nokogiri 1.0.5 → 1.0.6

data/History.ja.txt

@@ -0,0 +1,56 @@
+=== 1.0.6
+
+* 5つの修正
+
+  * XPath のパーサーが SyntaxError を生じさせ解析停止させる
+  * CSS のパーサーが SyntaxError を生じさせ解析停止させる
+  * filter() と not() hpricot の互換性を追加
+  * CSS が Node#search 経由で検索し、常時対応する事が出来るようになった
+  * CSS より XPath 変換がキャッシュに入れられるようになった
+
+=== 1.0.5
+
+* バグフィックス
+
+  * メーリンクリストを作成
+  * バグファイルを作成
+  * Windows 内で ENV['PATH'] が存在しない場合でも、存在出来るように設定完了
+  * Document 内の NodeSet#[] の結果をキャッシュする
+
+=== 1.0.4
+
+* バグフィックス
+
+  * 弱参照からドキュメント参照へのメモリー管理の変換
+  * メモリリークに接続
+  * ビルダーブロックが取り囲んでいるコンテキストから
+    メソッドの呼び出しをする事が出来る
+
+=== 1.0.3
+
+* 5つのバグ修正
+
+  * NodeSet が to_ary へ実行
+  * XML::Document#parent を除去
+  * GCバグ修正済み (Mike は最高！)
+  * 1.8.5互換性の為の RARRAY_LEN 除去
+  * inner_html 修正済み (Yahuda に感謝)
+
+=== 1.0.2
+
+* 1つのバグ修正
+
+  * extconf.rb は frex や racc を調べないはず
+
+=== 1.0.1
+
+* 1つのバグ修正
+
+  * extconf.rb が libdir や prefix を検索しない事を確認済み
+    それによって、ports libxml/ruby が正しくリンクする (lucsky に感謝!)
+
+=== 1.0.0 / 2008-07-13
+
+* 1つの偉大な増進
+
+  * ご誕生である

data/History.txt

@@ -1,3 +1,13 @@
+=== 1.0.6
+
+* 5 Bugfixes
+
+  * XPath Parser raises a SyntaxError on parse failure
+  * CSS Parser raises a SyntaxError on parse failure
+  * filter() and not() hpricot compatibility added
+  * CSS searches via Node#search are now always relative
+  * CSS to XPath conversion is now cached
+
 === 1.0.5
 
 * Bugfixes

data/Manifest.txt

@@ -1,3 +1,4 @@
+History.ja.txt
 History.txt
 Manifest.txt
 README.ja.txt
@@ -16,6 +17,8 @@ ext/nokogiri/xml_document.c
 ext/nokogiri/xml_document.h
 ext/nokogiri/xml_dtd.c
 ext/nokogiri/xml_dtd.h
+ext/nokogiri/xml_io.c
+ext/nokogiri/xml_io.h
 ext/nokogiri/xml_node.c
 ext/nokogiri/xml_node.h
 ext/nokogiri/xml_node_set.c
@@ -41,6 +44,7 @@ lib/nokogiri/css/generated_tokenizer.rb
 lib/nokogiri/css/node.rb
 lib/nokogiri/css/parser.rb
 lib/nokogiri/css/parser.y
+lib/nokogiri/css/syntax_error.rb
 lib/nokogiri/css/tokenizer.rb
 lib/nokogiri/css/tokenizer.rex
 lib/nokogiri/css/xpath_visitor.rb
@@ -74,6 +78,7 @@ lib/nokogiri/xml/sax/parser.rb
 lib/nokogiri/xml/syntax_error.rb
 lib/nokogiri/xml/text.rb
 lib/nokogiri/xml/xpath.rb
+lib/nokogiri/xml/xpath/syntax_error.rb
 lib/nokogiri/xml/xpath_context.rb
 lib/nokogiri/xslt.rb
 lib/nokogiri/xslt/stylesheet.rb
@@ -107,6 +112,7 @@ test/html/sax/test_parser.rb
 test/html/test_builder.rb
 test/html/test_document.rb
 test/test_convert_xpath.rb
+test/test_css_cache.rb
 test/test_gc.rb
 test/test_nokogiri.rb
 test/test_reader.rb

data/README.ja.txt

@@ -18,13 +18,23 @@ Nokogiri はHTMLやXMLやSAXやXSLTやReaderのパーサーです。
 
 検索出来たり、正確にCSS3とXPathをサポート出来たりする。
 
-これはスピッドテストです：
+これはスピードテストです：
 
   * http://gist.github.com/22176
 
 NokogiriはHpricotの代わりに使用出来る。
 その互換性は簡単に正しいCSSとXPathを使用する事が出来る。
 
+== SUPPORT:
+
+ノコギリのメーリングリストは:
+
+  * http://rubyforge.org/mailman/listinfo/nokogiri-talk
+
+バグファイルは:
+
+  * http://nokogiri.lighthouseapp.com/projects/19607-nokogiri/overview
+
 == SYNOPSIS:
 
   require 'nokogiri'
@@ -55,6 +65,7 @@ NokogiriはHpricotの代わりに使用出来る。
 
 * ruby 1.8 or 1.9
 * libxml
+* libxslt
 
 == INSTALL:
 

data/README.txt

@@ -67,6 +67,7 @@ The bug tracker is available here:
 
 * ruby 1.8 or 1.9
 * libxml
+* libxslt
 
 == INSTALL:
 

data/Rakefile

@@ -289,10 +289,9 @@ end
 
 # Evil evil hack.  Do not run tests when gem installs
 if ENV['RUBYARCHDIR']
-  class << Rake::Task[:default]
-    attr_writer :prerequisites
-  end
-  Rake::Task[:default].prerequisites = [:build]
+  prereqs = Rake::Task[:default].prerequisites
+  prereqs.clear
+  prereqs << :build
 end
 
 # vim: syntax=Ruby

data/ext/nokogiri/native.h

@@ -11,6 +11,7 @@
 #include <libxml/HTMLparser.h>
 #include <libxml/HTMLtree.h>
 
+#include <xml_io.h>
 #include <xml_document.h>
 #include <html_document.h>
 #include <xml_node.h>

data/ext/nokogiri/xml_document.c

@@ -62,6 +62,47 @@ static VALUE root(VALUE self)
   return Nokogiri_wrap_xml_node(root) ;
 }
 
+/*
+ * call-seq:
+ *  read_io(io, url, encoding, options)
+ *
+ * Create a new document from an IO object
+ */
+static VALUE read_io( VALUE klass,
+                      VALUE io,
+                      VALUE url,
+                      VALUE encoding,
+                      VALUE options )
+{
+  const char * c_url    = (url == Qnil) ? NULL : StringValuePtr(url);
+  const char * c_enc    = (encoding == Qnil) ? NULL : StringValuePtr(encoding);
+
+  xmlInitParser();
+
+  xmlDocPtr doc = xmlReadIO(
+      (xmlInputReadCallback)io_read_callback,
+      (xmlInputCloseCallback)io_close_callback,
+      (void *)io,
+      c_url,
+      c_enc,
+      NUM2INT(options)
+  );
+
+  if(doc == NULL) {
+    xmlFreeDoc(doc);
+    rb_raise(rb_eRuntimeError, "Couldn't create a document");
+    return Qnil;
+  }
+
+  return Nokogiri_wrap_xml_document(klass, doc);
+}
+
+/*
+ * call-seq:
+ *  read_memory(string, url, encoding, options)
+ *
+ * Create a new document from a String
+ */
 static VALUE read_memory( VALUE klass,
                           VALUE string,
                           VALUE url,
@@ -85,6 +126,12 @@ static VALUE read_memory( VALUE klass,
   return Nokogiri_wrap_xml_document(klass, doc);
 }
 
+/*
+ * call-seq:
+ *  new
+ *
+ * Create a new document
+ */
 static VALUE new(int argc, VALUE *argv, VALUE klass)
 {
   VALUE version;
@@ -97,7 +144,7 @@ static VALUE new(int argc, VALUE *argv, VALUE klass)
 
 /*
  *  call-seq:
- *    substitute_entities_set bool)
+ *    substitute_entities=(boolean)
  *
  *  Set the global XML default for substitute entities.
  */
@@ -109,7 +156,7 @@ static VALUE substitute_entities_set(VALUE klass, VALUE value)
 
 /*
  *  call-seq:
- *    substitute_entities_set bool)
+ *    load_external_subsets=(boolean)
  *
  *  Set the global XML default for load external subsets.
  */
@@ -122,9 +169,19 @@ static VALUE load_external_subsets_set(VALUE klass, VALUE value)
 VALUE cNokogiriXmlDocument ;
 void init_xml_document()
 {
-  VALUE klass = cNokogiriXmlDocument = rb_const_get(mNokogiriXml, rb_intern("Document"));
+  VALUE nokogiri = rb_define_module("Nokogiri");
+  VALUE xml = rb_define_module_under(nokogiri, "XML");
+  VALUE node = rb_define_class_under(xml, "Node", rb_cObject);
+
+  /*
+   * Nokogiri::XML::Document wraps an xml document.
+   */
+  VALUE klass = rb_define_class_under(xml, "Document", node);
+
+  cNokogiriXmlDocument = klass;
 
   rb_define_singleton_method(klass, "read_memory", read_memory, 4);
+  rb_define_singleton_method(klass, "read_io", read_io, 4);
   rb_define_singleton_method(klass, "new", new, -1);
   rb_define_singleton_method(klass, "substitute_entities=", substitute_entities_set, 1);
   rb_define_singleton_method(klass, "load_external_subsets=", load_external_subsets_set, 1);

data/ext/nokogiri/xml_dtd.c

@@ -108,6 +108,10 @@ void init_xml_dtd()
 {
   VALUE nokogiri = rb_define_module("Nokogiri");
   VALUE xml = rb_define_module_under(nokogiri, "XML");
+
+  /*
+   * Nokogiri::XML::DTD wraps DTD nodes in an XML document
+   */
   VALUE klass = rb_define_class_under(xml, "DTD", cNokogiriXmlNode);
 
   rb_define_method(klass, "notations", notations, 0);

data/ext/nokogiri/xml_io.c

@@ -0,0 +1,17 @@
+#include <xml_io.h>
+
+int io_read_callback(void * ctx, char * buffer, int len) {
+  VALUE io = (VALUE)ctx;
+  VALUE string = rb_funcall(io, rb_intern("read"), 1, INT2NUM(len));
+
+  if(Qnil == string) return 0;
+  VALUE length = rb_funcall(string, rb_intern("length"), 0);
+
+  memcpy(buffer, StringValuePtr(string), (unsigned int)NUM2INT(length));
+
+  return NUM2INT(length);
+}
+
+int io_close_callback(void * ctx) {
+  return 0;
+}

data/ext/nokogiri/xml_io.h

@@ -0,0 +1,9 @@
+#ifndef NOKOGIRI_XML_IO
+#define NOKOGIRI_XML_IO
+
+#include <native.h>
+
+int io_read_callback(void * ctx, char * buffer, int len);
+int io_close_callback(void * ctx);
+
+#endif

data/ext/nokogiri/xml_node.c

@@ -514,7 +514,17 @@ static VALUE new_from_str(VALUE klass, VALUE xml)
 
 VALUE Nokogiri_wrap_xml_node(xmlNodePtr node)
 {
-  VALUE rb_node = Qnil;
+  assert(node);
+  assert(node->doc);
+  assert(node->doc->_private);
+
+  VALUE index = INT2NUM((int)node);
+  VALUE document = (VALUE)node->doc->_private;
+
+  VALUE node_cache = rb_funcall(document, rb_intern("node_cache"), 0);
+  VALUE rb_node = rb_hash_aref(node_cache, index);
+
+  if(rb_node != Qnil) return rb_node;
 
   switch(node->type)
   {
@@ -544,11 +554,8 @@ VALUE Nokogiri_wrap_xml_node(xmlNodePtr node)
       rb_node = Data_Wrap_Struct(cNokogiriXmlNode, 0, 0, node) ;
   }
 
-  assert(node);
-  assert(node->doc);
-  assert(node->doc->_private);
-
-  rb_iv_set(rb_node, "@document",(VALUE)node->doc->_private);
+  rb_hash_aset(node_cache, index, rb_node);
+  rb_iv_set(rb_node, "@document", document);
   rb_funcall(rb_node, rb_intern("decorate!"), 0);
   return rb_node ;
 }

data/ext/nokogiri/xml_node_set.c

@@ -52,22 +52,7 @@ static VALUE index_at(VALUE self, VALUE number)
   if(i < 0)
     i = i + node_set->nodeNr;
 
-  VALUE document = rb_funcall(self, rb_intern("document"), 0);
-  if(Qnil == document)
-    rb_raise(rb_eRuntimeError, "You forgot to set a document.");
-
-  VALUE index = INT2NUM((int)node_set->nodeTab[i]);
-
-  VALUE node_cache = rb_funcall(document, rb_intern("node_cache"), 0);
-
-  VALUE node = rb_hash_aref(node_cache, index);
-
-  if(Qnil == node) {
-    node = Nokogiri_wrap_xml_node(node_set->nodeTab[i]);
-    rb_hash_aset(node_cache, index, node);
-  }
-
-  return node;
+  return Nokogiri_wrap_xml_node(node_set->nodeTab[i]);
 }
 
 static void deallocate(xmlNodeSetPtr node_set)

data/ext/nokogiri/xml_xpath_context.c

@@ -39,7 +39,9 @@ static VALUE evaluate(VALUE self, VALUE search_path)
   xmlChar* query = (xmlChar *)StringValuePtr(search_path);
   xmlXPathObjectPtr xpath = xmlXPathEvalExpression(query, ctx);
   if(xpath == NULL) {
-    rb_raise(rb_eRuntimeError, "Couldn't evaluate expression '%s'", query);
+    VALUE xpath = rb_const_get(mNokogiriXml, rb_intern("XPath"));
+    VALUE error = rb_const_get(xpath, rb_intern("SyntaxError"));
+    rb_raise(error, "Couldn't evaluate expression '%s'", query);
   }
 
   VALUE xpath_object = Nokogiri_wrap_xml_xpath(xpath);

data/lib/nokogiri/css.rb

@@ -1,6 +1,20 @@
 require 'nokogiri/css/node'
 require 'nokogiri/css/xpath_visitor'
-require 'nokogiri/css/generated_tokenizer'
 require 'nokogiri/css/generated_parser'
+require 'nokogiri/css/generated_tokenizer'
 require 'nokogiri/css/tokenizer'
 require 'nokogiri/css/parser'
+require 'nokogiri/css/syntax_error'
+
+module Nokogiri
+  module CSS
+    class << self
+      def parse string
+        Parser.new.parse string
+      end
+      def xpath_for string, options={}
+        Parser.new.xpath_for string, options
+      end
+    end
+  end
+end

data/lib/nokogiri/css/generated_tokenizer.rb

@@ -6,7 +6,7 @@
 
 module Nokogiri
 module CSS
-class GeneratedTokenizer
+class GeneratedTokenizer < GeneratedParser
   require 'strscan'
 
   class ScanError < StandardError ; end

data/lib/nokogiri/css/node.rb

@@ -11,9 +11,11 @@ module Nokogiri
         visitor.send(:"visit_#{type.to_s.downcase}", self)
       end
 
-      def to_xpath prefix = '//', preprocess = true
-        self.preprocess! if preprocess
-        prefix + XPathVisitor.new.accept(self)
+      def to_xpath prefix = nil, visitor = nil
+        prefix ||= '//'
+        visitor ||= XPathVisitor.new
+        self.preprocess!
+        prefix + visitor.accept(self)
       end
 
       def preprocess!

data/lib/nokogiri/css/parser.rb

@@ -1,23 +1,48 @@
 module Nokogiri
   module CSS
-    class Parser < GeneratedParser
+    class Parser < GeneratedTokenizer
       class << self
         def parse string
           new.parse(string)
         end
-      end
+        def xpath_for string, options={}
+          new.xpath_for(string, options)
+        end
 
-      def initialize
-        @tokenizer = Tokenizer.new
+        def set_cache setting
+          @cache_on = setting ? true : false
+        end
+        def cache_on?
+          @cache ||= {}
+          instance_variable_defined?('@cache_on') ? @cache_on : true
+        end
+        def check_cache string
+          return unless cache_on?
+          @cache[string]
+        end
+        def add_cache string, value
+          return value unless cache_on?
+          @cache[string] = value
+        end
+        def clear_cache
+          @cache = {}
+        end
       end
+      alias :parse :scan_str
+
+      def xpath_for string, options={}
+        v = self.class.check_cache(string)
+        return v unless v.nil?
 
-      def parse string
-        @tokenizer.scan string
-        do_parse
+        prefix = options[:prefix] || nil
+        visitor = options[:visitor] || nil
+        args = [prefix, visitor]
+        self.class.add_cache(string, parse(string).map {|ast| ast.to_xpath(prefix, visitor)})
       end
 
-      def next_token
-        @tokenizer.next_token
+      def on_error error_token_id, error_value, value_stack
+        after = value_stack.compact.last
+        raise SyntaxError.new("unexpected '#{error_value}' after '#{after}'")
       end
     end
   end

data/lib/nokogiri/css/syntax_error.rb

@@ -0,0 +1,6 @@
+module Nokogiri
+  module CSS
+    class SyntaxError < ::SyntaxError
+    end
+  end
+end

data/lib/nokogiri/css/tokenizer.rex

@@ -1,6 +1,6 @@
 module Nokogiri
 module CSS
-class GeneratedTokenizer
+class GeneratedTokenizer < GeneratedParser
 
 macro
   nl        \n|\r\n|\r|\f

data/lib/nokogiri/decorators/hpricot/node.rb

@@ -8,8 +8,7 @@ module Nokogiri
             convert_to_xpath(path)
           }.flatten.uniq
 
-          namespaces = document.xml? ? document.namespaces.merge(ns) : ns
-          super(*converted + [namespaces])
+          super(*converted + [ns])
         end
         def /(path); search(path) end
 
@@ -32,16 +31,15 @@ module Nokogiri
           rule = rule.to_s
           case rule
           when %r{^//}
-            [".#{rule}"]
+            [".#{Hpricot::XPathVisitor.xpath_namespace_helper(rule)}"]
           when %r{^/}
-            [rule]
+            [Hpricot::XPathVisitor.xpath_namespace_helper(rule)]
           when %r{^.//}
-            [rule]
+            [Hpricot::XPathVisitor.xpath_namespace_helper(rule)]
           else
-            ctx = CSS::Parser.parse(rule)
             visitor = CSS::XPathVisitor.new
             visitor.extend(Hpricot::XPathVisitor)
-            ctx.map { |ast| './/' + visitor.accept(ast.preprocess!) }
+            CSS.xpath_for(rule, :prefix => ".//", :visitor => visitor)
           end
         end
 

data/lib/nokogiri/decorators/hpricot/node_set.rb

@@ -2,12 +2,52 @@ module Nokogiri
   module Decorators
     module Hpricot
       module NodeSet
-        def filter rule
+
+        # Select nodes matching the supplied rule.
+        # Note that positional rules (like <tt>:nth()</tt>) aren't currently supported.
+        #
+        # example: 
+        #   node_set.filter('.ohmy')          # selects nodes from the set with class "ohmy"
+        #   node_set.filter('a#link2')        # selects nodes from the set with child node <a id='link2'>
+        #   node_set.filter('a[@id="link2"]') # selects nodes from the set with child node <a id='link2'>
+        def filter(rule)
+          filter_transformer( lambda {|j| j}, rule ) # identity transformer
+        end
+
+        # The complement to filter, select nodes <em>not</em> matching the supplied rule.
+        # Note that positional rules (like <tt>:nth()</tt>) aren't currently supported.
+        #
+        # See filter for examples.
+        #
+        # Also note that you can pass a XML::Node object instead of a
+        # rule to remove that object from the node set (if it is
+        # present):
+        #    node_set.not(node_to_exclude) # selects all nodes EXCEPT node_to_exclude
+        #
+        def not(rule)
+          filter_transformer( lambda {|j| !j}, rule ) # negation transformer
+        end
+
+      private
+        def filter_transformer(transformer, rule)
+          sub_set = XML::NodeSet.new(document)
+          document.decorate(sub_set)
+
+          if rule.is_a?(XML::Node)
+            each { |node| sub_set << node if transformer.call(node == rule) }
+            return sub_set
+          end
+
           ctx = CSS::Parser.parse(rule.to_s)
           visitor = CSS::XPathVisitor.new
           visitor.extend(Hpricot::XPathVisitor)
-          search('.//self::' + visitor.accept(ctx.first))
-        end
+          each do |node|
+            if transformer.call(node.at(".//self::" + visitor.accept(ctx.first)))
+              sub_set << node
+            end
+          end
+          sub_set
+        end          
       end
     end
   end

data/lib/nokogiri/decorators/hpricot/xpath_visitor.rb

@@ -11,6 +11,17 @@ module Nokogiri
           end
           super(node).gsub(/child::text\(\)/, 'normalize-space(child::text())')
         end
+
+        #  take a path like '//t:sam' and convert to xpath "*[name()='t:sam']"
+        def self.xpath_namespace_helper rule
+          rule.split(/\//).collect do |tag|
+            if match = tag.match(/^(\w+:\w+)(.*)/)
+              "*[name()='#{match[1]}']#{match[2]}"
+            else
+              tag
+            end
+          end.join("/")
+        end
       end
     end
   end

data/lib/nokogiri/hpricot.rb

@@ -16,6 +16,11 @@ module Nokogiri
         add_decorators(doc)
       end
 
+      def HTML(string)
+        doc = Nokogiri::HTML.parse(string)
+        add_decorators(doc)
+      end
+
       def make string
         doc = XML::Document.new
         ns = XML::NodeSet.new(doc)
@@ -40,7 +45,7 @@ module Nokogiri
         builder = Nokogiri::HTML::Builder.new(&block)
         Nokogiri::Hpricot.add_decorators(builder.doc)
       else
-        doc = Nokogiri::HTML.parse(*args)
+        doc = Nokogiri.parse(*args)
         Nokogiri::Hpricot.add_decorators(doc)
       end
     end

data/lib/nokogiri/version.rb

@@ -1,3 +1,3 @@
 module Nokogiri
-  VERSION = '1.0.5'
+  VERSION = '1.0.6'
 end

data/lib/nokogiri/xml.rb

@@ -46,7 +46,7 @@ module Nokogiri
       def parse string_or_io, url = nil, encoding = nil, options = 2159
         if string_or_io.respond_to?(:read)
           url ||= string_or_io.respond_to?(:path) ? string_or_io.path : nil
-          string_or_io = string_or_io.read
+          return Document.read_io(string_or_io, url, encoding, options)
         end
 
         # read_memory pukes on empty docs

data/lib/nokogiri/xml/node.rb

@@ -45,9 +45,7 @@ module Nokogiri
       def search *paths
         ns = paths.last.is_a?(Hash) ? paths.pop : {}
         xpath(*(paths.map { |path|
-          path =~ /^(\.\/|\/)/ ? path : CSS::Parser.parse(path).map { |ast|
-            ast.to_xpath
-          }
+          path =~ /^(\.\/|\/)/ ? path : CSS.xpath_for(path, :prefix => ".//")
         }.flatten.uniq) + [ns])
       end
       alias :/ :search
@@ -78,9 +76,7 @@ module Nokogiri
       end
 
       def css *rules
-        xpath(*(rules.map { |rule|
-          CSS::Parser.parse(rule).map { |ast| "." + ast.to_xpath }
-        }.flatten.uniq))
+        xpath(*(rules.map { |rule| CSS.xpath_for(rule, :prefix => ".//") }.flatten.uniq))
       end
 
       def at path, ns = {}
@@ -192,11 +188,6 @@ module Nokogiri
         children.each{|j| j.traverse(&block) }
         block.call(self)
       end
-
-      def == other
-        pointer_id == other.pointer_id
-      end
-      alias :eql? :==
     end
   end
 end

data/lib/nokogiri/xml/xpath.rb

@@ -1,3 +1,5 @@
+require 'nokogiri/xml/xpath/syntax_error'
+
 module Nokogiri
   module XML
     class XPath

data/lib/nokogiri/xml/xpath/syntax_error.rb

@@ -0,0 +1,8 @@
+module Nokogiri
+  module XML
+    class XPath
+      class SyntaxError < ::SyntaxError
+      end
+    end
+  end
+end

data/test/css/test_parser.rb

@@ -7,6 +7,10 @@ module Nokogiri
         @parser = Nokogiri::CSS::Parser.new
       end
 
+      def test_syntax_error_raised
+        assert_raises(CSS::SyntaxError) { @parser.parse("a[x=]") }
+      end
+
       def test_find_by_type
         ast = @parser.parse("a:nth-child(2)").first
         matches = ast.find_by_type(

data/test/hpricot/test_parser.rb

@@ -69,24 +69,23 @@ class TestParser < Nokogiri::TestCase
     assert_equal 'link1', (doc/:p/:a).first['id']
     assert_equal 'link1', doc.search('p').at('a').get_attribute('id')
 
-    ### Modified: We're not supporting the filter() function
-    #assert_equal 'link2', (doc/'p').filter('.ohmy').search('a').first.get_attribute('id')
-    #assert_equal (doc/'p')[2], (doc/'p').filter(':nth(2)')[0]
-    #assert_equal (doc/'p')[2], (doc/'p').filter(':nth(2)')[0]
-    #assert_equal (doc/'p')[2], (doc/'p').filter('[3]')[0]
-    #assert_equal 4, (doc/'p').filter('*').length
-    #assert_equal 4, (doc/'p').filter('* *').length
-    #eles = (doc/'p').filter('.ohmy')
-    #assert_equal 1, eles.length
-    #assert_equal 'ohmy', eles.first.get_attribute('class')
+    assert_equal 'link2', (doc/'p').filter('.ohmy').search('a').first.get_attribute('id')
+    assert_equal((doc/'p')[2], (doc/'p').filter('[text()="The third paragraph"]')[0])
+    assert_equal((doc/'p')[3], (doc/'p').filter('b')[0])
+    assert_equal((doc/'p')[1], (doc/'p').filter('a[@id="link2"]')[0])
+    assert_equal((doc/'p')[3], (doc/'p').filter('.last')[0])
+    assert_equal 4, (doc/'p').filter('*').length
+    assert_equal 3, (doc/'p').filter('* *').length
+    eles = (doc/'p').filter('.ohmy')
+    assert_equal 1, eles.length
+    assert_equal 'ohmy', eles.first.get_attribute('class')
     assert_equal 3, (doc/'p:not(.ohmy)').length
 
-    ### Modified: We're not supporting the not() function
-    #assert_equal 3, (doc/'p').not('.ohmy').length
-    #assert_equal 3, (doc/'p').not(eles.first).length
-    #assert_equal 2, (doc/'p').filter('[@class]').length
+    assert_equal 3, (doc/'p').not('.ohmy').length
+    assert_equal 3, (doc/'p').not(eles.first).length
+    assert_equal 2, (doc/'p').filter('[@class]').length
     assert_equal 'last final', (doc/'p[@class~="final"]').first.get_attribute('class')
-    #assert_equal 1, (doc/'p').filter('[@class~="final"]').length
+    assert_equal 1, (doc/'p').filter('[@class~="final"]').length
     assert_equal 2, (doc/'p > a').length
     assert_equal 1, (doc/'p.ohmy > a').length
     assert_equal 2, (doc/'p / a').length
@@ -249,17 +248,17 @@ class TestParser < Nokogiri::TestCase
   def test_class_search
     # test case sent by Chih-Chao Lam
     # Modified.  libxml corrects this differently than hpricot
-    doc = Nokogiri.Hpricot("<div class=xyz '>abc</div>")
+    doc = Nokogiri::Hpricot.HTML("<div class=xyz '>abc</div>")
     assert_equal 1, doc.search(".xyz").length
 
-    doc = Nokogiri.Hpricot("<div class=xyz>abc</div><div class=abc>xyz</div>")
+    doc = Nokogiri::Hpricot.HTML("<div class=xyz>abc</div><div class=abc>xyz</div>")
     assert_equal 1, doc.search(".xyz").length
     assert_equal 4, doc.search("*").length
   end
 
   def test_kleene_star
     # bug noticed by raja bhatia
-    doc = Nokogiri.Hpricot("<span class='small'>1</span><div class='large'>2</div><div class='small'>3</div><span class='blue large'>4</span>")
+    doc = Nokogiri::Hpricot.HTML("<span class='small'>1</span><div class='large'>2</div><div class='small'>3</div><span class='blue large'>4</span>")
     assert_equal 2, doc.search("*[@class*='small']").length
     assert_equal 2, doc.search("*.small").length
     assert_equal 2, doc.search(".small").length
@@ -267,11 +266,11 @@ class TestParser < Nokogiri::TestCase
   end
 
   def test_empty_comment
-    doc = Nokogiri.Hpricot("<p><!----></p>")
+    doc = Nokogiri::Hpricot.HTML("<p><!----></p>")
     doc = doc.search('//body').first
     assert doc.children[0].children[0].comment?
 
-    doc = Nokogiri.Hpricot("<p><!-- --></p>")
+    doc = Nokogiri::Hpricot.HTML("<p><!-- --></p>")
     doc = doc.search('//body').first
     assert doc.children[0].children[0].comment?
   end
@@ -332,9 +331,9 @@ class TestParser < Nokogiri::TestCase
   ####
   # Modified.  Added question.  Don't care.
   def test_procins
-    doc = Nokogiri.Hpricot("<?php print('hello') ?>\n<?xml blah='blah'?>")
+    doc = Nokogiri::Hpricot.HTML("<?php print('hello') ?>\n<?xml blah='blah'?>")
     assert_equal "php", doc.children[1].target
-    assert_equal "blah='blah'?", doc.children[2].content
+    assert_equal "blah='blah'?", doc.children[2].content #"# quote added so emacs ruby-mode parser doesn't barf
   end
 
   ####

data/test/html/test_document.rb

@@ -12,6 +12,38 @@ module Nokogiri
         assert html.html?
       end
 
+      def test_relative_css
+        html = Nokogiri::HTML(<<-eohtml)
+        <html>
+          <body>
+            <div>
+              <p>inside div tag</p>
+            </div>
+            <p>outside div tag</p>
+          </body>
+        </html>
+        eohtml
+        set = html.search('div').search('p')
+        assert_equal(1, set.length)
+        assert_equal('inside div tag', set.first.inner_text)
+      end
+
+      def test_multi_css
+        html = Nokogiri::HTML(<<-eohtml)
+        <html>
+          <body>
+            <div>
+              <p>p tag</p>
+              <a>a tag</a>
+            </div>
+          </body>
+        </html>
+        eohtml
+        set = html.css('p, a')
+        assert_equal(2, set.length)
+        assert_equal ['a tag', 'p tag'].sort, set.map { |x| x.content }.sort
+      end
+
       def test_inner_text
         html = Nokogiri::HTML(<<-eohtml)
         <html>

data/test/test_convert_xpath.rb

@@ -151,6 +151,11 @@ class TestConvertXPath < Nokogiri::TestCase
     end
   end
 
+  def test_compat_mode_namespaces
+    assert_equal(".//*[name()='t:sam']", @NH.convert_to_xpath("//t:sam").first)
+    assert_equal(".//*[name()='t:sam'][@rel='bookmark'][1]", @NH.convert_to_xpath("//t:sam[@rel='bookmark'][1]").first)
+  end
+
   ##
   #  'and' is not supported by hpricot
 #   def test_and

data/test/test_css_cache.rb

@@ -0,0 +1,50 @@
+require File.expand_path(File.join(File.dirname(__FILE__), "helper"))
+
+require 'rubygems'
+require 'mocha'
+
+class TestCssCache < Nokogiri::TestCase
+
+  def setup
+    @css = "a1 > b2 > c3"
+    @parse_result = Nokogiri::CSS.parse(@css)
+    @to_xpath_result = @parse_result.map {|ast| ast.to_xpath}
+    assert Nokogiri::CSS::Parser.cache_on?
+  end
+
+  def teardown
+    Nokogiri::CSS::Parser.clear_cache
+    Nokogiri::CSS::Parser.set_cache true
+  end
+
+  [ false, true ].each do |cache_setting|
+    define_method "test_css_cache_#{cache_setting ? "true" : "false"}" do
+      times = cache_setting ? 1 : 6
+      Nokogiri::CSS::Parser.set_cache cache_setting
+      
+      Nokogiri::CSS::Parser.any_instance.expects(:parse).with(@css).returns(@parse_result).times(times)
+      Nokogiri::CSS::Node.any_instance.expects(:to_xpath).returns(@to_xpath_result).times(times)
+      
+      Nokogiri::CSS.xpath_for(@css)
+      Nokogiri::CSS.xpath_for(@css)
+      Nokogiri::CSS::Parser.xpath_for(@css)
+      Nokogiri::CSS::Parser.xpath_for(@css)
+      Nokogiri::CSS::Parser.new.xpath_for(@css)
+      Nokogiri::CSS::Parser.new.xpath_for(@css)
+    end
+
+    define_method "test_hpricot_cache_#{cache_setting ? "true" : "false"}" do
+      times = cache_setting ? 1 : 2
+      Nokogiri::CSS::Parser.set_cache cache_setting
+
+      nh = Nokogiri.Hpricot("<html></html>")
+      Nokogiri::CSS::Parser.any_instance.expects(:parse).with(@css).returns(@parse_result).times(times)
+      Nokogiri::CSS::Node.any_instance.expects(:to_xpath).returns(@to_xpath_result).times(times)
+
+      nh.convert_to_xpath(@css)
+      nh.convert_to_xpath(@css)
+    end
+  end
+
+
+end

data/test/xml/test_document.rb

@@ -30,6 +30,8 @@ module Nokogiri
           xml = Nokogiri::XML(f)
         }
         assert xml.xml?
+        set = xml.search('//employee')
+        assert set.length > 0
       end
 
       def test_search_on_empty_documents
@@ -38,6 +40,12 @@ module Nokogiri
         assert_equal 0, ns.length
       end
 
+      def test_bad_xpath_raises_syntax_error
+        assert_raises(XML::XPath::SyntaxError) {
+          @xml.xpath('\\')
+        }
+      end
+
       def test_new_document_collect_namespaces
         doc = Nokogiri::XML::Document.new
         assert_equal({}, doc.collect_namespaces)

data/vendor/hoe.rb

@@ -6,7 +6,16 @@ require 'rake/gempackagetask'
 require 'rake/rdoctask'
 require 'rake/testtask'
 require 'rbconfig'
-require 'rubyforge'
+require 'uri'
+
+if ENV['RUBYARCHDIR']
+  class RubyForge
+    VERSION = 'awesome'
+  end
+else
+  require 'rubyforge'
+end
+
 require 'yaml'
 
 begin
@@ -621,8 +630,6 @@ class Hoe
 
     Rake::RDocTask.new(:docs) do |rd|
       rd.main = "README.txt"
-      rd.options << '-d' if
-        `which dot` =~ /\/dot/ unless ENV['NODOT'] unless WINDOZE
       rd.rdoc_dir = 'doc'
       files = spec.files.grep(rdoc_pattern)
       files -= ['Manifest.txt']

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: nokogiri
 version: !ruby/object:Gem::Version 
-  version: 1.0.5
+  version: 1.0.6
 platform: ruby
 authors: 
 - Aaron Patterson
@@ -10,7 +10,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2008-11-11 00:00:00 -08:00
+date: 2008-11-14 00:00:00 -08:00
 default_executable: 
 dependencies: []
 
@@ -23,11 +23,13 @@ executables: []
 extensions: 
 - Rakefile
 extra_rdoc_files: 
+- History.ja.txt
 - History.txt
 - Manifest.txt
 - README.ja.txt
 - README.txt
 files: 
+- History.ja.txt
 - History.txt
 - Manifest.txt
 - README.ja.txt
@@ -46,6 +48,8 @@ files:
 - ext/nokogiri/xml_document.h
 - ext/nokogiri/xml_dtd.c
 - ext/nokogiri/xml_dtd.h
+- ext/nokogiri/xml_io.c
+- ext/nokogiri/xml_io.h
 - ext/nokogiri/xml_node.c
 - ext/nokogiri/xml_node.h
 - ext/nokogiri/xml_node_set.c
@@ -71,6 +75,7 @@ files:
 - lib/nokogiri/css/node.rb
 - lib/nokogiri/css/parser.rb
 - lib/nokogiri/css/parser.y
+- lib/nokogiri/css/syntax_error.rb
 - lib/nokogiri/css/tokenizer.rb
 - lib/nokogiri/css/tokenizer.rex
 - lib/nokogiri/css/xpath_visitor.rb
@@ -104,6 +109,7 @@ files:
 - lib/nokogiri/xml/syntax_error.rb
 - lib/nokogiri/xml/text.rb
 - lib/nokogiri/xml/xpath.rb
+- lib/nokogiri/xml/xpath/syntax_error.rb
 - lib/nokogiri/xml/xpath_context.rb
 - lib/nokogiri/xslt.rb
 - lib/nokogiri/xslt/stylesheet.rb
@@ -137,6 +143,7 @@ files:
 - test/html/test_builder.rb
 - test/html/test_document.rb
 - test/test_convert_xpath.rb
+- test/test_css_cache.rb
 - test/test_gc.rb
 - test/test_nokogiri.rb
 - test/test_reader.rb
@@ -193,6 +200,7 @@ test_files:
 - test/html/test_builder.rb
 - test/html/test_document.rb
 - test/test_convert_xpath.rb
+- test/test_css_cache.rb
 - test/test_gc.rb
 - test/test_nokogiri.rb
 - test/test_reader.rb