Re: PIs considered harmful Was: XML-SW, a thought experiment

Elliotte Rusty Harold wrote:
> At 3:19 AM -0800 2/23/02, Piotr Kaminski wrote:
> 
> 
>>So what happens if you need processing instructions for your metadata --
>>nested PIs?  How do you validate your DC PIs against a schema?  What if
>>you want to indicate the author of the DC annotations?

> A PI can contain essentially arbitrary XML except for other PIs, so 
> there's no real limit to the structure here. You could put, for 
> example, an entire RDF document inside a PI and then pass it to an 
> XML parser to build a separate tree.  YOu could validate this XML in 
> the same way you'd validate any other XML.

Sounds great, thanks for this reminder!

I'd add that even though current schema languages have chosen to ignore 
PIs and comments, there is no reason why a XML Schema language couldn't 
validate (ie, check that they follow certain patterns and are included 
at certain locations in a document) them if there was a need for such a 
validation.

Actually, it should already be possible to do some validation on PIs 
with Schematron.

Eric
-- 
See you in Seattle.
                                        http://knowledgetechnologies.net/
------------------------------------------------------------------------
Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
http://xsltunit.org      http://4xt.org           http://examplotron.org
------------------------------------------------------------------------

Received on Saturday, 23 February 2002 15:02:12 UTC