What is its data structure? Is it XML-based? How can it distinguish between different content types, for example text, image, files etc.?