[
Lists Home |
Date Index |
Thread Index
]
Michael Kay wrote:
>
> James, I've completely lost your thread. You're using terminology I don't
> understand, and I don't know where it comes from. I've no idea what you mean
> by metacircularity, for example.
my first encounter with the concept "metacircularity" was in the context
of symbolic processing languages. in particular, scheme. as expressed in
that domain, "[a]n evaluator that is written in the same language that
it evaluates is said to be 'metacircular'."[1] i thought the word would
have made it into the OED, but as of 1989 it hadn't. oed-online costs
money, so i don't know if it's in there by now. try the link from [1]
instead. for another example, try [2].
xslt is concerned with metacircularity in that it involves
transformations for xml expressions which are expressed as xml
expressions. (i considered rephrasing that, but the point is cleared as written.)
as to the notion of extent which appears in the last paragraph of my
note, see [3] and/or [4]. in-scope namespaces are an attempt to afford
prefix/namespace-name bindings indefinite extent. given the operations
which one would like to perform on xml documents, this cannot attempt
cannot succeed unless the scope of the bindings is refined to correpond
to the scope of the possible operands. which, in the terms of the
infoset, would be character information items.
>
> The behavior of XSLT 1.0 stylesheets with respect to namespaces is
> well-defined, the results are predictable, the implementations are
> conformant with the specification, and on the whole the behavior meets user
> expectations. Now what exactly is your problem?
>
> > ...
> >
> > for example, when considering how to implement xslt, one can
> > conceive of
> > an abstract syntax for xslt expressions which permitted an
> > expression like
> >
the standard specifies this:
> > ({xsl}element ({}name "xsd:QName"
> > {}use-attribute-sets "nsp1:attset1 nsp2:attset2")
> > "nsp3:name")
> >
> > an implementation which were expressed in terms of this
> > abstract syntax
> > would be conformant, but flawed. a better implementation would specify
> > an abstract syntax which required an expression of the form
given this expression, and given that this expression may itself be the
result of an arbitrary computation, i cannot predict that a transform
applied to the expression of a transform will produce a usable result. i
would like to be able to expect it to. "in-scope" namespaces are an
attempt to be able to predict the result. they do not work in the
general case. i can expect to be able to predict the result to be usable
only if the abstract syntax is restricted to one of the following forms.
> >
> > ({xsl}element ({}name "QName"
> > {}namespace "xsd"
> > {}use-attribute-sets
> > ({namespace-name-1}attset1
> > {namespace-name-2}attset2))
> > {namespace-name-3}name )
> >
> > or
> >
> > ({xsl}element ({}name {xsd}QName
> > {}use-attribute-sets
> > ({namespace-name-1}attset1
> > {namespace-name-2}attset2))
> > {namespace-name-3}name )
> >
neither of these is conformant.
that is not a good state of affairs. to make matters worse, XSLT is but
one manifestation of the problem.
--------------------------------------------
[1] Abelson,Sussman,Sussman, Structure and Interpretation of Computer
Programs, MITP 1966, p362ff (or
http://mitpress.mit.edu/sicp/full-text/sicp/book/node76.html)
[2] Baker, "Metacircular Semantics for Common Lisp Special Forms" (http://linux.rice.edu/~rahul/hbaker/MetaCircular.html)
[3] Steele, Common Lisp the Language, 2ndEd, Chapter 3, "Scope and
Extent" (http://www.tnt.uni-hannover.de/js/soft/case/lang/lisp/cltl/clm/node43.html)
[4]
http://www.xanalys.com/software_tools/reference/HyperSpec/Body/26_glo_e.htm#extent,
in particular v/s
http://www.xanalys.com/software_tools/reference/HyperSpec/Body/26_glo_s.htm#scope,
and in the context of http://www.xanalys.com/software_tools/reference/HyperSpec/Body/d_specia.htm
|
