It seems when using an XML parser that it does the conversion automatically. So, it seems it is not possible to get the original content from the node as the XML parser applies some processing to it. If someone knows about a way to get it, I would be happy to learn more about it.