr)node);
 
   rb_node = Nokogiri_wrap_xml_node(klass, (xmlNodePtr)node);
   rb_obj_call_init(rb_node, argc, argv);
 
-  if(rb_block_given_p()) rb_yield(rb_node);
+  if (rb_block_given_p()) { rb_yield(rb_node); }
 
   return rb_node;
 }

data/ext/nokogiri/xml_node.c

@@ -833,37 +833,40 @@ static VALUE set(VALUE self, VALUE property, VALUE value)
 static VALUE get(VALUE self, VALUE rattribute)
 {
   xmlNodePtr node;
-  xmlChar* value = 0;
-  VALUE rvalue ;
-  char* attribute = 0;
-  char *colon = 0, *attr_name = 0, *prefix = 0;
+  const xmlChar *value = 0;
+  VALUE rvalue;
+  xmlChar *colon;
+  const xmlChar *attribute, *attr_name, *prefix;
   xmlNsPtr ns;
 
   if (NIL_P(rattribute)) return Qnil;
 
   Data_Get_Struct(self, xmlNode, node);
-  attribute = strdup(StringValueCStr(rattribute));
+  attribute = xmlCharStrdup(StringValueCStr(rattribute));
 
-  colon = strchr(attribute, ':');
+  colon = (xmlChar *)xmlStrchr(attribute, (const xmlChar)':');
   if (colon) {
-    (*colon) = 0 ; /* create two null-terminated strings of the prefix and attribute name */
-    prefix = attribute ;
-    attr_name = colon + 1 ;
-    ns = xmlSearchNs(node->doc, node, (const xmlChar *)(prefix));
+    /* split the attribute string into separate prefix and name by
+     * null-terminating the prefix at the colon */
+    prefix = attribute;
+    attr_name = colon + 1;
+    (*colon) = 0;
+
+    ns = xmlSearchNs(node->doc, node, prefix);
     if (ns) {
-      value = xmlGetNsProp(node, (xmlChar*)(attr_name), ns->href);
+      value = xmlGetNsProp(node, attr_name, ns->href);
     } else {
       value = xmlGetProp(node, (xmlChar*)StringValueCStr(rattribute));
     }
   } else {
-    value = xmlGetNoNsProp(node, (xmlChar*)attribute);
+    value = xmlGetNoNsProp(node, attribute);
   }
 
-  free(attribute);
+  xmlFree((void *)attribute);
   if (!value) return Qnil;
 
   rvalue = NOKOGIRI_STR_NEW2(value);
-  xmlFree(value);
+  xmlFree((void *)value);
 
   return rvalue ;
 }

data/ext/nokogiri/xml_sax_push_parser.c

@@ -3,7 +3,7 @@
 static void deallocate(xmlParserCtxtPtr ctx)
 {
   NOKOGIRI_DEBUG_START(ctx);
-  if(ctx != NULL) {
+  if (ctx != NULL) {
     NOKOGIRI_SAX_TUPLE_DESTROY(ctx->userData);
     xmlFreeParserCtxt(ctx);
   }
@@ -30,12 +30,12 @@ static VALUE native_write(VALUE self, VALUE _chunk, VALUE _last_chunk)
 
   Data_Get_Struct(self, xmlParserCtxt, ctx);
 
-  if(Qnil != _chunk) {
+  if (Qnil != _chunk) {
     chunk = StringValuePtr(_chunk);
     size = (int)RSTRING_LEN(_chunk);
   }
 
-  if(xmlParseChunk(ctx, chunk, size, Qtrue == _last_chunk ? 1 : 0)) {
+  if (xmlParseChunk(ctx, chunk, size, Qtrue == _last_chunk ? 1 : 0)) {
     if (!(ctx->options & XML_PARSE_RECOVER)) {
       xmlErrorPtr e = xmlCtxtGetLastError(ctx);
       Nokogiri_error_raise(NULL, e);
@@ -59,17 +59,18 @@ static VALUE initialize_native(VALUE self, VALUE _xml_sax, VALUE _filename)
 
   Data_Get_Struct(_xml_sax, xmlSAXHandler, sax);
 
-  if(_filename != Qnil) filename = StringValueCStr(_filename);
+  if (_filename != Qnil) { filename = StringValueCStr(_filename); }
 
   ctx = xmlCreatePushParserCtxt(
-      sax,
-      NULL,
-      NULL,
-      0,
-      filename
-  );
-  if(ctx == NULL)
+          sax,
+          NULL,
+          NULL,
+          0,
+          filename
+        );
+  if (ctx == NULL) {
     rb_raise(rb_eRuntimeError, "Could not create a parser context");
+  }
 
   ctx->userData = NOKOGIRI_SAX_TUPLE_NEW(ctx, self);
 
@@ -91,12 +92,53 @@ static VALUE set_options(VALUE self, VALUE options)
   xmlParserCtxtPtr ctx;
   Data_Get_Struct(self, xmlParserCtxt, ctx);
 
-  if (xmlCtxtUseOptions(ctx, (int)NUM2INT(options)) != 0)
+  if (xmlCtxtUseOptions(ctx, (int)NUM2INT(options)) != 0) {
     rb_raise(rb_eRuntimeError, "Cannot set XML parser context options");
+  }
 
   return Qnil;
 }
 
+/*
+ * call-seq:
+ *  replace_entities
+ *
+ * Should this parser replace entities?  & will get converted to '&' if
+ * set to true
+ */
+static VALUE get_replace_entities(VALUE self)
+{
+  xmlParserCtxtPtr ctx;
+  Data_Get_Struct(self, xmlParserCtxt, ctx);
+
+  if (0 == ctx->replaceEntities) {
+    return Qfalse;
+  } else {
+    return Qtrue;
+  }
+}
+
+/*
+ * call-seq:
+ *  replace_entities=(boolean)
+ *
+ * Should this parser replace entities?  & will get converted to '&' if
+ * set to true
+ */
+static VALUE set_replace_entities(VALUE self, VALUE value)
+{
+  xmlParserCtxtPtr ctx;
+  Data_Get_Struct(self, xmlParserCtxt, ctx);
+
+  if (Qfalse == value) {
+    ctx->replaceEntities = 0;
+  } else {
+    ctx->replaceEntities = 1;
+  }
+
+  return value;
+}
+
 VALUE cNokogiriXmlSaxPushParser ;
 void init_xml_sax_push_parser()
 {
@@ -112,4 +154,6 @@ void init_xml_sax_push_parser()
   rb_define_private_method(klass, "native_write", native_write, 2);
   rb_define_method(klass, "options", get_options, 0);
   rb_define_method(klass, "options=", set_options, 1);
+  rb_define_method(klass, "replace_entities", get_replace_entities, 0);
+  rb_define_method(klass, "replace_entities=", set_replace_entities, 1);
 }

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

@@ -36,6 +36,16 @@ module Nokogiri
           ctx.parse_with self
         end
 
+        ###
+        # Parse given +io+
+        def parse_io io, encoding = 'UTF-8'
+          check_encoding(encoding)
+          @encoding = encoding
+          ctx = ParserContext.io(io, ENCODINGS[encoding])
+          yield ctx if block_given?
+          ctx.parse_with self
+        end
+
         ###
         # Parse a file with +filename+
         def parse_file filename, encoding = 'UTF-8'

data/lib/nokogiri/version.rb

@@ -1,6 +1,6 @@
 module Nokogiri
   # The version of Nokogiri you are using
-  VERSION = '1.7.2'
+  VERSION = '1.8.0'
 
   class VersionInfo # :nodoc:
     def jruby?
@@ -12,9 +12,10 @@ module Nokogiri
     end
 
     def loaded_parser_version
-      LIBXML_PARSER_VERSION.scan(/^(\d+)(\d\d)(\d\d)(?!\d)/).first.collect{ |j|
-        j.to_i
-      }.join(".")
+      LIBXML_PARSER_VERSION.
+        scan(/^(\d+)(\d\d)(\d\d)(?!\d)/).first.
+        collect(&:to_i).
+        join(".")
     end
 
     def compiled_parser_version

data/lib/nokogiri/xml/document.rb

@@ -33,10 +33,11 @@ module Nokogiri
       # +block+ (optional) is passed a configuration object on which
       # parse options may be set.
       #
-      # When parsing untrusted documents, it's recommended that the
-      # +nonet+ option be used, as shown in this example code:
-      #
-      #   Nokogiri::XML::Document.parse(xml_string) { |config| config.nonet }
+      # By default, Nokogiri treats documents as untrusted, and so
+      # does not attempt to load DTDs or access the network. See
+      # Nokogiri::XML::ParseOptions for a complete list of options;
+      # and that module's DEFAULT_XML constant for what's set (and not
+      # set) by default.
       #
       # Nokogiri.XML() is a convenience method which will call this method.
       #
@@ -239,10 +240,10 @@ module Nokogiri
       undef_method :namespace_definitions, :line, :add_namespace
 
       def add_child node_or_tags
-        raise "Document already has a root node" if root && root.name != 'nokogiri_text_wrapper'
+        raise "A document may not have multiple root nodes." if (root && root.name != 'nokogiri_text_wrapper') && !(node_or_tags.comment? || node_or_tags.processing_instruction?)
         node_or_tags = coerce(node_or_tags)
         if node_or_tags.is_a?(XML::NodeSet)
-          raise "Document cannot have multiple root nodes" if node_or_tags.size > 1
+          raise "A document may not have multiple root nodes." if node_or_tags.size > 1
           super(node_or_tags.first)
         else
           super
@@ -273,9 +274,8 @@ module Nokogiri
           (string_or_io.respond_to?(:eof?) && string_or_io.eof?)
       end
 
-      def implied_xpath_contexts # :nodoc:
-        ["//"]
-      end
+      # @private
+      IMPLIED_XPATH_CONTEXTS = [ '//'.freeze ].freeze # :nodoc:
 
       def inspect_attributes
         [:name, :children]

data/lib/nokogiri/xml/node.rb

@@ -154,7 +154,7 @@ module Nokogiri
       def prepend_child node_or_tags
         if first = children.first
           # Mimic the error add_child would raise.
-          raise RuntimeError, "Document already has a root node" if document? && !node_or_tags.processing_instruction?
+          raise RuntimeError, "Document already has a root node" if document? && !(node_or_tags.comment? || node_or_tags.processing_instruction?)
           first.__send__(:add_sibling, :previous, node_or_tags)
         else
           add_child(node_or_tags)
@@ -172,6 +172,7 @@ module Nokogiri
         add_child node_or_tags
         self
       end
+
       ###
       # Insert +node_or_tags+ before this Node (as a sibling).
       # +node_or_tags+ can be a Nokogiri::XML::Node, a ::DocumentFragment, a ::NodeSet, or a string containing markup.
@@ -180,7 +181,7 @@ module Nokogiri
       #
       # Also see related method +before+.
       def add_previous_sibling node_or_tags
-        raise ArgumentError.new("A document may not have multiple root nodes.") if (parent && parent.document?) && !node_or_tags.processing_instruction?
+        raise ArgumentError.new("A document may not have multiple root nodes.") if (parent && parent.document?) && !(node_or_tags.comment? || node_or_tags.processing_instruction?)
 
         add_sibling :previous, node_or_tags
       end
@@ -193,7 +194,7 @@ module Nokogiri
       #
       # Also see related method +after+.
       def add_next_sibling node_or_tags
-        raise ArgumentError.new("A document may not have multiple root nodes.") if (parent && parent.document?) && !node_or_tags.processing_instruction?
+        raise ArgumentError.new("A document may not have multiple root nodes.") if (parent && parent.document?) && !(node_or_tags.comment? || node_or_tags.processing_instruction?)
 
         add_sibling :next, node_or_tags
       end
@@ -622,7 +623,7 @@ module Nokogiri
         encoding = options[:encoding] || document.encoding
         options[:encoding] = encoding
 
-        outstring = ""
+        outstring = String.new
         if encoding && outstring.respond_to?(:force_encoding)
           outstring.force_encoding(Encoding.find(encoding))
         end
@@ -818,9 +819,8 @@ Requires a Node, NodeSet or String argument, and cannot accept a #{data.class}.
         EOERR
       end
 
-      def implied_xpath_contexts # :nodoc:
-        [".//"]
-      end
+      # @private
+      IMPLIED_XPATH_CONTEXTS = [ './/'.freeze ].freeze # :nodoc:
 
       def add_child_node_and_reparent_attrs node # :nodoc:
         add_child_node node