[
Lists Home |
Date Index |
Thread Index
]
----- Original Message -----
From: "Ronald Bourret" <rpbourret@rpbourret.com>
To: <xml-dev@lists.xml.org>
Sent: Thursday, January 17, 2002 11:50 PM
Subject: Re: SV: SV: [xml-dev] XML=WAP? And DOA?
> Nicolas LEHUEN wrote:
>
> > How do you process elements that you don't know ?
>
> Several ways:
>
> 1) Work on the structure, not the tag name. Examples are DOM, SAX, and
> tools such as XMLWriter.
>
> 2) Interpret tag names based on information supplied programmatically at
> run time. Examples are XSLT and XQuery. DTD-based editors also fit this
> category.
100% OK for 1) and 2). That's what I call 'lexical-level' processing. No
knowledge of the semantics of the data is required. An XSLT engine doesn't
"know" the meaning of the document it manipulates, only the stylesheet
author does.
> 3) Present tag information to the user and let them decide. An example
> is an XML-based search engine that lets people refine queries based on
> tag names.
I never thought of that one. We could present the tag name to the user,
assuming that she will find out what it means, but we could try to fetch
some human-readable information about the tag from the Web. I see that this
RDDL thread is still hot :)...
> What you can't do, and I don't think you ever will be able to do, is
> have somebody ship you a random XML document and have your system
> automatically recognize it as (for example) an invoice, then enter the
> contents of the document into your invoice system. It just won't happen
> without human interaction.
That's what I wanted you to write :). 'The No Silver-Bullet Manifesto'. This
should be written in every book about XML, and printed on the forehead of
marketing people. This would help the XML hype cool down before it
backlashes.
Regards,
Nicolas
|
