Fix Unicode bug

Fix Diego Restrepo's bug (see Rev 184).
Update to latest HTML5lib.

vendor/plugins/HTML5lib/test/test_input_stream.rb

@@ -0,0 +1,26 @@
+require File.join(File.dirname(__FILE__), 'preamble')
+require "test/unit"
+require "html5/inputstream"
+
+class TestHtml5Inputstream < Test::Unit::TestCase
+  def test_newline_in_queue
+    stream = HTML5::HTMLInputStream.new("\nfoo")
+    stream.unget(stream.char)
+    assert_equal [1, 0], stream.position
+  end
+  
+  def test_buffer_boundary
+    stream = HTML5::HTMLInputStream.new("abcdefghijklmnopqrstuvwxyz" * 50, :encoding => 'windows-1252')
+    1022.times{stream.char}
+    assert_equal "i", stream.char
+  end
+  
+  def test_chars_until
+    stream = HTML5::HTMLInputStream.new("aaaaaaab")
+    assert_equal "aaaaaaa", stream.chars_until("b")
+
+    stream = HTML5::HTMLInputStream.new("aaaaaaab")
+    assert_equal "aaaaaaab", stream.chars_until("c")
+    
+  end
+end