I like how much effort has gone into this package. I would like it better if the class were split into separate functional parts e.g. the part that unpacks the .docx into a set of DOMDocument objects could be a separate class, because I would use that for other purposes (other than converting the docx-xml to html). But bravo for this solution!!