Finding containing XML node based on character offset

Question

Based on a character offset into an XML document, how can I find the containing node of that character position?

For example consider the XML: