From: Philippe Verdy (firstname.lastname@example.org)
Date: Wed Mar 09 2005 - 17:12:38 CST
From: "Richard T. Gillam" <email@example.com>
> >Anyone who does digitizing of epigraphic sources is familiar with the
> various text markup schemes; but no markup
>>addresses plain text integrity.
> You haven't demonstrated why plain text integrity is a requirement.
> You've demonstrated why XML is inconvenient. Not only is inconvenient
> not the same thing as unusable, but XML isn't necessarily your only
> choice here (although it might be that the established standards for
> this kind of thing all use XML). There are lots of methods of carrying
> around out-of-band styling information and metadata.
Yes but XML is not the only choice. You can create non strictly hierarchical
markup with SGML instead of XML.
SGML parsing is much more complicate than XML, but SGML has been used since
much longer in the edition and news industries, and it allows creating
several parallel hierarchies on the same document.
With XML, the only way to do that is to split the document into as many leaf
nodes as needed, give them each a distinct id, and then create separate
hierarchies with references to the actual node ids. In fact, this method is
also used but with more refinement when handling a set of CSS rules over a
XHTML document, except that elements are not required to be given all a
distinct id, and matching elements by type or class is possible.
You could use the same method in XML text (out of XHTML context) by
assigning classnames to text nodes, and then create a separate markup trees
referencing all nodes with the same class(es)... Classes are working in the
reverse direction of ids: text elements will reference one or more nodes in
the markup tree, but with ids this is the markup tree that references the
This archive was generated by hypermail 2.1.5 : Wed Mar 09 2005 - 17:13:26 CST