* Hash code for PI, CDATA and comments
* Differences Text in PI, CDATA and Comments
* Best difference in text, PI, CDATA <diff:added> <diff:removed> with longest common sequence
* Support for XPath expressions in identifiers
* Debug the segfault when building as a dynamic library
* Binders : Java (JAXP, libxmlj, JNI ?), Python (Swig ?)
