Gareth and Ed responded to my post:
Gareth:
If the document has sequential page numbering and/or cross-references between the volumes then it would be very difficult to print each volume of the document in turn.
The best bet is to produce a monolithic PDF and then split it into chunks. If the bookmarks are getting lost then that means you need to use a better PDF split tool that retains the bookmarks. We have written code for prior projects which achieve this via libraries such as iText or PDFBox. I don't believe any of the Arbortext tools have a function for PDF splitting.
BTW the other problem you might have is intra-document hyperlinks. If volume 1 contains hyperlinks to other volumes then after the document is split those hyperlinks would no longer work. It would be possible to programmatically fix those type of hyperlinks but I don't believe any PDF tools will do that out the box. You would have to write custom code.
Ed:
If you have some way of flagging where the Volumes begin and end, and have a separate output DTD with the Volume structures, XSL should work for transforming your source into multiple output volumes. Sounds like a lot of work, but transforming to a more versatile output doctype/DTD/Schema may be helpful for other reasons.
