Canonicalization failed with some latin2 characters 'čćžšđČĆŽŠĐ'(leters with
caron, ... ).

Release 1.3.0 don't have such problem.

Code which demonstrates bug:

// parse document
DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
dbf.setNamespaceAware(true);
DocumentBuilder db = dbf.newDocumentBuilder();
// text contains some latin2 characters 'čćžšđČĆŽŠĐ'
String text = new
String("<text>\u010D\u0107\u017E\u0161\u0111\u010C\u0106\u017D\u0160\u0110</text>");
Document doc = db.parse(new ByteArrayInputStream(text.getBytes("UTF-8")));
Element e_latin2 = doc.getDocumentElement();
Canonicalizer20010315WithComments c14 = new Canonicalizer20010315WithComments();
byte[] canon_bin = c14.engineCanonicalizeSubTree(e_latin2);

if (Arrays.equals(text.getBytes("UTF-8"), canon_bin))
   System.out.println("OK");
else
   System.out.println("Failed");