Re: About Simple Types

A revised schema recommendation 1.1 is being prepared by the workgroup. As 
you can see in the draft available at [1], the lexical forms are indeed 
being described more formally using EBNF and regexps [2].  I suggest you 
look at some of the specific definitions therein.

Of course, the schema datatypes provide value spaces along with lexical 
forms.  Characteristics such as "order" are usually defined on the value 
spaces, and only indirectly on the lexical forms.  So, the suggestion to 
use BNF for the lexical forms is a good one and is essentially being 
adopted.  There are other important characteristics of the schema 
datatypes that are not well expressed using variations on BNF.

Noah

[1] http://www.w3.org/TR/2005/WD-xmlschema11-2-20050224/datatypes.html
[2] 
http://www.w3.org/TR/2005/WD-xmlschema11-2-20050224/datatypes.html#Intro

--------------------------------------
Noah Mendelsohn 
IBM Corporation
One Rogers Street
Cambridge, MA 02142
1-617-693-4036
--------------------------------------








"Thomas Li" <tli@corporola.com>
Sent by: www-xml-schema-comments-request@w3.org
04/25/05 10:32 AM

 
        To:     <www-xml-schema-comments@w3.org>
        cc:     (bcc: Noah Mendelsohn/Cambridge/IBM)
        Subject:        About Simple Types


Comments:
 
For XML Schema, I think it is equivalent to context free grammar, CFG. So 
the simple types should be regular expression for defining tokens in 
language, complex types should be BNF productions. Both regular expression 
and BNF or CFG are well defined and concise. Why don't we adopt and 
standardize them? 
 
The central task, I think, is to make XML schema shorter/concise and 
expressive as CFG using known terms.
 
I have a feeling of standard/specification/recommendation and/or term 
explosion.
 
Thank you for consideration.
 
Thomas Li, Ph. D.
 
Corporola Inc.

Received on Tuesday, 26 April 2005 02:23:36 UTC