2012/6/7 John Maddock <boost.regex@virgin.net>

This is slightly off topic.

...

I have a question how pdf's are build within the boost build system. As far as I could figure it out the build system uses a proprietary solution called RenderX XEP or so. I'm wondering if there are open source solutions out there that could be used to generate boost's documentation as pdf files?

Apache FOP is the only open source solution I know of: http://xmlgraphics.apache.org/**fop/ <http://xmlgraphics.apache.org/fop/>

Unfortunately it still hasn't got past a 1.0 release, and last I checked it's rendering was still far inferior to XEP - basically building with XEP is pretty much a "fire and forget" process, while building with FOP quite often either mangles the docs or uses rather unfortunate pagination choices :(

John.

I don't know if is in the same category of software you're discussing, but i found a library sometime ago for generating PDFs from C++: http://libpdfxx.sourceforge.net/doc/index.html Seems it's not maintained for a while, though. Regards, Júlio.

on Thu Jun 07 2012, John Maddock <boost.regex-AT-virgin.net> wrote:

...

I don't know if is in the same category of software you're discussing, but i found a library sometime ago for generating PDFs from C++: http://libpdfxx.sourceforge.net/doc/index.html

Nope, doesn't even come close - converting from FO XML and getting all the pagination correct in the output is a hugely complex job - probably why there's only ever been one solution that's even come close to getting it right - and even they (XEP that is) don't handle everything.

Did we every try using http://dblatex.sourceforge.net/ and then sending it through pdflatex? -- Dave Abrahams BoostPro Computing http://www.boostpro.com

Thank you all for your insight. I was just curious. Regards, Christian

...

Nope, doesn't even come close - converting from FO XML and getting all the pagination correct in the output is a hugely complex job - probably why there's only ever been one solution that's even come close to getting it right - and even they (XEP that is) don't handle everything.

And BTW - who cares XEP is open source or not. We're only interested in the result (and the cost) not the software.

Exactly.

what about using pandoc to convert docbook to pdf ? Has anyone looked into this?

Nope, web page looked interesting so I gave it a try: Boost.Math: pdflatex couldn't handle some of the Unicode characters generated and bailed out. Boost.Type_traits: generated invalid tex file (pdflatex bailed out). Boost.Config: generated a PDF file, but wasn't much better than a plain text dump - you got bold and italics converted, but other than that all document structure was lost. John.

[Please do not mail me a copy of your followup] boost-users@lists.boost.org spake the secret code <m2txyn6jzk.fsf@pluto.luannocracy.com> thusly:

on Thu Jun 07 2012, Christian Henning <chhenning-AT-gmail.com> wrote:

...

solution called RenderX XEP or so. I'm wondering if there are open source solutions out there that could be used to generate boost's documentation as pdf files?

The best way to do this continues to be to go through LaTeX.

Is there a recipe for how to do this? -- "The Direct3D Graphics Pipeline" free book <http://tinyurl.com/d3d-pipeline> The Computer Graphics Museum <http://computergraphicsmuseum.org> The Terminals Wiki <http://terminals.classiccmp.org> Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>

PDFlib has a free version.

http://www.pdflib.com/

Then, there's GNU PDF:

http://www.gnupdf.org/Main_Page

I highly recommend PDFNet from PDFTron; it is commercial but their free evaluation version lasts a long time (specially on Linux).

http://www.pdftron.com/pdfnet/

Stop already! Sorry, but those are libraries and we need a complete solution for docbook to PDF transformation, as I said in another message, the hard part isn't writing the PDF, it's paginating the document, that's why no one has written a home-grown solution, now would it be sensible to try when there are already complete working solutions IMO. John.