nokogiri 1.4.1 → 1.4.2

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

@@ -4,16 +4,6 @@ module Nokogiri
       
       attr_accessor :cstruct # :nodoc:
 
-      def node_set # :nodoc:
-        ptr = cstruct[:nodesetval] if cstruct[:nodesetval]
-        ptr = LibXML.xmlXPathNodeSetCreate(nil) if ptr.null?
-
-        set = XML::NodeSet.new(@document)
-        set.cstruct = LibXML::XmlNodeSet.new(ptr)
-        set.document = @document
-        set
-      end
-
     end
   end
 end

data/lib/nokogiri/ffi/xml/xpath_context.rb

@@ -43,8 +43,24 @@ module Nokogiri
 
         xpath = XML::XPath.new
         xpath.cstruct = LibXML::XmlXpathObject.new(xpath_ptr)
-        xpath.document = cstruct[:doc]
-        xpath
+        xpath.document = cstruct.document.ruby_doc
+
+        case xpath.cstruct[:type]
+        when LibXML::XmlXpathObject::XPATH_NODESET
+          if xpath.cstruct[:nodesetval].null?
+            NodeSet.new(xpath.document)
+          else
+            NodeSet.wrap(xpath.cstruct[:nodesetval], xpath.document)
+          end
+        when LibXML::XmlXpathObject::XPATH_STRING
+          xpath.cstruct[:stringval]
+        when LibXML::XmlXpathObject::XPATH_NUMBER
+          xpath.cstruct[:floatval]
+        when LibXML::XmlXpathObject::XPATH_BOOLEAN
+          0 != xpath.cstruct[:boolval]
+        else
+          NodeSet.new(xpath.document)
+        end
       end
 
       def self.new(node) # :nodoc:
@@ -66,8 +82,8 @@ module Nokogiri
       def ruby_funcall(name, xpath_handler) # :nodoc:
         lambda do |ctx, nargs|
           parser_context = LibXML::XmlXpathParserContext.new(ctx)
-          context = parser_context.context
-          doc = context.doc.ruby_doc
+          context_cstruct = parser_context.context
+          document = context_cstruct.document.ruby_doc
 
           params = []
 
@@ -81,10 +97,7 @@ module Nokogiri
             when LibXML::XmlXpathObject::XPATH_NUMBER
               params.unshift obj[:floatval]
             when LibXML::XmlXpathObject::XPATH_NODESET
-              ns_ptr = LibXML::XmlNodeSet.new(obj[:nodesetval])
-              set = NodeSet.allocate
-              set.cstruct = ns_ptr
-              params.unshift set
+              params.unshift NodeSet.wrap(obj[:nodesetval], document)
             else
               char_ptr = params.unshift LibXML.xmlXPathCastToString(obj)
               string = char_ptr.read_string
@@ -110,7 +123,7 @@ module Nokogiri
           when NilClass.to_s
             ;
           when Array.to_s
-            node_set = XML::NodeSet.new(doc, result)
+            node_set = XML::NodeSet.new(document, result)
             LibXML.xmlXPathReturnNodeSet(
               ctx,
               LibXML.xmlXPathNodeSetMerge(nil, node_set.cstruct)

data/lib/nokogiri/ffi/xslt/stylesheet.rb

@@ -30,6 +30,9 @@ module Nokogiri
       end
 
       def transform(document, params=[]) # :nodoc:
+        params = params.to_a.flatten if params.is_a?(Hash)
+        raise(TypeError) unless params.is_a?(Array)
+
         param_arr = FFI::MemoryPointer.new(:pointer, params.length + 1, false)
 
         # Keep the MemoryPointer instances alive until after the call

data/lib/nokogiri/html/document.rb

@@ -9,7 +9,7 @@ module Nokogiri
           node['http-equiv'] =~ /Content-Type/i
         }
 
-        /charset\s*=\s*([\w\d-]+)/i.match(meta['content'])[1]
+        /charset\s*=\s*([\w-]+)/i.match(meta['content'])[1]
       end
 
       ###
@@ -49,7 +49,7 @@ module Nokogiri
       ####
       # Create a Nokogiri::XML::DocumentFragment from +tags+
       def fragment tags = nil
-        DocumentFragment.new(self, tags)
+        DocumentFragment.new(self, tags, self.root)
       end
 
       class << self
@@ -68,7 +68,9 @@ module Nokogiri
           yield options if block_given?
 
           if string_or_io.respond_to?(:encoding)
-            encoding ||= string_or_io.encoding.name
+            unless string_or_io.encoding.name == "ASCII-8BIT"
+              encoding ||= string_or_io.encoding.name
+            end
           end
 
           if string_or_io.respond_to?(:read)

data/lib/nokogiri/html/document_fragment.rb

@@ -1,15 +1,36 @@
 module Nokogiri
   module HTML
     class DocumentFragment < Nokogiri::XML::DocumentFragment
-
-      class << self
-        ####
-        # Create a Nokogiri::XML::DocumentFragment from +tags+
-        def parse tags
-          self.new(HTML::Document.new, tags)
-        end
+      ####
+      # Create a Nokogiri::XML::DocumentFragment from +tags+
+      def self.parse tags
+        doc = HTML::Document.new
+        doc.encoding = 'UTF-8'
+        self.new(doc, tags)
       end
 
+      def initialize document, tags = nil, ctx = nil
+        return self unless tags
+
+        children = if ctx
+                     ctx.parse("<div>#{tags.strip}</div>").first.children
+                   else
+                     ###
+                     # This is a horrible hack, but I don't care
+                     if tags.strip =~ /^<body/i
+                       path = "/html/body"
+                     else
+                       path = "/html/body/node()"
+                     end
+
+                     HTML::Document.parse(
+                       "<html><body>#{tags.strip}</body></html>",
+                       nil,
+                       document.encoding
+                     ).xpath(path)
+                   end
+        children.each { |child| child.parent = self }
+      end
     end
   end
 end

data/lib/nokogiri/version.rb

@@ -1,11 +1,15 @@
 module Nokogiri
   # The version of Nokogiri you are using
-  VERSION = '1.4.1'
+  VERSION = '1.4.2'
 
   # More complete version information about libxml
   VERSION_INFO = {}
   VERSION_INFO['warnings']              = []
-  VERSION_INFO['nokogiri']              = VERSION
+  VERSION_INFO['nokogiri']              = Nokogiri::VERSION
+  VERSION_INFO['ruby']                  = {}
+  VERSION_INFO['ruby']['version']       = ::RUBY_VERSION
+  VERSION_INFO['ruby']['platform']      = ::RUBY_PLATFORM
+  VERSION_INFO['ruby']['jruby']         = ::JRUBY_VERSION if RUBY_PLATFORM == "java"
   if defined?(LIBXML_VERSION)
     VERSION_INFO['libxml']              = {}
     VERSION_INFO['libxml']['binding']   = 'extension'

data/lib/nokogiri/version_warning.rb

@@ -1,11 +1,14 @@
 module Nokogiri
-  if self.is_2_6_16? && !defined?(I_KNOW_I_AM_USING_AN_OLD_AND_BUGGY_VERSION_OF_LIBXML2)
-    warn <<-eom
+  if self.is_2_6_16?
+    VERSION_INFO['warnings'] << "libxml 2.6.16 is old and buggy."
+    if !defined?(I_KNOW_I_AM_USING_AN_OLD_AND_BUGGY_VERSION_OF_LIBXML2)
+      warn <<-eom
 HI.  You're using libxml2 version 2.6.16 which is over 4 years old and has
 plenty of bugs.  We suggest that for maximum HTML/XML parsing pleasure, you
 upgrade your version of libxml2 and re-install nokogiri.  If you like using
 libxml2 version 2.6.16, but don't like this warning, please define the constant
 I_KNOW_I_AM_USING_AN_OLD_AND_BUGGY_VERSION_OF_LIBXML2 before requring nokogiri.
-    eom
+      eom
+    end
   end
 end

data/lib/nokogiri/xml.rb

@@ -1,7 +1,6 @@
 require 'nokogiri/xml/pp'
 require 'nokogiri/xml/parse_options'
 require 'nokogiri/xml/sax'
-require 'nokogiri/xml/fragment_handler'
 require 'nokogiri/xml/node'
 require 'nokogiri/xml/attribute_decl'
 require 'nokogiri/xml/element_decl'
@@ -11,6 +10,7 @@ require 'nokogiri/xml/namespace'
 require 'nokogiri/xml/attr'
 require 'nokogiri/xml/dtd'
 require 'nokogiri/xml/cdata'
+require 'nokogiri/xml/text'
 require 'nokogiri/xml/document'
 require 'nokogiri/xml/document_fragment'
 require 'nokogiri/xml/processing_instruction'

data/lib/nokogiri/xml/builder.rb

@@ -194,6 +194,36 @@ module Nokogiri
     #   </root>
     #
     # Note the "foo:object" tag.
+    #
+    # == Document Types
+    #
+    # To create a document type (DTD), access use the Builder#doc method to get
+    # the current context document.  Then call Node#create_internal_subset to
+    # create the DTD node.
+    #
+    # For example, this Ruby:
+    #
+    #   builder = Nokogiri::XML::Builder.new do |xml|
+    #     xml.doc.create_internal_subset(
+    #       'html',
+    #       "-//W3C//DTD HTML 4.01 Transitional//EN",
+    #       "http://www.w3.org/TR/html4/loose.dtd"
+    #     )
+    #     xml.root do
+    #       xml.foo
+    #     end
+    #   end
+    #   
+    #   puts builder.to_xml
+    #
+    # Will output this xml:
+    #
+    #   <?xml version="1.0"?>
+    #   <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
+    #   <root>
+    #     <foo/>
+    #   </root>
+    #
     class Builder
       # The current Document object being built
       attr_accessor :doc
@@ -275,8 +305,7 @@ module Nokogiri
       ###
       # Create a CDATA Node with content of +string+
       def cdata string
-        node = Nokogiri::XML::CDATA.new(@doc, string.to_s)
-        insert(node)
+        insert(doc.create_cdata(string))
       end
 
       ###
@@ -311,29 +340,12 @@ module Nokogiri
         if @context && @context.respond_to?(method)
           @context.send(method, *args, &block)
         else
-          node = @doc.create_element(method.to_s.sub(/[_!]$/, '')) { |n|
+          node = @doc.create_element(method.to_s.sub(/[_!]$/, ''),*args) { |n|
             # Set up the namespace
             if @ns
               n.namespace = @ns
               @ns = nil
             end
-
-            args.each do |arg|
-              case arg
-              when Hash
-                arg.each { |k,v|
-                  key = k.to_s
-                  if key =~ /^xmlns(:\w+)?$/
-                    ns_name = key.split(":", 2)[1]
-                    n.add_namespace_definition ns_name, v
-                    next
-                  end
-                  n[k.to_s] = v.to_s
-                }
-              else
-                n.content = arg
-              end
-            end
           }
           insert(node, &block)
         end
@@ -345,14 +357,15 @@ module Nokogiri
       def insert(node, &block)
         node.parent = @parent
         if block_given?
-          @parent = node
+          old_parent = @parent
+          @parent    = node
           @arity ||= block.arity
           if @arity <= 0
             instance_eval(&block)
           else
             block.call(self)
           end
-          @parent = node.parent
+          @parent = old_parent
         end
         NodeBuilder.new(node, self)
       end

data/lib/nokogiri/xml/document.rb

@@ -1,6 +1,6 @@
 module Nokogiri
   module XML
-    ####
+    ##
     # Nokogiri::XML::Document is the main entry point for dealing with
     # XML documents.  The Document is created by parsing an XML document.
     # See Nokogiri.XML()
@@ -8,7 +8,7 @@ module Nokogiri
     # For searching a Document, see Nokogiri::XML::Node#css and
     # Nokogiri::XML::Node#xpath
     class Document < Node
-      ###
+      ##
       # Parse an XML file.  +thing+ may be a String, or any object that
       # responds to _read_ and _close_ such as an IO, or StringIO.
       # +url+ is resource where this document is located.  +encoding+ is the
@@ -36,13 +36,39 @@ module Nokogiri
       # A list of Nokogiri::XML::SyntaxError found when parsing a document
       attr_accessor :errors
 
-      def initialize *args
+      def initialize *args # :nodoc:
+        @errors     = []
         @decorators = nil
       end
 
-      # Create an element with +name+
-      def create_element name, &block
-        Nokogiri::XML::Element.new(name, self, &block)
+      ##
+      # Create an element with +name+, and optionally setting the content and attributes.
+      #
+      #   doc.create_element "div" # <div></div>
+      #   doc.create_element "div", :class => "container" # <div class='container'></div>
+      #   doc.create_element "div", "contents" # <div>contents</div>
+      #   doc.create_element "div", "contents", :class => "container" # <div class='container'>contents</div>
+      #   doc.create_element "div" { |node| node['class'] = "container" } # <div class='container'></div>
+      #
+      def create_element name, *args, &block
+        elm = Nokogiri::XML::Element.new(name, self, &block)
+        args.each do |arg|
+          case arg
+          when Hash
+            arg.each { |k,v|
+              key = k.to_s
+              if key =~ /^xmlns(:\w+)?$/
+                ns_name = key.split(":", 2)[1]
+                elm.add_namespace_definition ns_name, v
+                next
+              end
+              elm[k.to_s] = v.to_s
+            }
+          else
+            elm.content = arg
+          end
+        end
+        elm
       end
 
       # Create a text node with +text+
@@ -50,6 +76,11 @@ module Nokogiri
         Nokogiri::XML::Text.new(text.to_s, self, &block)
       end
 
+      # Create a CDATA element containing +text+
+      def create_cdata text
+        Nokogiri::XML::CDATA.new(self, text.to_s)
+      end
+
       # The name of this document.  Always returns "document"
       def name
         'document'
@@ -60,7 +91,7 @@ module Nokogiri
         self
       end
 
-      ###
+      ##
       # Recursively get all namespaces from this node and its subtree and
       # return them as a hash.
       #
@@ -94,7 +125,7 @@ module Nokogiri
         @decorators[key] ||= []
       end
 
-      ###
+      ##
       # Validate this Document against it's DTD.  Returns a list of errors on
       # the document or +nil+ when there is no DTD.
       def validate
@@ -102,7 +133,7 @@ module Nokogiri
         internal_subset.validate self
       end
 
-      ###
+      ##
       # Explore a document with shortcut methods.
       def slop!
         unless decorators(XML::Node).include? Nokogiri::Decorators::Slop
@@ -113,7 +144,7 @@ module Nokogiri
         self
       end
 
-      ###
+      ##
       # Apply any decorators to +node+
       def decorate node
         return unless @decorators
@@ -131,16 +162,17 @@ module Nokogiri
         root ? root.namespaces : {}
       end
 
-      ####
+      ##
       # Create a Nokogiri::XML::DocumentFragment from +tags+
       # Returns an empty fragment if +tags+ is nil.
       def fragment tags = nil
-        DocumentFragment.new(self, tags)
+        DocumentFragment.new(self, tags, self.root)
       end
 
       undef_method :swap, :parent, :namespace, :default_namespace=
       undef_method :add_namespace_definition, :attributes
       undef_method :namespace_definitions, :line, :add_namespace
+      undef_method :parse, :in_context
 
       def add_child child
         raise "Document already has a root node" if root

data/lib/nokogiri/xml/document_fragment.rb

@@ -1,18 +1,22 @@
 module Nokogiri
   module XML
     class DocumentFragment < Nokogiri::XML::Node
-      def initialize document, tags=nil
-        if tags
-          if self.kind_of?(Nokogiri::HTML::DocumentFragment)
-            HTML::SAX::Parser.new(FragmentHandler.new(self, tags)).parse(tags)
-          else
-            wrapped = "<div>#{tags.strip}</div>"
-            XML::SAX::Parser.new(FragmentHandler.new(self, wrapped)).parse(wrapped)
-            div = self.child
-            div.children.each { |child| child.parent = self }
-            div.unlink
-          end
-        end
+      ##
+      #  Create a new DocumentFragment from +tags+.
+      #
+      #  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
+        return self unless tags
+
+        children = if ctx
+                     ctx.parse(tags.strip)
+                   else
+                     XML::Document.parse("<root>#{tags.strip}</root>") \
+                       .xpath("/root/node()")
+                   end
+        children.each { |child| child.parent = self }
       end
 
       ###