- a setting for the whole export - first parse the text frames and retain where the page breaks are (which caracter index for each frame) - get the list of the images that are to be distributed (top or bottom of the page)