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

Re: XPath brain teaser

• To: Richard Tobin <richard@c...>,xml-dev@l...
• Subject: Re: XPath brain teaser
• From: Jonathan Robie <jonathan.robie@d...>
• Date: Wed, 27 Aug 2003 11:11:47 -0400
• In-reply-to: <200308270056.BAA19379@s...>
• References: <0b1301c36c23$36de0140$6401a8c0@snobird>

At 08:56 PM 8/26/2003, Richard Tobin wrote:
> >This is easier to understand if you think of it in terms of set construction
> >and iterators. The two expressions with abbreviations eliminated are:
> >
> >/descendant-or-self::node()/para[1]
> >/descendant::para[1]
>
>You might ask why // means the first of these, rather than the more
>obvious second.  The answer (I think) is precisely so that expressions
>like //para[1] will mean what they do; it's more common to want to
>treat the first paragraph in each section specially than the first
>paragraph in the while document.

Another reason is that we wanted the definition of // to work for both 
elements and attributes:

         //@id
         //para

The expanded definition of // works for both these cases. It's rather hard 
to think of anything  based on the XPath axes that does this any better.

Jonathan 

• References:
  • Re: XPath brain teaser
    • From: "Bob Foster" <bob@o...>
  • Re: XPath brain teaser
    • From: Richard Tobin <richard@c...>

• Prev by Date: Re: RE: xml taxonomy
• Next by Date: RE: XML CMM and ISO9000 compliance? - was A standard approach to glueing together reusableXML fragments in prose?
• Previous by thread: Re: XPath brain teaser
• Next by thread: RE: Re: XML and the Relational Model
• Index(es):
  • Date
  • Thread