Re: A (long) question on implementing fo:retrieve-table-marker
On Sep 2, 2008, at 11:21, Jean-François El Fouly wrote: Just another hint I thought of: To begin with, I've built a simple example / test case related to my customer's needs but fairly based on the one in 6.13.1.1.2: http://www.w3.org/TR/xsl/#d0e14681 This example looks good and though not trivial to understand at first sight, it is a common-sense use of fo:retrieve-table-marker. It has a sort of conditional row that is displayed (or not) in the table footer and is defined just after table-body: *fo:marker marker-class-name=continued fo:table-row fo:table-cell fo:blockTable continued.../fo:block /fo:table-cell /fo:table-row /fo:marker * Due to the problems you ran into so far, I'd advise first trying to get the retrieval of a simple fo:block into the header/footer working. That would already help most people out who need this, I think. Once that works, extending the approach to also work with entire table-parts should prove a relatively simple exercise. Cheers Andreas
Re: A (long) question on implementing fo:retrieve-table-marker
On Sep 2, 2008, at 11:21, Jean-François El Fouly wrote: ... and another detail, to facilitate studying the process through debugging: snip / file:/D:/Java/fop-0.95beta/test-rtm.fo:533:112: Error(533/112): fo:retrieve-tab ^^^ For your own convenience, trim that testcase down! ;-) Seriously, try to generate extra space by means of space-before and/ or -after to generate multiple pages. Use an extra small page-size to limit the amount of needed content to trigger page-breaks... Generating dummy content, leading to a FO of 500+ lines, will not make the job any easier. Something I've also found helpful, in case you're unaware of it: specify ids on the FOs to easily track the corresponding FObj and their related LayoutManager in a debugger. Andreas
