libxml-ruby 2.2.2 → 2.3.0

data/test/c14n/result/1-1-without-comments/example-5

@@ -0,0 +1,3 @@
+<doc attrExtEnt="entExt">
+   Hello, world!
+</doc>

data/test/c14n/result/1-1-without-comments/example-6

@@ -0,0 +1 @@
+<doc>©</doc>

data/test/c14n/result/1-1-without-comments/example-7

@@ -0,0 +1 @@
+<e1 xmlns="http://www.ietf.org" xmlns:w3c="http://www.w3.org"><e3 xmlns="" id="E3" xml:space="preserve"></e3></e1>

data/test/c14n/result/1-1-without-comments/example-8

@@ -0,0 +1 @@
+<e1 xmlns="http://www.ietf.org" xmlns:w3c="http://www.w3.org" xml:base="something/else"><e3 xmlns="" id="E3" xml:base="bar/foo" xml:space="preserve"></e3></e1>

data/test/c14n/result/with-comments/example-1

@@ -0,0 +1,6 @@
+<?xml-stylesheet href="doc.xsl"
+   type="text/xsl"   ?>
+<doc>Hello, world!<!-- Comment 1 --></doc>
+<?pi-without-data?>
+<!-- Comment 2 -->
+<!-- Comment 3 -->

data/test/c14n/result/with-comments/example-2

@@ -0,0 +1,11 @@
+<doc>
+   <clean>   </clean>
+   <dirty>   A   B   </dirty>
+   <mixed>
+      A
+      <clean>   </clean>
+      B
+      <dirty>   A   B   </dirty>
+      C
+   </mixed>
+</doc>

data/test/c14n/result/with-comments/example-3

@@ -0,0 +1,14 @@
+<doc>
+   <e1></e1>
+   <e2></e2>
+   <e3 id="elem3" name="elem3"></e3>
+   <e4 id="elem4" name="elem4"></e4>
+   <e5 xmlns="http://www.uvic.ca" xmlns:a="http://www.w3.org" xmlns:b="http://www.ietf.org" attr="I'm" attr2="all" b:attr="sorted" a:attr="out"></e5>
+   <e6 xmlns:a="http://www.w3.org">
+       <e7 xmlns="http://www.ietf.org">
+           <e8 xmlns="">
+               <e9 xmlns:a="http://www.ietf.org" attr="default"></e9>
+           </e8>
+       </e7>
+   </e6>
+</doc>

data/test/c14n/result/with-comments/example-4

@@ -0,0 +1,9 @@
+<doc>
+   <text>First line&#xD;
+Second line</text>
+   <value>2</value>
+   <compute>value&gt;"0" &amp;&amp; value&lt;"10" ?"valid":"error"</compute>
+   <compute expr="value>&quot;0&quot; &amp;&amp; value&lt;&quot;10&quot; ?&quot;valid&quot;:&quot;error&quot;">valid</compute>
+   <norm attr=" '    &#xD;&#xA;&#x9;   ' "></norm>
+   <normId id="' &#xD;&#xA;&#x9; '"></normId>
+</doc>

data/test/c14n/result/with-comments/example-5

@@ -0,0 +1,4 @@
+<doc attrExtEnt="entExt">
+   Hello, world!
+</doc>
+<!-- Let world.txt contain "world" (excluding the quotes) -->

data/test/c14n/result/with-comments/example-6

@@ -0,0 +1 @@
+<doc>©</doc>

data/test/c14n/result/with-comments/example-7

@@ -0,0 +1 @@
+<e1 xmlns="http://www.ietf.org" xmlns:w3c="http://www.w3.org"><e3 xmlns="" id="E3" xml:space="preserve"></e3></e1>

data/test/c14n/result/without-comments/example-1

@@ -0,0 +1,4 @@
+<?xml-stylesheet href="doc.xsl"
+   type="text/xsl"   ?>
+<doc>Hello, world!</doc>
+<?pi-without-data?>

data/test/c14n/result/without-comments/example-2

@@ -0,0 +1,11 @@
+<doc>
+   <clean>   </clean>
+   <dirty>   A   B   </dirty>
+   <mixed>
+      A
+      <clean>   </clean>
+      B
+      <dirty>   A   B   </dirty>
+      C
+   </mixed>
+</doc>

data/test/c14n/result/without-comments/example-3

@@ -0,0 +1,14 @@
+<doc>
+   <e1></e1>
+   <e2></e2>
+   <e3 id="elem3" name="elem3"></e3>
+   <e4 id="elem4" name="elem4"></e4>
+   <e5 xmlns="http://www.uvic.ca" xmlns:a="http://www.w3.org" xmlns:b="http://www.ietf.org" attr="I'm" attr2="all" b:attr="sorted" a:attr="out"></e5>
+   <e6 xmlns:a="http://www.w3.org">
+       <e7 xmlns="http://www.ietf.org">
+           <e8 xmlns="">
+               <e9 xmlns:a="http://www.ietf.org" attr="default"></e9>
+           </e8>
+       </e7>
+   </e6>
+</doc>

data/test/c14n/result/without-comments/example-4

@@ -0,0 +1,9 @@
+<doc>
+   <text>First line&#xD;
+Second line</text>
+   <value>2</value>
+   <compute>value&gt;"0" &amp;&amp; value&lt;"10" ?"valid":"error"</compute>
+   <compute expr="value>&quot;0&quot; &amp;&amp; value&lt;&quot;10&quot; ?&quot;valid&quot;:&quot;error&quot;">valid</compute>
+   <norm attr=" '    &#xD;&#xA;&#x9;   ' "></norm>
+   <normId id="' &#xD;&#xA;&#x9; '"></normId>
+</doc>

data/test/c14n/result/without-comments/example-5

@@ -0,0 +1,3 @@
+<doc attrExtEnt="entExt">
+   Hello, world!
+</doc>

data/test/c14n/result/without-comments/example-6

@@ -0,0 +1 @@
+<doc>©</doc>

data/test/c14n/result/without-comments/example-7

@@ -0,0 +1 @@
+<e1 xmlns="http://www.ietf.org" xmlns:w3c="http://www.w3.org"><e3 xmlns="" id="E3" xml:space="preserve"></e3></e1>

data/test/tc_canonicalize.rb

@@ -0,0 +1,125 @@
+# encoding: UTF-8
+require './test_helper'
+require 'test/unit'
+
+class TestCanonicalize < Test::Unit::TestCase
+  # (www.w3.org) 3.1 PIs, Comments, and Outside of Document Element
+  # http://www.w3.org/TR/xml-c14n#Example-OutsideDoc
+  def test_canonicalize_with_w3c_c14n_3_1
+    given_doc = XML::Document.file('c14n/given/example-1.xml')
+
+    # With Comments
+    expected_with_comments = IO.read('c14n/result/with-comments/example-1')
+    assert_equal(expected_with_comments, given_doc.canonicalize(:comments => true))
+
+    # Without Comments
+    expected_without_comments = IO.read('c14n/result/without-comments/example-1')
+    assert_equal(expected_without_comments, given_doc.canonicalize)
+    assert_equal(expected_without_comments, given_doc.canonicalize(:comments => false))
+
+    # Without Comments (XML_C14N_1_1)
+    expected_1_1_without_comments = IO.read('c14n/result/1-1-without-comments/example-1')
+    mode = XML::Document::XML_C14N_1_1
+    assert_equal(expected_1_1_without_comments, given_doc.canonicalize(:mode => mode))
+  end#test_canonicalize_with_w3c_c14n_3_1
+
+
+  # (www.w3.org) 3.2 Whitespace in Document Content
+  # http://www.w3.org/TR/xml-c14n#Example-WhitespaceInContent
+  def test_canonicalize_with_w3c_c14n_3_2
+    given_doc = XML::Document.file('c14n/given/example-2.xml')
+    expected = IO.read('c14n/result/without-comments/example-2')
+    assert_equal(expected, given_doc.canonicalize)
+
+    expected_1_1_without_comments = IO.read('c14n/result/1-1-without-comments/example-2')
+    mode = XML::Document::XML_C14N_1_1
+    assert_equal(expected_1_1_without_comments, given_doc.canonicalize(:mode => mode))
+  end
+
+  # (www.w3.org) 3.3 Start and End Tags
+  # http://www.w3.org/TR/xml-c14n#Example-SETags
+  # (2012-02-20) Test failing due to missing 'attr' in 'e9' node.
+  #   - Embedded DTD not parsed out of XML, therefore default attributes are
+  #     not applied to canonicalization.
+  def test_canonicalize_with_w3c_c14n_3_3
+    given_doc = XML::Document.file('c14n/given/example-3.xml')
+    expected = IO.read('c14n/result/without-comments/example-3')
+    assert_equal(expected, given_doc.canonicalize)
+
+    expected_1_1_without_comments = IO.read('c14n/result/1-1-without-comments/example-3')
+    mode = XML::Document::XML_C14N_1_1
+    assert_equal(expected_1_1_without_comments, given_doc.canonicalize(:mode => mode))
+  end
+
+  # (www.w3.org) 3.4 Character Modifications and Character References
+  # http://www.w3.org/TR/xml-c14n#Example-Chars
+  def test_canonicalize_with_w3c_c14n_3_4
+    given_doc = XML::Document.file('c14n/given/example-4.xml')
+    expected = IO.read('c14n/result/without-comments/example-4')
+    assert_equal(expected, given_doc.canonicalize)
+
+    expected_1_1_without_comments = IO.read('c14n/result/1-1-without-comments/example-4')
+    mode = XML::Document::XML_C14N_1_1
+    assert_equal(expected_1_1_without_comments, given_doc.canonicalize(:mode => mode))
+  end
+
+  # (www.w3.org) 3.5 Entity References
+  # http://www.w3.org/TR/xml-c14n#Example-Entities
+  # (2012-02-20) Failing likely due to a logic error
+  #   - libxml2(c14n.c:1788) XML_ENTITY_REF_NODE is invalid node for parsing.
+  def test_canonicalize_with_w3c_c14n_3_5
+    given_doc = XML::Document.file('c14n/given/example-5.xml')
+
+    # With Comments
+    expected_with_comments = IO.read('c14n/result/with-comments/example-5')
+
+    # TODO - CANNOT COMPLETE TEST unless libxml2 supports additional node types.
+    #assert_equal(expected_with_comments, given_doc.canonicalize(:comments => true))
+
+    # Without Comments
+    expected_without_comments = IO.read('c14n/result/without-comments/example-5')
+    # TODO - CANNOT COMPLETE TEST unless libxml2 supports additional node types.
+    #assert_equal(expected_without_comments, given_doc.canonicalize(:comments => false))
+    expected_1_1_without_comments = IO.read('c14n/result/1-1-without-comments/example-5')
+    mode = XML::Document::XML_C14N_1_1
+
+    # TODO - CANNOT COMPLETE TEST unless libxml2 supports additional node types.
+    #assert_equal(expected_1_1_without_comments, given_doc.canonicalize(:mode => mode))
+  end
+
+  # (www.w3.org) 3.6 UTF-8 Encoding
+  # http://www.w3.org/TR/xml-c14n#Example-UTF8
+  def test_canonicalize_with_w3c_c14n_3_6
+    given_doc = XML::Document.file('c14n/given/example-6.xml')
+    if defined?(Encoding)
+      expected = IO.read('c14n/result/without-comments/example-6', :encoding => Encoding::UTF_8)
+    else
+      expected = IO.read('c14n/result/without-comments/example-6')
+    end
+    assert_equal(expected, given_doc.canonicalize)
+
+    if defined?(Encoding)
+      expected_1_1_without_comments = IO.read('c14n/result/1-1-without-comments/example-6', :encoding => Encoding::UTF_8)
+    else
+      expected_1_1_without_comments = IO.read('c14n/result/1-1-without-comments/example-6')
+    end
+    mode = XML::Document::XML_C14N_1_1
+    assert_equal(expected_1_1_without_comments, given_doc.canonicalize(:mode => mode))
+  end
+
+  # (www.w3.org) 3.7 Document Subsets
+  # http://www.w3.org/TR/xml-c14n#Example-DocSubsets
+  def test_canonicalize_with_w3c_c14n_3_7
+    # Non Canonicalized Document
+    given_doc = XML::Document.file('c14n/given/example-7.xml')
+    expected = IO.read('c14n/result/without-comments/example-7')
+
+    e1_node = given_doc.find_first('ietf:e1', 'ietf:http://www.ietf.org')
+
+    # Select current node, all child nodes, all attributes and namespace nodes
+    subdoc_nodes = e1_node.find("(.//.|.//@id|namespace::*)")
+
+    # TODO - This fails because the namespace nodes aren't taken into account
+    # assert_equal(expected, given_doc.canonicalize(:nodes => subdoc_nodes))
+  end
+end

data/test/tc_document.rb

@@ -31,22 +31,6 @@ class TestDocument < Test::Unit::TestCase
     }
   end
 
-  def test_canonicalize
-    doc = XML::Parser.string(<<-EOS).parse
-      <?xml-stylesheet href="doc.xsl" type="text/xsl"?>
-      <doc>Hello, world!<!-- Comment 1 --></doc>
-    EOS
-
-    expected = "<?xml-stylesheet href=\"doc.xsl\" type=\"text/xsl\"?>\n<doc>Hello, world!</doc>"
-    assert_equal(expected, doc.canonicalize)
-
-    expected = "<?xml-stylesheet href=\"doc.xsl\" type=\"text/xsl\"?>\n<doc>Hello, world!<!-- Comment 1 --></doc>"
-    assert_equal(expected, doc.canonicalize(true))
-
-    expected = "<?xml-stylesheet href=\"doc.xsl\" type=\"text/xsl\"?>\n<doc>Hello, world!</doc>"
-    assert_equal(expected, doc.canonicalize(false))
-  end
-
   def test_compression
     if XML.enabled_zlib?
       0.upto(9) do |i|
@@ -88,7 +72,6 @@ class TestDocument < Test::Unit::TestCase
                  @doc.to_s(:indent => false))
   end
 
-
   def test_doc_node_type
     assert_equal(XML::Node::DOCUMENT_NODE, XML::Document.new.node_type)
   end
@@ -131,6 +114,7 @@ class TestDocument < Test::Unit::TestCase
     error = assert_raise(XML::Error) do
       doc2.root << node
     end
+
     assert_equal(" Nodes belong to different documents.  You must first import the node by calling XML::Document.import.",
                  error.to_s)
 
@@ -139,4 +123,4 @@ class TestDocument < Test::Unit::TestCase
     assert_equal("<nums><two/><one/></nums>",
                  doc2.root.to_s(:indent => false))
   end
-end
+end

data/test/tc_encoding.rb

@@ -1,6 +1,5 @@
 # encoding: UTF-8
 
-if defined?(Encoding)
 require './test_helper'
 require 'test/unit'
 
@@ -26,16 +25,20 @@ class TestEncoding < Test::Unit::TestCase
     Encoding.default_internal = nil
   end
 
+  def file_for_encoding(encoding)
+    file_name = "model/bands.#{encoding.name.downcase}.xml"
+    File.join(File.dirname(__FILE__), file_name)
+  end
+
   def load_encoding(encoding)
     @encoding = encoding
-    file_name = "model/bands.#{@encoding.name.downcase}.xml"
+    file = file_for_encoding(encoding)
 
     # Strip spaces to make testing easier
     XML.default_keep_blanks = false
-    file = File.join(File.dirname(__FILE__), file_name)
     @doc = XML::Document.file(file)
   end
-  
+
   def test_encoding
     doc = XML::Document.new
     assert_equal(XML::Encoding::NONE, doc.encoding)
@@ -115,5 +118,4 @@ class TestEncoding < Test::Unit::TestCase
     assert_equal("109 246 116 108 101 121 95 99 114 252 101",
                  name.bytes.to_a.join(" "))
   end
-end
 end

data/test/tc_encoding_sax.rb

@@ -0,0 +1,115 @@
+# encoding: UTF-8
+require './test_helper'
+require 'test/unit'
+
+class SaxEncodingCallbacks
+  attr_reader :encoding
+  #include Test::Unit::Assertions
+  include MiniTest::Assertions
+
+  def initialize
+    @encoding = Encoding::UTF_8
+  end
+
+  # Called for a CDATA block event.
+  def on_cdata_block(cdata)
+    assert_equal(self.encoding, cdata.encoding)
+  end
+
+  # Called for a characters event.
+  def on_characters(chars)
+    assert_equal(self.encoding, chars.encoding)
+  end
+
+  # Called for a comment event.
+  def on_comment(msg)
+    assert_equal(self.encoding, msg.encoding)
+  end
+
+  # Called for a end document event.
+  def on_end_document
+  end
+
+  # Called for a end element event.
+  def on_end_element_ns(name, prefix, uri)
+    assert_equal(self.encoding, name.encoding)
+    assert_equal(self.encoding, prefix.encoding) if prefix
+    assert_equal(self.encoding, uri.encoding) if uri
+  end
+
+  # Called for parser errors.
+  def on_error(msg)
+    assert_equal(self.encoding, msg.encoding)
+  end
+
+  # Called for an external subset event.
+  def on_external_subset(name, external_id, system_id)
+    assert_equal(self.encoding, name.encoding)
+    assert_equal(self.encoding, external_id.encoding)
+    assert_equal(self.encoding, system_id.encoding)
+  end
+
+  # Called for an external subset notification event.
+  def on_has_external_subset
+  end
+
+  # Called for an internal subset notification event.
+  def on_has_internal_subset
+  end
+
+  # Called for an internal subset event.
+  def on_internal_subset(name, external_id, system_id)
+    assert_equal(self.encoding, name.encoding)
+    assert_equal(self.encoding, external_id.encoding)
+    assert_equal(self.encoding, system_id.encoding)
+  end
+
+  # Called for 'is standalone' event.
+  def on_is_standalone
+  end
+
+  # Called for an processing instruction event.
+  def on_processing_instruction(target, data)
+    assert_equal(self.encoding, target.encoding)
+    assert_equal(self.encoding, data.encoding)
+  end
+
+  # Called for a reference event.
+  def on_reference(name)
+    assert_equal(self.encoding, name.encoding)
+  end
+
+  # Called for a start document event.
+  def on_start_document
+  end
+
+  # Called for a start element event.
+  def on_start_element_ns(name, attributes, prefix, uri, namespaces)
+    assert_equal(self.encoding, name.encoding) if name
+    assert_equal(self.encoding, prefix.encoding) if prefix
+    assert_equal(self.encoding, uri.encoding) if uri
+  end
+end
+
+class TestEncodingSax < Test::Unit::TestCase
+  def setup
+    Encoding.default_internal = nil
+  end
+
+  def file_for_encoding(encoding)
+    file_name = "model/bands.#{encoding.name.downcase}.xml"
+    File.join(File.dirname(__FILE__), file_name)
+  end
+
+  def test_encoding_iso_8859_1
+    parser = XML::SaxParser.file(file_for_encoding(Encoding::ISO_8859_1))
+    parser.callbacks = SaxEncodingCallbacks.new
+    parser.parse
+  end
+
+  def test_encoding_utf8
+    parser = XML::SaxParser.file(file_for_encoding(Encoding::UTF_8))
+    parser.callbacks = SaxEncodingCallbacks.new
+    parser.parse
+  end
+end