- From: Norman Walsh <ndw@nwalsh.com>
- Date: Tue, 01 Dec 2009 16:38:05 -0500
- To: public-xml-processing-model-comments@w3.org
- Message-ID: <m24ooa4bpe.fsf@nwalsh.com>
"Henry S. Thompson" <ht@inf.ed.ac.uk> writes:
> While we're getting our wishlist out (see Norm's previous posting) for
> obscure facilities which we need for pipelines we're writing, I've
> finally been converted to the restricting-values-to-strings-is-crazy
> position.
Welcome. Don't worry, we're friendly.
> But here's why we're _not_ going to do this in V1: it would take a
> substantial amount of work to decide how to get things _back_ from
> such values.
Really? I'd have thought string($expr) was the obvious answer.
> Would we allow, e.g.,
>
> <p:declare-step type="my:demo">
> <p:output name="result" sequence="true" primary="true"/>
> <p:option name="attrSet"/>
>
> <p:identity select="$attrSet"/>
> <p:for-each>
> <p:load>
> <p:with-option name="href" select="."/>
> </p:load>
> </p:for-each>
> </p:pipeline>
>
> To be useful, such a pipeline would need to be allowed. But that's
> too big a change for this late in the game. . .
I was so convinced that it was both necessary and easy to do that I
implemented sequences of nodes as values in XML Calabash. If you use
the "-X general-values" option, XML Calabash is a non-conforming
processor that (probably) does what you want.
Be seeing you,
norm
--
Norman Walsh <ndw@nwalsh.com> | In science, "fact" can only mean
http://nwalsh.com/ | "confirmed to such a degree that it
| would be perverse to withhold
| provisional assent." I suppose that
| apples might start to rise tomorrow,
| but the possibility does not merit
| equal time in physics
| classrooms.--Stephen J. Gould
Received on Tuesday, 1 December 2009 21:38:43 UTC