Well, I am not entirely sure, but if you mess a lot with styling in Word, this can happen. If you select a word and change it and then change it back (not via undo), I believe this will put a span around the selection it as it is styling applied to that specific selection. It is not checked if the styling is exactly the same as the parent. So, if they done this a lot (perhaps via S&R), disaster will strike... If it is a Word document originally and you get your hands on it, I would be interested to know what comes out of it after conversion to HTML and ePUB via my tools. In theory it should remove all that crud.
Depending on the styling in the ePUB, you could try importing the ePUB and re-export to ePUB via the tools. As stylenames would be retained, replacing the stylesheet afterwards should give you a head start.
|