[xquery-talk] Collections - family relationships

[David Lee]

Given those truths ...
If you do have a sequence from fn:collection() or another function you *can* find the position of a document in that collection and its 'sibling'

Note: this may be a horribly inefficient thing to do so buyer beware .... 

let $c := fn:collection(),
     $doc := local:pick-a-random-doc( $c ),
     $prev := $c[ fn:index-of( $c , $doc ) - 1 ]   (: Might be bad to do on the first or  only document :)
   return $c


this will answer the question of "what is the previous document in the sequence provided by fn:collection() ...
( you can replace fn:collection() with any function that produces a sequence of documents )




----------------------------------------
David A. Lee
dlee at calldei.com
http://www.xmlsh.org

-----Original Message-----
From: talk-bounces at x-query.com [mailto:talk-bounces at x-query.com] On Behalf Of Michael Kay
Sent: Sunday, January 05, 2014 3:54 PM
To: ihe.onwuka at gmail.com
Cc: talk at x-query.com
Subject: Re: [xquery-talk] Collections - family relationships


On 5 Jan 2014, at 16:50, Ihe Onwuka <ihe.onwuka at gmail.com> wrote:

> If x is the document element of a document in a collection, is it the sibling of y that is the document element of another document in the same collection?

No.
> 
> If not why not?

"Why" questions are very difficult to answer. Do you want a historically accurate answer (was the question debated at a WG meeting, who argued which position, how did the vote go, and why did individual members vote as they did?) - because it's very rarely possible to give one. Or do you want a post-hoc justification (can you think of a sensible reason that anyone might have designed it this way?)

One post-hoc justification is that collections are unordered, whereas axes are always ordered.

Another is that documents may belong to more than one collection.

Michael Kay
Saxonica


_______________________________________________
talk at x-query.com
http://x-query.com/mailman/listinfo/talk