RE: [transquery-discuss] Interoperability and xsl:param (was Re: Ananalogy)

From: Leigh Dodds (ldodds@ingenta.com)
Date: Thu Dec 06 2001 - 13:17:38 CET


> -----Original Message-----
> From: Francis Norton [mailto:francis@redrice.com]
> Sent: 06 December 2001 11:47
> To: Evan Lenz
> Cc: cutlass; transquery-discuss@lists.oasis-open.org
> Subject: Re: [transquery-discuss] Interoperability and xsl:param (was
> Re: Ananalogy)
>
[...]
> Interestingly Vadim Draluk has started a discussion
> [http://lists.w3.org/Archives/Public/www-ql/2001OctDec/0024.html] about
> the challenge of addressing web services in XML Query. Since the inputs
> to web services are typically xml docs such as SOAP messages as opposed
> to URLs.
>
> TranQuery nicely sidesteps the fact that neither XSLT nor apparently XML
> Query has a standard way of invoking those web services which require
> document-structured requests.

Draluk notes that he hopes 'no one would suggest putting them
[the document-structured requests] into the URL parameter of the
document() function'.

However I've been seeing that approach appear in a number of
contexts that describe themselves as interoperability
frameworks/specifications.
E.g. OpenURL and OAI Metadata Harvesting Protocol.

Responses may be still be XML documents, but it's simpler to
whip up simple standardised parameterised URL interfaces to do queries.

Frustrating but seemingly necessary.

However this is probably off-topic for TransQuery. To guide things
back to the main theme, I'm seeing interest in both the definition of
a top-level param (x:input) AND some (probably non-normative)
discussion of different processing models on how that parameter
gets populated.

--

As an aside, it occured to me that one could re-write any stylesheet that uses the document() function to use x:input, so long as the XSLT processor exposes those documents in the input collection.

I'm not sure whether there's any value in this, other than factoring out the location of the documents from the processing stylesheet?

But this would necessitate being able to define (externally) that a particular stylesheet required a particular collection of documents. Might be useful if one wanted to re-use the same stylesheet to process data from different sources. E.g. several different URL addressable data sources that may vary in their interfaces (if only by domain name), but respond with the same XML format.

You could achieve the same end with a custom URI scheme however.

Cheers,

L.



This archive was generated by hypermail 2.1.4 : Fri Feb 22 2002 - 11:35:58 CET