Re: Three Access Language Paradigms

Peter Murray-Rust (peter@ursus.demon.co.uk)
Wed, 19 Nov 1997 09:20:48

At 23:54 18/11/97 -0800, Tim Bray wrote:
[...]
>
>Let me rephrase Henry's comment: I suggest that those who are proposing
>brave new query language worlds go have a look at XLL. It *may* be the
>case that XLL xpointers hit a good 80-20 point in terms of what we'd
>like in a query language and in ease of implementation. -Tim
>
I support this. I have found TEI Xpointers in XLL *extremely* useful - they
have revolutionised my thinking about XML documents. Essentially, in many
cases, the 'document is the database' (for smallish applications). I also
use them inside JUMBO for navigating within known structures (e.g. seeing
whether an element has certain relatives and, if so, taking some action.)

For certain purposes the TEI Xpointer is limited. Initial discussions
suggested:

- SPACE (for coordinate systems such as images, tables)
- some sort of regexp
- FOREIGN for adding your own methods on.
- and I'd be happy to see something for numeric and other types values

I was in favour of these (I have to use them somehow in JUMBO), but it was
made clear that Xpointers were intended as an addressing scheme and not a
query language. I respect this distinction, but it would be very nice to be
able to extend TEI syntax to allow this.

As I understand it, TEI syntax (a la XLL) is confined the use in HREFs
within elements with XML-LINK attributes. Any of the proposed extensions is
(rightly) illegal there. But it would be possible to extend TEI for use
*elsewhere* (e.g. in querying documents) and I would be very happy to see
keywords of the sort above added *for query purposes*.

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