As increasing amount of data is published in the form of XML, copyright protection of XML data is becoming an important requirement for many applications. While digital watermarking is a widely used measure to protect digital data from copyright offences, the complex and flexible construction of XML data poses a number of challenges to digital watermarking, such as re-organization and alteration attacks. To overcome these challenges, the watermarking scheme has to be based on the usability of data and the underlying semantics like key attributes and functional dependencies. In this paper, we describe WmXML, a system for watermarking XML documents. It generates queries from essential semantics to identify the available watermarking bandwidth in XML documents, and integrates query rewriting technique to overcome the threats from data re-organization and alteration. In the demonstration, we will showcase the use of WmXML and its effectiveness in countering various attacks.