This is the mail archive of the docbook-apps@lists.oasis-open.org mailing list .


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Direct DocBook XML to LaTeX?


>From: "Prikryl,Petr" <PRIKRYLP@skil.cz>
>To: docbook-apps@lists.oasis-open.org
>Subject: Direct DocBook XML to LaTeX? (was ... Re: Conceptual model
          for processing DocBook SGML and XML)
>Date: Wed, 09 Jan 2002 12:10:12 +0100
>
>I would also appreciate direct conversion to sources for LaTeX, >because 
>the LaTeX is widely used and matured.  It is possible to generate 
>PostScript, PDF, HTML, text...  Many users who already
>use LaTeX do not know DocBook, but it is likely they will use it
>soon.  Adding only one tool -- a XSLT processor -- would give them
>complete chain of tools if the LaTeX were produced.

Many DocBook processing tools can already generate all of those output 
formats, and since some information is lost between DocBook -> LaTeX, the 
HTML output produced directly from DocBook is better.

Currently, I use OpenJade with JadeTeX, to produce PDF and PostScript 
output.  I don't like it, though, since you still run into many of TeX's 
limitations (like having to iterate, to get your ToC and XRefs to converge).

I don't know a whole lot about XSL-FO (yet), but I think it preserves nearly 
all of the meaningful information (for further processing) that DocBook -> 
LaTeX could capture, as well as some things for which LaTeX has no concept.

As far as I can see, there are really only 4 reasons for sticking w/ LaTeX, 
as a back end:
1. writing XSLT to transform DocBook to LaTeX, since you don't know
   XSL-FO (IMO, this isn't worthwhile, but have a look at db2latex,
   if this interests you)
2. Math support - I've heard of people using TeX to render math
   embedded in DocBook.  I don't know what tools there are, for
   doing this (PasiveTeX, maybe?).  Ultimately, such hacks, while
   still perhaps necessary, should be replaced by embedding MathML,
   once suitable infrastructure exists to render it.
3. integration w/ specialized LaTeX backend tools
4. documentation (perhaps you could generate document chunks from
   DocBook, that you could use \input{} to include from a LaTeX
   document.  IMO, the only value of this would be to allow a more
   graceful transition from LaTeX-based docs to DocBook.  However,
   I think it'd be more trouble than its worth.

As you can see, they're all fairly short-term and transitional, in nature.  
Personally, I think TeX shows its age (and the fact that Knuth wasn't really 
interested in completely solving the problem of page layout), and I look 
forward to getting completely away from it, and never looking back.


Matt Gruenke


_________________________________________________________________
Chat with friends online, try MSN Messenger: http://messenger.msn.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]