nokogiri 1.3.3-x86-mingw32 → 1.4.0-x86-mingw32

data/lib/nokogiri/xml/pp.rb

@@ -0,0 +1,2 @@
+require 'nokogiri/xml/pp/node'
+require 'nokogiri/xml/pp/character_data'

data/lib/nokogiri/xml/pp/character_data.rb

@@ -0,0 +1,18 @@
+module Nokogiri
+  module XML
+    module PP
+      module CharacterData
+        def pretty_print pp # :nodoc:
+          nice_name = self.class.name.split('::').last
+          pp.group(2, "#(#{nice_name} ", ')') do
+            pp.pp text
+          end
+        end
+
+        def inspect # :nodoc:
+          "#<#{self.class.name}:#{sprintf("0x%x",object_id)} #{text.inspect}>"
+        end
+      end
+    end
+  end
+end

data/lib/nokogiri/xml/pp/node.rb

@@ -0,0 +1,56 @@
+module Nokogiri
+  module XML
+    module PP
+      module Node
+        def inspect # :nodoc:
+          attributes = inspect_attributes.reject { |x|
+            begin
+              attribute = send x
+              !attribute || (attribute.respond_to?(:empty?) && attribute.empty?)
+            rescue NoMethodError
+              true
+            end
+          }.map { |attribute|
+            "#{attribute.to_s.sub(/_\w+/, 's')}=#{send(attribute).inspect}"
+          }.join ' '
+          "#<#{self.class.name}:#{sprintf("0x%x", object_id)} #{attributes}>"
+        end
+
+        def pretty_print pp # :nodoc:
+          nice_name = self.class.name.split('::').last
+          pp.group(2, "#(#{nice_name}:#{sprintf("0x%x", object_id)} {", '})') do
+
+            pp.breakable
+            attrs = inspect_attributes.map { |t|
+              [t, send(t)] if respond_to?(t)
+            }.compact.find_all { |x|
+              if x.last
+                if [:attribute_nodes, :children].include? x.first
+                  !x.last.empty?
+                else
+                  true
+                end
+              end
+            }
+
+            pp.seplist(attrs) do |v|
+              if [:attribute_nodes, :children].include? v.first
+                pp.group(2, "#{v.first.to_s.sub(/_\w+$/, 's')} = [", "]") do
+                  pp.breakable
+                  pp.seplist(v.last) do |item|
+                    pp.pp item
+                  end
+                end
+              else
+                pp.text "#{v.first} = "
+                pp.pp v.last
+              end
+            end
+            pp.breakable
+
+          end
+        end
+      end
+    end
+  end
+end

data/lib/nokogiri/xml/reader.rb

@@ -54,6 +54,14 @@ module Nokogiri
         }.flatten)].merge(namespaces || {})
       end
 
+      ###
+      # Get a list of attributes for the current node
+      def attribute_nodes
+        nodes = attr_nodes
+        nodes.each { |v| v.instance_variable_set(:@_r, self) }
+        nodes
+      end
+
       ###
       # Move the cursor through the document yielding each node to the block
       def each(&block)

data/lib/nokogiri/xml/sax.rb

@@ -1,4 +1,4 @@
 require 'nokogiri/xml/sax/document'
-require 'nokogiri/xml/sax/legacy_handlers'
+require 'nokogiri/xml/sax/parser_context'
 require 'nokogiri/xml/sax/parser'
 require 'nokogiri/xml/sax/push_parser'

data/lib/nokogiri/xml/sax/document.rb

@@ -68,6 +68,11 @@ module Nokogiri
       # You can use this event handler for any SAX style parser included with
       # Nokogiri.  See Nokogiri::XML::SAX, and Nokogiri::HTML::SAX.
       class Document
+        ###
+        # Called when an XML declaration is parsed
+        def xmldecl version, encoding, standalone
+        end
+
         ###
         # Called when document starts parsing
         def start_document
@@ -97,7 +102,16 @@ module Nokogiri
         # +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
         end
 
         ###
@@ -106,6 +120,9 @@ module Nokogiri
         # +prefix+ is the namespace prefix associated with the element
         # +uri+ is the associated namespace URI
         def end_element_namespace name, prefix = nil, uri = nil
+          ###
+          # Deal with SAX v1 interface
+          end_element [prefix, name].compact.join(':')
         end
 
         ###

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

@@ -67,7 +67,7 @@ module Nokogiri
         attr_accessor :encoding
 
         # Create a new Parser with +doc+ and +encoding+
-        def initialize(doc = Nokogiri::XML::SAX::Document.new, encoding = 'ASCII')
+        def initialize doc = Nokogiri::XML::SAX::Document.new, encoding = 'UTF-8'
           @encoding = encoding
           @document = doc
           @warned   = false
@@ -76,11 +76,11 @@ module Nokogiri
         ###
         # Parse given +thing+ which may be a string containing xml, or an
         # IO object.
-        def parse thing
+        def parse thing, &block
           if thing.respond_to?(:read) && thing.respond_to?(:close)
-            parse_io(thing)
+            parse_io(thing, &block)
           else
-            parse_memory(thing)
+            parse_memory(thing, &block)
           end
         end
 
@@ -88,7 +88,9 @@ module Nokogiri
         # Parse given +io+
         def parse_io io, encoding = 'ASCII'
           @encoding = encoding
-          native_parse_io io, ENCODINGS[@encoding] || ENCODINGS['ASCII']
+          ctx = ParserContext.io(io, ENCODINGS[encoding])
+          yield ctx if block_given?
+          ctx.parse_with self
         end
 
         ###
@@ -97,11 +99,16 @@ module Nokogiri
           raise ArgumentError unless filename
           raise Errno::ENOENT unless File.exists?(filename)
           raise Errno::EISDIR if File.directory?(filename)
-          native_parse_file filename
+          ctx = ParserContext.file filename
+          yield ctx if block_given?
+          ctx.parse_with self
         end
 
-        private
-        include Nokogiri::XML::SAX::LegacyHandlers
+        def parse_memory data
+          ctx = ParserContext.memory data
+          yield ctx if block_given?
+          ctx.parse_with self
+        end
       end
     end
   end

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

@@ -0,0 +1,16 @@
+module Nokogiri
+  module XML
+    module SAX
+      ###
+      # Context for XML SAX parsers.  This class is usually not instantiated
+      # by the user.  Instead, you should be looking at
+      # Nokogiri::XML::SAX::Parser
+      class ParserContext
+        def self.new thing, encoding = 'UTF-8'
+          [:read, :close].all? { |x| thing.respond_to?(x) } ?
+            io(thing, Parser::ENCODINGS[encoding]) : memory(thing)
+        end
+      end
+    end
+  end
+end

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

@@ -54,9 +54,6 @@ module Nokogiri
         def finish
           write '', true
         end
-
-        private
-        include Nokogiri::XML::SAX::LegacyHandlers
       end
     end
   end

data/lib/nokogiri/xml/syntax_error.rb

@@ -4,6 +4,10 @@ module Nokogiri
     # This class provides information about XML SyntaxErrors.  These
     # exceptions are typically stored on Nokogiri::XML::Document#errors.
     class SyntaxError < ::Nokogiri::SyntaxError
+      def initialize message
+        self.message = message
+      end
+
       ###
       # return true if this is a non error
       def none?

data/lib/nokogiri/xslt/stylesheet.rb

@@ -18,7 +18,7 @@ module Nokogiri
       # +params+ is an array of strings used as XSLT parameters.
       # returns serialized document
       def apply_to document, params = []
-        serialize(transform(document, params))
+        transform(document, params).to_s
       end
     end
   end

data/test/css/test_nthiness.rb

@@ -1,4 +1,4 @@
-require File.expand_path(File.join(File.dirname(__FILE__), '..', "helper"))
+require "helper"
 
 module Nokogiri
   module CSS

data/test/css/test_parser.rb

@@ -1,4 +1,4 @@
-require File.expand_path(File.join(File.dirname(__FILE__), '..', "helper"))
+require "helper"
 
 module Nokogiri
   module CSS

data/test/css/test_tokenizer.rb

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-require File.expand_path(File.join(File.dirname(__FILE__), '..', "helper"))
+require "helper"
 
 module Nokogiri
   module CSS

data/test/css/test_xpath_visitor.rb

@@ -1,4 +1,4 @@
-require File.expand_path(File.join(File.dirname(__FILE__), '..', "helper"))
+require "helper"
 
 module Nokogiri
   module CSS

data/test/ffi/test_document.rb

@@ -1,4 +1,4 @@
-require File.expand_path(File.join(File.dirname(__FILE__), "..", "helper"))
+require "helper"
 
 if defined?(Nokogiri::LibXML)
 

data/test/files/shift_jis.html

@@ -0,0 +1,10 @@
+<html>
+  <head>
+    <meta http-equiv="Content-Type" content="text/html; charset=Shift_JIS">
+    <title>����ɂ��́I</title>
+  </head>
+  <body>
+    <h1>This is a Shift_JIS File</h1>
+    <h2>����ɂ��́I</h2>
+  </body>
+</html>

data/test/files/staff.dtd

@@ -0,0 +1,10 @@
+<!ENTITY ent1 "es">
+<!ENTITY ent2 "1900 Dallas Road">
+<!ENTITY ent3 "Texas">
+<!ENTITY ent4 "<entElement domestic='Yes'>Element data</entElement><?PItarget PIdata?>">
+<!ENTITY ent5 PUBLIC "entityURI" "entityFile" NDATA notation1>
+<!ENTITY ent1 "This entity should be discarded">
+<!ELEMENT br EMPTY>
+<!ATTLIST br width CDATA "0">
+<!NOTATION notation1 PUBLIC "notation1File">
+<!NOTATION notation2 SYSTEM "notation2File">

data/test/helper.rb

@@ -1,8 +1,10 @@
-Process.setrlimit(Process::RLIMIT_CORE, Process::RLIM_INFINITY) unless RUBY_PLATFORM =~ /(java|mswin|mingw)/i
+#Process.setrlimit(Process::RLIMIT_CORE, Process::RLIM_INFINITY) unless RUBY_PLATFORM =~ /(java|mswin|mingw)/i
 $VERBOSE = true
 require 'rubygems'
 require 'test/unit'
+require 'fileutils'
 require 'tempfile'
+require 'pp'
 
 %w(../lib ../ext).each do |path|
   $LOAD_PATH.unshift(File.expand_path(File.join(File.dirname(__FILE__), path)))
@@ -14,7 +16,7 @@ warn "#{__FILE__}:#{__LINE__}: libxml version info: #{Nokogiri::VERSION_INFO.ins
 
 module Nokogiri
   class TestCase < Test::Unit::TestCase
-    ASSETS_DIR      = File.join(File.dirname(__FILE__), 'files')
+    ASSETS_DIR      = File.expand_path File.join(File.dirname(__FILE__), 'files')
     XML_FILE        = File.join(ASSETS_DIR, 'staff.xml')
     XSLT_FILE       = File.join(ASSETS_DIR, 'staff.xslt')
     EXSLT_FILE      = File.join(ASSETS_DIR, 'exslt.xslt')
@@ -22,6 +24,7 @@ module Nokogiri
     HTML_FILE       = File.join(ASSETS_DIR, 'tlm.html')
     NICH_FILE       = File.join(ASSETS_DIR, '2ch.html')
     SHIFT_JIS_XML   = File.join(ASSETS_DIR, 'shift_jis.xml')
+    SHIFT_JIS_HTML  = File.join(ASSETS_DIR, 'shift_jis.html')
     PO_XML_FILE     = File.join(ASSETS_DIR, 'po.xml')
     PO_SCHEMA_FILE  = File.join(ASSETS_DIR, 'po.xsd')
     ADDRESS_SCHEMA_FILE = File.join(ASSETS_DIR, 'address_book.rlx')
@@ -39,7 +42,7 @@ module Nokogiri
     def teardown
       if ENV['NOKOGIRI_GC']
         STDOUT.putc '!'
-        GC.start 
+        GC.start
       end
     end
 
@@ -63,6 +66,12 @@ module Nokogiri
         attr_reader :end_elements, :end_document_called
         attr_reader :data, :comments, :cdata_blocks, :start_elements_namespace
         attr_reader :errors, :warnings, :end_elements_namespace
+        attr_reader :xmldecls
+
+        def xmldecl version, encoding, standalone
+          @xmldecls = [version, encoding, standalone].compact
+          super
+        end
 
         def start_document
           @start_document_called = true

data/test/html/sax/test_parser.rb

@@ -1,4 +1,4 @@
-require File.expand_path(File.join(File.dirname(__FILE__), '..', '..', "helper"))
+require "helper"
 
 module Nokogiri
   module HTML

data/test/html/sax/test_parser_context.rb

@@ -0,0 +1,48 @@
+# -*- coding: utf-8 -*-
+
+require "helper"
+
+module Nokogiri
+  module HTML
+    module SAX
+      class TestParserContext < Nokogiri::SAX::TestCase
+        def test_from_io
+          assert_nothing_raised do
+            ctx = ParserContext.new StringIO.new('fo'), 'UTF-8'
+          end
+        end
+
+        def test_from_string
+          assert_nothing_raised do
+            ctx = ParserContext.new 'blah blah'
+          end
+        end
+
+        def test_parse_with
+          ctx = ParserContext.new 'blah'
+          assert_raises ArgumentError do
+            ctx.parse_with nil
+          end
+        end
+
+        def test_parse_with_sax_parser
+          assert_nothing_raised do
+            xml = "<root />"
+            ctx = ParserContext.new xml
+            parser = Parser.new Doc.new
+            ctx.parse_with parser
+          end
+        end
+
+        def test_from_file
+          assert_nothing_raised do
+            ctx = ParserContext.file HTML_FILE, 'UTF-8'
+            parser = Parser.new Doc.new
+            ctx.parse_with parser
+          end
+        end
+      end
+    end
+  end
+end
+

data/test/html/test_builder.rb

@@ -1,8 +1,14 @@
-require File.expand_path(File.join(File.dirname(__FILE__), '..', "helper"))
+require "helper"
 
 module Nokogiri
   module HTML
     class TestBuilder < Nokogiri::TestCase
+      def test_top_level_function_builds
+        foo = nil
+        Nokogiri() { |xml| foo = xml }
+        assert_instance_of Nokogiri::HTML::Builder, foo
+      end
+
       def test_builder_with_explicit_tags
         html_doc = Nokogiri::HTML::Builder.new {
           div.slide(:class => 'another_class') {
@@ -119,7 +125,7 @@ module Nokogiri
         assert_equal('<html><body><b>bold tag</b></body></html>',
                      builder.doc.root.to_html.chomp.gsub(/>\s*</, '><'))
       end
-      
+
       def test_instance_eval_with_delegation_to_block_context
         class << self
           def foo

data/test/html/test_document.rb

@@ -1,4 +1,4 @@
-require File.expand_path(File.join(File.dirname(__FILE__), '..', "helper"))
+require "helper"
 
 module Nokogiri
   module HTML
@@ -8,6 +8,11 @@ module Nokogiri
         @html = Nokogiri::HTML.parse(File.read(HTML_FILE))
       end
 
+      def test_fragment
+        fragment = @html.fragment
+        assert_equal 0, fragment.children.length
+      end
+
       def test_document_takes_config_block
         options = nil
         Nokogiri::HTML(File.read(HTML_FILE), HTML_FILE) do |cfg|
@@ -66,8 +71,12 @@ module Nokogiri
         assert_equal @html.to_s, html.to_s
       end
 
+      ###
+      # Nokogiri::HTML returns an empty Document when given a blank string GH#11
       def test_empty_string_returns_empty_doc
         doc = Nokogiri::HTML('')
+        assert_instance_of Nokogiri::HTML::Document, doc
+        assert_nil doc.root
       end
 
       unless %w[2 6] === LIBXML_VERSION.split('.')[0..1]
@@ -109,6 +118,14 @@ module Nokogiri
         assert_equal 'EUC-JP', @html.meta_encoding
       end
 
+      def test_meta_encoding_without_head
+        html = Nokogiri::HTML('<html><body>foo</body></html>')
+        assert_nil html.meta_encoding
+
+        html.meta_encoding = 'EUC-JP'
+        assert_nil html.meta_encoding
+      end
+
       def test_root_node_parent_is_document
         parent = @html.root.parent
         assert_equal @html, parent
@@ -295,6 +312,11 @@ module Nokogiri
         assert_equal('<p>Helloworld!</p>', node.inner_html.gsub(/\s/, ''))
       end
 
+      def test_round_trip
+        doc = Nokogiri::HTML(@html.inner_html)
+        assert_equal @html.root.to_html, doc.root.to_html
+      end
+
       def test_fragment_contains_text_node
         fragment = Nokogiri::HTML.fragment('fooo')
         assert_equal 1, fragment.children.length