[
Lists Home |
Date Index |
Thread Index
]
> Paul Prescod <paul@prescod.net> wrote:
> * there is XHTML
> * there is XSL-FO
> * there is a new Open Office project which will surely define a new
> word processor format
> * SVG now has textflow features
>
> The XHTML working group _could_ take it upon themselves to attack this
> particular problem.
>
From the set of existing schemas for "word-processor" like documents
it is interesting to split schemas that are "fully structured" like XML DocBook
from schemas that are only "semi-structured" like XHTML or the OpenOffice format.
Here the term "semi-structured" means that the document sections may not
be encapsulated properly. For example in XHTML it is allowed (even if not recommanded) to
write:
...
<h1>A section title</h1>
<p> This is the section content...</p>
<h2> A subsection </h2>
<p> The content of the subsection </p2>
<h1>Another section title</h1>
<p> This is the other section content </p>
...
Therefore it is possible to produce a linear structure of the document instead of
the intended hierachical structure. Proper use of XHTML would add <div> elements
to isolate section and encapsulate subsections. However this is not
controled by the schema, only by authoring guidelines.
Another potential classification is to separate layout-oriented schemas
like XSL-FO and SVG from semantic-schema like Docbook (XHTML being
inbetween).
Since it is difficult to retrieve the hierarchical structure back from its linear
form, and nearly impossible to retrieve semantic information that has been
translated in typographical attributes, I think we need to be very careful
when choosing an authoring schema. Eventually there are not many options
if you want to save the structure and the logic of your content in your XML
document.
Thierry
---------------------------------------------------
- IXELIS - Systèmes d'Information Sémantique
17 rue des Cèdres, F67200 Strasbourg (France).
Tél : +33 (0)3 88 27 81 39
Mobile : 06 30 07 65 73
E-mail : mail@ixelis.net
Fax : 03 88 27 81 39
---------------------------------------------------
|
