I'm not sure what you're trying to do here but XSL is not used by FOP. For convenience, FOP can accept FO input or XML and XSL input but it only uses the FO. I'm using embedded code to transform XML with an XSL, and for debugging I wanted to see the FO, so I split out the FOP and call the transform twice. I'm not sure what you mean by calculations on page number. If you're letting FOP overrun blocks to show one thing on multiple pages you can get the page count after the transform, and you can call the transform again if you retain your input. If you can do what I do and just know what fits on a page, you can call the transform once for each page. I started doing that to save memory and I can generate really large reports. I just have to pass a value in with each input to use the xsl:initial-page-number so the first page of the second document can come out identified as page 2. I retain the PDFs in memory, no physical files have to be written, and I create PDF objects with them using the pdfbox project which is able to create a new PDF and copy in the pages. I tested this on a report with over 1800 pages. Trying to get FOP to put that all in one document crashed out of memory long before that.
________________________________ From: Theresa Jayne Forster [mailto:[email protected]] Sent: Friday, June 03, 2011 9:04 AM To: [email protected] Subject: RE: Problems since upgrading, On that note, is it possible to chain XSL files on the fop input, or even make it double process, Ie multi pass processing to allow for calculations on page number etc? Kindest regards Theresa Forster Senior Software Developer
