[
Lists Home |
Date Index |
Thread Index
]
On Friday, 25 April 2003 at 01:54, Dare Obasanjo wrote:
> Or he could just define a mapping between the input XML and his
> relational tables using a number of XML mapping technologies that target
> his database of choice. For instance if he was a user Microsoft's SQL
> Server 2000 this could easily be handled by SQLXML
> (http://msdn.microsoft.com/sqlxml).
This is the point of the article though: every different schema has a
different mapping which is fine until they start being refined and
refactored over time. Schema mapping works well for static xml, but
isn't suited to extensible, modular xml.
> Also in we are close to having the
> major relational databases providing native XML storage and query which
> makes this point moot.
Storing the XML in a database doesn't remove the problem of mapping
into the business, it just postpones it.
> XSLT is your friend.
In the scenario I discuss XSLT is isomorphic to SAX which is
isomorphic to DOM. Thye all require you to change the processor when
you extend the XML.
> Soon we'll have XQuery and this will be even more
> of a moot point.
Not sure if this is true. Can anyone demonstrate that I don't need
knowledge of the structure of the document to query it?
- Ian <iand@internetalchemy.org>
"The great tragedy of Science - the slaying of a beautiful hypothesis by an ugly fact."
|
