libxml-ruby 1.1.4 → 2.0.0

data/script/benchmark/throughput

@@ -0,0 +1,41 @@
+#!/usr/bin/env ruby
+
+require 'rubygems'
+require 'hpricot'
+require 'xml'
+require 'rexml/document'
+
+ITERATIONS = 3
+NESTED_ITERATIONS = 5
+
+def bm(name, filename, &block)
+  text = File.open(filename).read
+  length = text.length / 1024.0 / 1024.0
+  puts "#{filename}: #{name} (#{(length * 1024).round} kb)"
+  for j in 0 .. NESTED_ITERATIONS
+    s = Time.now.to_f
+    for i in 0 .. ITERATIONS
+      block.call(text)
+    end
+    timer = Time.now.to_f - s
+    puts "\t#{length * ITERATIONS / timer} MB/s"
+  end
+end
+
+def bm_suite(filenames)
+  filenames.each do |filename|    
+    bm("LIBXML THROUGHPUT:", filename) do |text|
+      XML::Document.file(filename)
+    end
+    
+    bm("HPRICOT THROUGHPUT:", filename) do |text|
+      Hpricot.XML(text)
+    end
+    
+    bm("REXML THROUGHPUT:", filename) do |text|
+      REXML::Document.new(text)
+    end
+  end
+end
+
+bm_suite("hamlet.xml")

data/script/test

@@ -0,0 +1,6 @@
+#!/usr/bin/env ruby
+
+%w{test lib ext/libxml}.each{ |path| $LOAD_PATH.unshift(path) }
+
+load './test/test_suite.rb'
+

data/test/etc_doc_to_s.rb

@@ -1,19 +1,21 @@
-require 'xml'
-
-1.times do |count|
-
-  xml_doc = XML::Document.new()
-  xml_doc.encoding = "UTF-8"
-  xml_doc.root = XML::Node.new("Request")
-
-  1000.times do |index|
-    xml_doc.root << node = XML::Node.new("row")
-    node["user_id"] = index.to_s
-    node << "600445"
-  end
-
-  xml_str = xml_doc.to_s
-  print "\r#{count}"
-  $stdout.flush
-end
-puts "\n"
+# encoding: UTF-8
+
+require './test_helper'
+
+1.times do |count|
+
+  xml_doc = XML::Document.new()
+  xml_doc.encoding = "UTF-8"
+  xml_doc.root = XML::Node.new("Request")
+
+  1000.times do |index|
+    xml_doc.root << node = XML::Node.new("row")
+    node["user_id"] = index.to_s
+    node << "600445"
+  end
+
+  xml_str = xml_doc.to_s
+  print "\r#{count}"
+  $stdout.flush
+end
+puts "\n"

data/test/ets_doc_file.rb

@@ -1,15 +1,17 @@
-require 'xml'
-
-# This is related to bug 8337, complaint is on amd64/fbsd
-# unknown if it happens on other amd64/os combos
-
-Process.setrlimit(Process::RLIMIT_NOFILE,10005)
-
-(1..10000).each{|time|
-  XML::Document.file(File.join(File.dirname(__FILE__),'ets_test.xml'))
-  if time % 100 == 0
-    print "\r#{time}"  
-    $stdout.flush
-  end
-}
-puts "\n"
+# encoding: UTF-8
+
+require './test_helper'
+
+# This is related to bug 8337, complaint is on amd64/fbsd
+# unknown if it happens on other amd64/os combos
+
+Process.setrlimit(Process::RLIMIT_NOFILE,10005)
+
+(1..10000).each{|time|
+  XML::Document.file(File.join(File.dirname(__FILE__),'ets_test.xml'))
+  if time % 100 == 0
+    print "\r#{time}"  
+    $stdout.flush
+  end
+}
+puts "\n"

data/test/ets_doc_to_s.rb

@@ -1,21 +1,23 @@
-require 'xml'
-
-100.times do |count|
-
-  xml_doc = XML::Document.new()
-  xml_doc.encoding = "UTF-8"
-  xml_doc.root = XML::Node.new("Request")
-
-  1000.times do |index|
-
-    xml_doc.root << node = XML::Node.new("row")
-    node["user_id"] = index.to_s
-    node << "600445"
-
-  end
-
-  xml_str = xml_doc.to_s
-  print "\r#{count}"
-  $stdout.flush
-end
-puts "\n"
+# encoding: UTF-8
+
+require './test_helper'
+
+100.times do |count|
+
+  xml_doc = XML::Document.new()
+  xml_doc.encoding = "UTF-8"
+  xml_doc.root = XML::Node.new("Request")
+
+  1000.times do |index|
+
+    xml_doc.root << node = XML::Node.new("row")
+    node["user_id"] = index.to_s
+    node << "600445"
+
+  end
+
+  xml_str = xml_doc.to_s
+  print "\r#{count}"
+  $stdout.flush
+end
+puts "\n"

data/test/ets_gpx.rb

@@ -1,26 +1,28 @@
-require 'libxml'
-
-100.times do
-  doc = XML::Document.new
-  doc.encoding = 'UTF-8'
-
-  root = XML::Node.new 'gpx'
-  root['version'] = '1.0'
-  root['creator'] = 'OpenStreetMap.org'
-  root['xmlns'] = "http://www.topografix.com/GPX/1/0/"
-
-  doc.root = root
-
-  track = XML::Node.new 'trk'
-  doc.root << track
-
-  trkseg = XML::Node.new 'trkseg'
-  track << trkseg
-
-  1.upto(1000) do |n|
-    trkpt = XML::Node.new 'trkpt'
-    trkpt['lat'] = n.to_s
-    trkpt['lon'] = n.to_s
-    trkseg << trkpt
-  end
-end
+# encoding: UTF-8
+
+require 'libxml'
+
+100.times do
+  doc = XML::Document.new
+  doc.encoding = 'UTF-8'
+
+  root = XML::Node.new 'gpx'
+  root['version'] = '1.0'
+  root['creator'] = 'OpenStreetMap.org'
+  root['xmlns'] = "http://www.topografix.com/GPX/1/0/"
+
+  doc.root = root
+
+  track = XML::Node.new 'trk'
+  doc.root << track
+
+  trkseg = XML::Node.new 'trkseg'
+  track << trkseg
+
+  1.upto(1000) do |n|
+    trkpt = XML::Node.new 'trkpt'
+    trkpt['lat'] = n.to_s
+    trkpt['lon'] = n.to_s
+    trkseg << trkpt
+  end
+end

data/test/ets_node_gc.rb

@@ -1,21 +1,23 @@
-require 'xml'
-
-# test of bug 13310, clears MEM2
-
-include GC
-
-inner = XML::Node.new('inner')
-save = nil
-1.times do
-  outer = XML::Node.new('outer')
-  outer.child = inner
-  1.times do
-    doc = XML::Document.new
-    doc.root = outer
-    # Uncomment the following line and it won't crash
-    save = doc
-  end
-  garbage_collect
-end
-garbage_collect
-puts inner
+# encoding: UTF-8
+
+require './test_helper'
+
+# test of bug 13310, clears MEM2
+
+include GC
+
+inner = XML::Node.new('inner')
+save = nil
+1.times do
+  outer = XML::Node.new('outer')
+  outer.child = inner
+  1.times do
+    doc = XML::Document.new
+    doc.root = outer
+    # Uncomment the following line and it won't crash
+    save = doc
+  end
+  garbage_collect
+end
+garbage_collect
+puts inner

data/test/ets_tsr.rb

@@ -1,9 +1,11 @@
-require 'xml'
-
-10_000.times {|n|
-j=XML::Node.new2(nil,"happy#{n}")
-  10.times {|r|
-    j1=XML::Node.new("happy#{r}","farts")
-    j.child=j1
-  }
-}
+# encoding: UTF-8
+
+require './test_helper'
+
+10_000.times {|n|
+j=XML::Node.new2(nil,"happy#{n}")
+  10.times {|r|
+    j1=XML::Node.new("happy#{r}","farts")
+    j.child=j1
+  }
+}

data/test/model/bands.iso-8859-1.xml

@@ -0,0 +1,5 @@
+<?xml version="1.0" encoding="iso-8859-1"?>
+<bands genre="metal">
+  <m�tley_cr�e country="us">M�tley Cr�e is an American heavy metal band formed in Los Angeles, California in 1981.</m�tley_cr�e>
+  <iron_maiden country="uk">Iron Maiden is a British heavy metal band formed in 1975.</iron_maiden>
+</bands>

data/test/model/bands.utf-8.xml

@@ -0,0 +1,5 @@
+<?xml version="1.0" encoding="utf-8"?>
+<bands genre="metal">
+  <mötley_crüe country="us">Mötley Crüe is an American heavy metal band formed in Los Angeles, California in 1981.</mötley_crüe>
+  <iron_maiden country="uk">Iron Maiden is a British heavy metal band formed in 1975.</iron_maiden>
+</bands>

data/test/rb-magic-comment.rb

@@ -0,0 +1,33 @@
+#!/usr/local/ruby/bin/ruby
+# encoding: UTF-8
+
+ARGV.reject! { |path| !File.file?(path) }
+(puts DATA.read.gsub("$0", File.basename($0)); exit 1) if ARGV.empty?
+
+ARGV.each do |path|
+  ls = IO.readlines(path)
+  ix = ls[0] !~ /^#!/ ? 0 : 1
+  next if ls[ix] =~ /#.*?coding\s*[:=]\s*\S/
+  ls.insert ix, "# encoding: UTF-8\n", ("\n" unless ls[ix] =~ /^#?\s*\n/)
+  open(path, 'w') { |f| f.write ls.join }
+end
+
+__END__
+Add the unicode magic comment to ruby files.
+
+* Files are modified in-place.
+* If a file already contains a magic comment, it's skipped.
+* Comment is added below the shebang if it exists, otherwise as the first line.
+
+Usage:
+  $0 file-1 [file-2 ... file-n]
+
+Examples:
+  # Add comment to standalone ruby files:
+  $0 foo.rb bar.rb
+
+  # Add comment to an entire ruby library or app:
+  find . -name '*.rb' -type f -print0 | xargs -0 $0
+
+  # Add comment to app files in a rails application:
+  find app config lib -name '*.rb' -type f -print0 | xargs -0 $0

data/test/tc_attr.rb

@@ -1,170 +1,181 @@
-require 'xml'
-require 'test/unit'
-
-class AttrNodeTest < Test::Unit::TestCase
-  def setup()
-    xp = XML::Parser.string(<<-EOS)
-    <CityModel
-      xmlns="http://www.opengis.net/examples"
-      xmlns:city="http://www.opengis.net/examples"
-      xmlns:gml="http://www.opengis.net/gml"
-      xmlns:xlink="http://www.w3.org/1999/xlink"
-      xmlns:xsi="http://www.w3.org/2000/10/XMLSchema-instance"
-      xsi:schemaLocation="http://www.opengis.net/examples city.xsd">
-      <type>City</type>
-      <cityMember name="Cambridge" 
-                  xlink:type="simple"
-                  xlink:title="Trinity Lane"
-                  xlink:href="http://www.foo.net/cgi-bin/wfs?FeatureID=C10239"
-                  gml:remoteSchema="city.xsd#xpointer(//complexType[@name='RoadType'])"/>
-    </CityModel>
-    EOS
-    
-    @doc = xp.parse
-  end
-  
-  def teardown()
-    @doc = nil
-  end
-  
-  def city_member
-    @doc.find('/city:CityModel/city:cityMember').first
-  end
-  
-  def test_types
-    attribute = city_member.attributes.get_attribute('name')
-    assert_instance_of(XML::Attr, attribute)
-    assert_equal('attribute', attribute.node_type_name)
-  end
-  
-  def test_name
-    attribute = city_member.attributes.get_attribute('name')
-    assert_equal('name', attribute.name)
-    
-    attribute = city_member.attributes.get_attribute('href')
-    assert_equal('href', attribute.name)
-    assert_equal('xlink', attribute.ns.prefix)
-    assert_equal('http://www.w3.org/1999/xlink', attribute.ns.href)
-    
-    attribute = city_member.attributes.get_attribute_ns('http://www.w3.org/1999/xlink', 'href')
-    assert_equal('href', attribute.name)
-    assert_equal('xlink', attribute.ns.prefix)
-    assert_equal('http://www.w3.org/1999/xlink', attribute.ns.href)
-  end
-  
-  def test_value
-    attribute = city_member.attributes.get_attribute('name')
-    assert_equal('Cambridge', attribute.value)
-    
-    attribute = city_member.attributes.get_attribute('href')
-    assert_equal('http://www.foo.net/cgi-bin/wfs?FeatureID=C10239', attribute.value)
-  end
-  
-  def test_set_value
-    attribute = city_member.attributes.get_attribute('name')
-    attribute.value = 'London'
-    assert_equal('London', attribute.value)
-    
-    attribute = city_member.attributes.get_attribute('href')
-    attribute.value = 'http://i.have.changed'
-    assert_equal('http://i.have.changed', attribute.value)
-  end
-  
-  def test_set_nil
-    attribute = city_member.attributes.get_attribute('name')
-    assert_raise(TypeError) do
-      attribute.value = nil
-    end
-  end
-  
-  def test_create
-    attributes = city_member.attributes
-    assert_equal(5, attributes.length)
-    
-    attr = XML::Attr.new(city_member, 'size', '50,000')
-    assert_instance_of(XML::Attr, attr)
-    
-    attributes = city_member.attributes
-    assert_equal(6, attributes.length)
-    
-    assert_equal(attributes['size'], '50,000')
-  end
-  
-  def test_create_on_node
-    attributes = city_member.attributes
-    assert_equal(5, attributes.length)
-    
-    attributes['country'] = 'England'
-    
-    attributes = city_member.attributes
-    assert_equal(6, attributes.length)
-    
-    assert_equal(attributes['country'], 'England')
-  end
-  
-  def test_create_ns
-    assert_equal(5, city_member.attributes.length)
-    
-    ns = XML::NS.new(city_member, 'my_namepace', 'http://www.mynamespace.com')
-    attr = XML::Attr.new(city_member, 'rating', 'rocks', ns)
-    assert_instance_of(XML::Attr, attr)
-    assert_equal('rating', attr.name)
-    assert_equal('rocks', attr.value)
-    
-    attributes = city_member.attributes
-    assert_equal(6, attributes.length)
-    
-    assert_equal('rocks', city_member['rating'])
-  end
-
-  def test_remove
-    attributes = city_member.attributes
-    assert_equal(5, attributes.length)
-    
-    attribute = attributes.get_attribute('name')
-    assert_not_nil(attribute.parent)
-    assert(attribute.parent?)
-    
-    attribute.remove!
-    assert_equal(4, attributes.length)
-    assert_nil(attribute.parent)
-    assert(!attribute.parent?)
-  end
-  
-  def test_first
-    attribute = city_member.attributes.first
-    assert_instance_of(XML::Attr, attribute)
-    assert_equal('name', attribute.name)
-    assert_equal('Cambridge', attribute.value)
-    
-    attribute = attribute.next
-    assert_instance_of(XML::Attr, attribute)
-    assert_equal('type', attribute.name)
-    assert_equal('simple', attribute.value)
-
-    attribute = attribute.next
-    assert_instance_of(XML::Attr, attribute)
-    assert_equal('title', attribute.name)
-    assert_equal('Trinity Lane', attribute.value)
-    
-    attribute = attribute.next
-    assert_instance_of(XML::Attr, attribute)
-    assert_equal('href', attribute.name)
-    assert_equal('http://www.foo.net/cgi-bin/wfs?FeatureID=C10239', attribute.value)
-
-    attribute = attribute.next
-    assert_instance_of(XML::Attr, attribute)
-    assert_equal('remoteSchema', attribute.name)
-    assert_equal("city.xsd#xpointer(//complexType[@name='RoadType'])", attribute.value)
-    
-    attribute = attribute.next
-    assert_nil(attribute)
-  end
-
-  def test_no_attributes
-    element = @doc.find('/city:CityModel/city:type').first
-
-    assert_not_nil(element.attributes)
-    assert_equal(0, element.attributes.length)
-  end
-end
+# encoding: UTF-8
+
+require './test_helper'
+require 'test/unit'
+
+class AttrNodeTest < Test::Unit::TestCase
+  def setup
+    xp = XML::Parser.string(<<-EOS)
+    <CityModel
+      xmlns="http://www.opengis.net/examples"
+      xmlns:city="http://www.opengis.net/examples"
+      xmlns:gml="http://www.opengis.net/gml"
+      xmlns:xlink="http://www.w3.org/1999/xlink"
+      xmlns:xsi="http://www.w3.org/2000/10/XMLSchema-instance"
+      xsi:schemaLocation="http://www.opengis.net/examples city.xsd">
+      <type>City</type>
+      <cityMember name="Cambridge" 
+                  xlink:type="simple"
+                  xlink:title="Trinity Lane"
+                  xlink:href="http://www.foo.net/cgi-bin/wfs?FeatureID=C10239"
+                  gml:remoteSchema="city.xsd#xpointer(//complexType[@name='RoadType'])"/>
+    </CityModel>
+    EOS
+    
+    @doc = xp.parse
+  end
+  
+  def teardown
+    @doc = nil
+  end
+  
+  def city_member
+    @doc.find('/city:CityModel/city:cityMember').first
+  end
+
+  def test_doc
+    assert_not_nil(@doc)
+    assert_equal(XML::Encoding::NONE, @doc.encoding)
+  end
+  
+  def test_types
+    attribute = city_member.attributes.get_attribute('name')
+    assert_instance_of(XML::Attr, attribute)
+    assert_equal('attribute', attribute.node_type_name)
+  end
+  
+  def test_name
+    attribute = city_member.attributes.get_attribute('name')
+    assert_equal('name', attribute.name)
+    assert_equal(Encoding::ASCII_8BIT, attribute.name.encoding) if defined?(Encoding)
+    
+    attribute = city_member.attributes.get_attribute('href')
+    assert_equal('href', attribute.name)
+    assert_equal('xlink', attribute.ns.prefix)
+    assert_equal('http://www.w3.org/1999/xlink', attribute.ns.href)
+    
+    attribute = city_member.attributes.get_attribute_ns('http://www.w3.org/1999/xlink', 'href')
+    assert_equal('href', attribute.name)
+    assert_equal('xlink', attribute.ns.prefix)
+    assert_equal('http://www.w3.org/1999/xlink', attribute.ns.href)
+  end
+  
+  def test_value
+    attribute = city_member.attributes.get_attribute('name')
+    assert_equal('Cambridge', attribute.value)
+    assert_equal(Encoding::ASCII_8BIT, attribute.value.encoding) if defined?(Encoding)
+    
+    attribute = city_member.attributes.get_attribute('href')
+    assert_equal('http://www.foo.net/cgi-bin/wfs?FeatureID=C10239', attribute.value)
+  end
+  
+  def test_set_value
+    attribute = city_member.attributes.get_attribute('name')
+    attribute.value = 'London'
+    assert_equal('London', attribute.value)
+    assert_equal(Encoding::ASCII_8BIT, attribute.value.encoding) if defined?(Encoding)
+    
+    attribute = city_member.attributes.get_attribute('href')
+    attribute.value = 'http://i.have.changed'
+    assert_equal('http://i.have.changed', attribute.value)
+    assert_equal(Encoding::ASCII_8BIT, attribute.value.encoding) if defined?(Encoding)
+  end
+  
+  def test_set_nil
+    attribute = city_member.attributes.get_attribute('name')
+    assert_raise(TypeError) do
+      attribute.value = nil
+    end
+  end
+  
+  def test_create
+    attributes = city_member.attributes
+    assert_equal(5, attributes.length)
+    
+    attr = XML::Attr.new(city_member, 'size', '50,000')
+    assert_instance_of(XML::Attr, attr)
+    
+    attributes = city_member.attributes
+    assert_equal(6, attributes.length)
+    
+    assert_equal(attributes['size'], '50,000')
+  end
+  
+  def test_create_on_node
+    attributes = city_member.attributes
+    assert_equal(5, attributes.length)
+    
+    attributes['country'] = 'England'
+    
+    attributes = city_member.attributes
+    assert_equal(6, attributes.length)
+    
+    assert_equal(attributes['country'], 'England')
+  end
+  
+  def test_create_ns
+    assert_equal(5, city_member.attributes.length)
+    
+    ns = XML::Namespace.new(city_member, 'my_namepace', 'http://www.mynamespace.com')
+    attr = XML::Attr.new(city_member, 'rating', 'rocks', ns)
+    assert_instance_of(XML::Attr, attr)
+    assert_equal('rating', attr.name)
+    assert_equal('rocks', attr.value)
+    
+    attributes = city_member.attributes
+    assert_equal(6, attributes.length)
+    
+    assert_equal('rocks', city_member['rating'])
+  end
+
+  def test_remove
+    attributes = city_member.attributes
+    assert_equal(5, attributes.length)
+    
+    attribute = attributes.get_attribute('name')
+    assert_not_nil(attribute.parent)
+    assert(attribute.parent?)
+    
+    attribute.remove!
+    assert_equal(4, attributes.length)
+    assert_nil(attribute.parent)
+    assert(!attribute.parent?)
+  end
+  
+  def test_first
+    attribute = city_member.attributes.first
+    assert_instance_of(XML::Attr, attribute)
+    assert_equal('name', attribute.name)
+    assert_equal('Cambridge', attribute.value)
+    
+    attribute = attribute.next
+    assert_instance_of(XML::Attr, attribute)
+    assert_equal('type', attribute.name)
+    assert_equal('simple', attribute.value)
+
+    attribute = attribute.next
+    assert_instance_of(XML::Attr, attribute)
+    assert_equal('title', attribute.name)
+    assert_equal('Trinity Lane', attribute.value)
+    
+    attribute = attribute.next
+    assert_instance_of(XML::Attr, attribute)
+    assert_equal('href', attribute.name)
+    assert_equal('http://www.foo.net/cgi-bin/wfs?FeatureID=C10239', attribute.value)
+
+    attribute = attribute.next
+    assert_instance_of(XML::Attr, attribute)
+    assert_equal('remoteSchema', attribute.name)
+    assert_equal("city.xsd#xpointer(//complexType[@name='RoadType'])", attribute.value)
+    
+    attribute = attribute.next
+    assert_nil(attribute)
+  end
+
+  def test_no_attributes
+    element = @doc.find('/city:CityModel/city:type').first
+
+    assert_not_nil(element.attributes)
+    assert_equal(0, element.attributes.length)
+  end
+end