Re: [xml-dev] XPath and a continuous, uniform information space

[Uche Ogbuji <uche@ogbuji.net>]

On Sat, Aug 17, 2013 at 11:07 AM, Hans-Juergen Rennau <hrennau@yahoo.de> wrote:

Intriguing, Uche! You write:

"My own idea is to expand beyond that to support callback-type processing within the XML parse phase, triggered according to node patterns, which could support actions such as retrieving resources from the Web to produce additional or modified nodes.  I prefer that separation, and I can also imagine other uses for such a system, including additional and more efficient schema and transform mechanisms."

So you speak of "node patterns", "processing" and "actions". Do you think of associating nodes with something dynamical, something reminding of functions - the definition of a processing, to be triggered by the XML parser coming along?

There are many programming analogues that work for what I mean.  Sure functions (i.e. call-backs) would fit the bill.  I think structured programming co-routines are even closer to the mark.

I would hope that, if standardized, such an approach would find a way to not rely too heavily one one programming paradigm or another, which has been the curse of recent XML-related specs where the designers have clearly been thinking in terms of how a feature would be implemented in their pet language, thereby squandering the value of XML and data expression in general that it can be designed to outlast the mechanics of its processing.  

Seven years old but I still feel compelled to point to my longer argument at:

http://www.ibm.com/developerworks/xml/library/x-think38/index.html

(tl;dr = skip to "Separating information from process" section)

 

And I wonder if such associations could be regarded as resources themselves (or part of the resources), so that the result would resemble an object in the OO sense - data plus behaviour. Or perhaps I misunderstood? ...

I do think that a resource-centric approach should be taken to bind the triggering patterns to the processor's actions, but I would not call it OO-style encapsulation/polymorphism.  I was an OO fanatic back in the mid-90s but XML was one of the very technologies that taught me that even if you have a system of late-binding from pattern to actions, you do not have to, and in most cases should not think of the pattern and the bound action as a bundled unit of computing.  They should be very independent, and only conveniently joined for the moment of executing the triggered action.

Hope that makes sense.  In other words, I think you're getting the concepts rightly, but I think it can be dangerous to try to express them by analogy to mainstream programming concepts.

--
Uche Ogbuji                       http://uche.ogbuji.net
Founding Partner, Zepheira        http://zepheira.com
http://wearekin.org
http://www.thenervousbreakdown.com/author/uogbuji/
http://copia.ogbuji.net
http://www.linkedin.com/in/ucheogbuji
http://twitter.com/uogbuji