In message <v01540b0aafc81699ccd6@[137.111.90.66]> ross@mpce.mq.edu.au (Ross Moore) writes:
[...]
>
> Currently I'm putting the finishing touches on the latest version of LaTeX2HTML.
This is a noble effort.
>
> Later this year I hope to tackle LaTeXML for which I would like to be
> able to use
> existing DTDs as much as possible --- especially for portions of MathML ---
> rather than having to write my own.
I have always admired the (La)TeX virtual community of volunteers and presumably
they will be keen to learn about XML and how it applies to LaTeX. In which case
this represents a significant pool of potential XML-friendly hackers :-)
I'm thinking as I write, but it seems as if there should be 'a' LaTeX DTD
(possibly modular), which interoperates with the MathML DTD. I think it's
important to keep them distinct because there are many people who don't use
LaTeX for maths, but as a general authoring tool. Since MathML specifically
mentions TeX as a NOTATION, and as isomorphic to mathML in some parts, the
clear separation of all components (LaTeXML, MathML, TeX) is critical.
>
> Having a reliable HTML --> XML ought to be an option too.
>
> Indeed this would probably be the easiest way to go for a first working version,
> given the effort that has already gone into LaTeX2HTML .
I'd agree. LaTeX is an excellent tool, but it doesn't have the full structuring
power of XML unless it's specifically thought of at the start. I speak from
experience as I wrote a complex book in LaTeX, with outputs as *.dvi, *.html,
and several implied conditional sections. That was before I discovered the
point of SGML - I spent many midnights writing programs to restructure the
book :-(
>
> Ultimately a scheme will be needed whereby (partial) DTDs can be
> constructed automatically
> from any \newenvironment commands that the user devises for the LaTeX
> typeset version.
Yes - I think that a current LaTeX user can probably devise structuring
like this that makes the transformation much easier. Among the things that are
difficult to convert are paragraph/line breaks (when not explicitly marked up)
>
>
> I'd love to hear from anyone else interested in:
>
> 1. converting existing LaTeX documents into XML ;
I'd agree that LaTeX->HTML/XML is a useful start. One discussion would
be whether one had to have a DTD that supported all constructs in the LaTeX
manual, or whether there was a more generic DIV-like container. Another would
be how to support user-defined macros. Also, would you work on the authored
document, or some later normalised/expanded version (I've lost touch with
Latex2html, but I assume that it works on some normalised version which has
lost the author's macros).
For scientific technical documents this is a highly desirable goal :-)
>
> 2. using LaTeX syntax as a front-end to XML for documents on the Web .
Do you mean transforming XML documents into LaTeX (I tend to think of this
as a back-end) or as a way of authoring XML documents using LaTeX? The latter
is rather similar to (1). The second will require a transformation engine
which most people would approach through DSSSL styleheets, I imagine.
P.
-- Peter Murray-Rust, domestic net connection Virtual School of Molecular Sciences http://www.vsms.nottingham.ac.uk/