Support for Unicode XML/HTML identifiers

Technically HTML only allows for ASCII names but restricting that
actually requires more work than just allowing it.

ext/ragel/base_lexer.rl

@@ -48,7 +48,9 @@
 
     newline    = '\r\n' | '\n' | '\r';
     whitespace = [ \t];
-    ident_char = [a-zA-Z0-9\-_\.];
+
+    unicode    = any - ascii;
+    ident_char = unicode | [a-zA-Z0-9\-_\.];
     identifier = ident_char+;
 
     whitespace_or_newline = whitespace | newline;

spec/oga/xml/lexer/elements_spec.rb

@@ -315,4 +315,31 @@ describe Oga::XML::Lexer do
     ]
   end
 
+  it 'lexes an element with a name containing Unicode characters' do
+    lex('<foobár />').should == [
+      [:T_ELEM_NAME, 'foobár', 1],
+      [:T_ELEM_END, nil, 1]
+    ]
+  end
+
+  it 'lexes an element with a name containing an underscore' do
+    lex('<foo_bar />').should == [
+      [:T_ELEM_NAME, 'foo_bar', 1],
+      [:T_ELEM_END, nil, 1]
+    ]
+  end
+
+  it 'lexes an element with a name containing a dash' do
+    lex('<foo-bar />').should == [
+      [:T_ELEM_NAME, 'foo-bar', 1],
+      [:T_ELEM_END, nil, 1]
+    ]
+  end
+
+  it 'lexes an element with a name containing numbers' do
+    lex('<foo123 />').should == [
+      [:T_ELEM_NAME, 'foo123', 1],
+      [:T_ELEM_END, nil, 1]
+    ]
+  end
 end