The bug is fixed now. The span handling code only detected span changes but did not react properly on page breaks inside spanned sections.
On 07.05.2006 06:17:15 Adam wrote: > Dear FOP-dev group. > > Firstly would just like to say that the current work being done on FOP > is fantastic. I've been able to use it in all manor of places from help > guides to dynamic content creation using Cocoon. :) > > Re: http://issues.apache.org/bugzilla/show_bug.cgi?id=39443 > I've done some more testing and the results seem to lean towards ( I > suspect) the page layout engine. Different renderers like print, TIFF > and SVG are showing the same problem as stated in the report. > > Could the problem be that when a fo:block is set to span='all' and the > block flows over to the next page its span="all" attribute setting is > lost? Its holds the width correctly (span all width) but the columns are > reset back to the default page layout. It happens when 2,3,4,etc columns > are set. > > A workaround, I made each original row into a single table, so that the > FO was just a table after table in the same block. It still had issues > like above. I then put each table in to its own span all block > interleaved with a <fo:block>​</fo:block> , a zero width space. > This produced the results I needed. Haven't tested what would happen if > a single row spanned 2 or more pages. > > Does this make sense? Is this a bug like I thought? > > Regards > Adam > > Jeremias Maerki