Comments on [Editorial Draft] Abstract Component References

Suppose that some abstract resource (e.g., a schema or query) has multiple 
representations.  They have the same semantics, but have very different syntax 
and structures.  In such cases, how can fragment identifiers identify abstract 
components?

1) XQuery

Other than the current syntax, an XML-syntax has been planned.

Moreover, normalization to XQuery core causes a significant change.  For example, 
some XPath expressions are converted into "For" statements.

2) RELAX NG

It has two representations: an XML-syntax and compact syntax.  Some constructs 
of the XML-syntax has to be tweaked in the compact syntax.  For example, 
namespace declarations have to be moved to the top-level.

Just like XQuery, RELAX NG uses normalization.  A normalized schema is very 
different from the original.

Cheers,

Makoto

-- 
MURATA Makoto <murata@hokkaido.email.ne.jp>

Received on Saturday, 1 November 2003 20:48:33 UTC