[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: xml-entity PI

• To: XML-Dev <xml-dev@l...>
• Subject: Re: xml-entity PI
• From: "G. Ken Holman" <gkholman@C...>
• Date: Sat, 23 Oct 2004 14:10:42 -0400
• In-reply-to: <1098466497.728.104.camel@localhost>

At 2004-10-22 03:34 -0700, James William Pye wrote:
>Yes, I do have something specific in mind, a pair of PIs:
>
><?xml-entity name='ent' value='val'?>
>and
><?xml-entities type='text/xml-dtd' href='someURI'?>
><?xml-entities type='text/xml-entities' href='someURI.xml'?>

Processing instructions exist to communicate with the processing 
application, yet entities are an aspect of the syntax and processing 
applications never see the syntax, only the XML processor in processing 
applications see the syntax.

Moreover, no XML file with entity references using the above declarations 
would be well-formed because existing XML processors seeing the entity 
references would not have associated XML 1.0 or XML 1.1 declarations for 
them.  Therefore, you cannot call such a file an XML file.

An example of where a standardized PI is used is stylesheet association 
http://www.w3.org/1999/06/REC-xml-stylesheet-19990629 where the processing 
application has the choice of following up on the "hint" and has nothing to 
do with the syntactic representation of the information.

An example of where new standardized processing is proposed before 
application processing is XInclude 
http://www.w3.org/TR/2004/PR-xinclude-20040930 yet this is done without 
violating any syntactic restrictions of an XML document and only adds a 
processing model so that applications with processors respecting XInclude 
will produce the same result information set in advance of the application 
acting on the information.

>With the many validation mechanisms available today, it seems a bit
>obtuse to rely solely on DTDs for sources of entities.

Ummmmmmm ... with an existing method of doing entities why introduce a 
different way of doing them?  Any different way would be an incompatible 
way.  Your proposed way would render all existing XML processors incapable 
of reading the document, since you've violated syntactic rules that are 
already standardized.  It wouldn't be XML.

Examples like stylesheet association and XInclude add features and 
functionality without breaking anything.

I hope this helps.

............................... Ken

--
World-wide on-site corporate, govt. & user group XML/XSL training.
G. Ken Holman                 mailto:gkholman@C...
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/x/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Breast Cancer Awareness  http://www.CraneSoftwrights.com/x/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal

• Follow-Ups:
  • Re: xml-entity PI
    • From: James William Pye <flaw@r...>

• References:
  • xml-entity PI
    • From: James William Pye <flaw@r...>

• Prev by Date: Re: Conformance testing and the XML-test-suite
• Next by Date: Re: xml-entity PI
• Previous by thread: xml-entity PI
• Next by thread: Re: xml-entity PI
• Index(es):
  • Date
  • Thread