We are very grateful to Eve for having produced the markup specification.
Unfortunately she is a victim of her success in that rec.xml [my shorthand
for the spec] is the first 'really crunchy official piece of XML' that we
can get to grips with for learning and developing our tools. This is why a
DTD and its associated semantics/documentation is so important :-). [I
would also expect that 'spec.dtd' might be re-usable in other contexts.]
>
[...]
>
>We had a lengthy discussion of whether our production markup should be more
>semantic and less presentational. It's so much work to make the markup
>simulate the EBNF and to make the filters handle this, that we decided not
>to go further in that direction. I do agree that the production markup is
>less than "pure" in this area.
>
My interest is similar - but complementary - to Michael's; I am interested
in the terminology. Thus I want to be able to abstract the terms [there are
62 termdefs] in the document and produce a model for their structure (e.g.
entailment by containment, by linking and so on.) In this way I can create
a graphical interactive map of the concepts in the XML spec and have
already created a prototype. I would like to know, for example, whether all
terms are defined by <termdef> or whether there are some which are simply
defined by <term>foo bar</term>. There appears to be some duplication here
as well; thus a termdef has an attribute naming the term, but it is also
often contained within a <term> later in the 'description'. [And there is
at least one case where </termdef> occurs in mid-sentence - I suspect this
isn't intended.]
P.
Peter Murray-Rust, Director Virtual School of Molecular Sciences, domestic
net connection
VSMS http://www.nottingham.ac.uk/vsms, Virtual Hyperglossary
http://www.venus.co.uk/vhg