• To: Michael Kay <mike@s...>
• Subject: Re: Collection-Valued Subexpressions?
• From: 'Alan Gutierrez' <alan-xml-dev@e...>
• Date: Sun, 28 Aug 2005 19:55:10 -0400
• Cc: xml-dev@l...
• Mail-followup-to: Michael Kay <mike@s...>, xml-dev@l...
• User-agent: Mutt/1.4.1i

* Michael Kay <mike@s...> [2005-08-28 14:24]:
>  
> >     The problem with XPath 1.0 appears to be that:
> > 
> >         /a[b = 5] 
> > 
> >     Returns all matches while.. 
> > 
> >         /a[b + 1 = 6]
> > 
> >     .. returns only the first match.
> > 
> >     The fix in XPath 2.0 appears to be to return an error for the
> >     latter if there is more than one match.
> > 
> >     Why is it not possible to return all /a[b + 1 = 6] ?
> 
> For that to work, b+1 would have to return a set of numbers, one for each b.
> What then would b+c return? We did consider (very early on) a suggestion
> that it should return the set 
> 
> for $b in b, $c in c return $b + $c
> 
> but we decided this would quickly produce unmanageable results; and in any
> case it would be impossible to define it in a way that was backwards
> compatible.

    Now I see. Thank you.

--
Alan Gutierrez - alan@e...
    - http://engrm.com/blogometer/index.html
    - http://engrm.com/blogometer/rss.2.0.xml

• Prev by Date: RE: Collection-Valued Subexpressions?
• Next by Date: RE: Names As Types
• Previous by thread: Collection-Valued Subexpressions?
• Next by thread: Substitution Groups - Use 'em or lose 'em?
• Index(es):
  • Date
  • Thread