Re: [xsl] Defining and operating on types (Was: Re: [xsl] Why no namespace node KindTest?)

[Florent Georges <darkman_spam@xxxxxxxx>]

Dimitre Novatchev wrote:

>   2. In the meantime of approximately 6 years or so, for
>      any generic-type function that might be interested in
>      manipulating the types of its arguments, one way to
>      pass both the argument and its type is using a tuple,
>      consisting of the constructor function for that type,
>      followed by the actual arguments to the constructor,
>      from which to instantiate the value of the argument.

  If we are speaking of XSLT 3.0, I'd see instead types as
first-class objects, and no need for a tuple, the type being
a property in the data model, accessible throught a function.

>      Passing a type-constructor as an argument is an
>      extremely powerful design pattern!

  For sure!

  But before that, I think nested sequences (or similar)
could be very useful.  That would permit to build libraries
that define their own structures and the functions to
manipulate them (so their own types and the API to
manipulate such black boxes).

  I wonder if this could be implemented as an extension,
something like EXSLT for XSLT 2.0, as it'd have a intimate
relation to the data model itself.

  Regards,

--drkm























	
 p5.vert.ukl.yahoo.com uncompressed/chunked Sun Aug 27 20:13:41 GMT 2006 
	
		
___________________________________________________________________________ 
Dicouvrez un nouveau moyen de poser toutes vos questions quelque soit le sujet ! 
Yahoo! Questions/Riponses pour partager vos connaissances, vos opinions et vos expiriences. 
http://fr.answers.yahoo.com