[Home] [By Thread] [By Date] [Recent Entries]

  • From: "Rushforth, Peter" <Peter.Rushforth@N...>
  • To: "Michael Kay" <mike@s...>,"Costello, Roger L." <costello@m...>,<xml-dev@l...>
  • Date: Mon, 9 Feb 2009 10:50:34 -0500

> Of course, using child elements is better from a purist point 
> of view. But if you look at how SVG or GML represent 
> polygons, doing it with child elements would give a 
> horrendous explosion of document size and parsing time.
> 
> (However, the micro-syntax that SVG and GML use isn't XSD 
> list syntax, because they are dealing with a sequence of 
> pairs of numbers. So you can't take advantage of 
> schema-awareness in XSLT in quite the same way; to process 
> these vocabularies you do have to parse the data at 
> application level. But that can be done very easily, you only 
> have to write the function library to handle the syntax once.)

GML 2.x uses the xs:string data type for coordinates, where the strings 
are space-separated tuples (or triples) of x,y(,z) coordinates.

GML 3.x uses XSD xs:list syntax to represent coordinates, with the
number 
of dimensions to be interpeted of the list stored in an "dimension" 
attribute.  This allows XSLT/XPATH 2 GML users to benefit from 
schema awareness.
 
Cheers,
Peter Rushforth


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index]


Site Map | Privacy Policy | Terms of Use | Trademarks
Free Stylus Studio XML Training:
W3C Member