[xsl] Re: sort / unsort

[Dimitre Novatchev <dnovatchev at yahoo dot com>]

--- "McKeever, Marty" <marty dot mckeever at bankofamerica dot com>
wrote:

 
> Interesting problem, i've been wracking my brain to solve.
> 
> >From a group of nodes, output the first five in a sorted order, and
> dump the
> remainder in document order.
> 
> that is:
> 	<foo>I</foo>
> 	<foo>F</foo>
> 	<foo>E</foo>
> 	<foo>D</foo>
> 	<foo>A</foo>
> 	<foo>C</foo>
> 	<foo>B</foo>
> 	<foo>H</foo>
> 	<foo>G</foo>
> 
> would output:
> 	<foo>A</foo>
> 	<foo>B</foo>
> 	<foo>C</foo>
> 	<foo>D</foo>
> 	<foo>E</foo>
> 	<foo>I</foo>
> 	<foo>F</foo>
> 	<foo>H</foo>
> 	<foo>G</foo>
> note that only the first 5 are sorted.
> the remainder are in document order - less the 5 sorted items.
> 
> TIA,
> Marty


Marty,

You can use one of my generic templates for partial sort:

http://www.topxml.com/code/default.asp?p=3&id=v20011001235959

its complexity is O(N*log2(k)), for finding the first k elements out of
N.

It doesn't produce a nodeset with copies of the k nodes -- just a
string containing the positions of these k nodes in the original
node-set.

This is also convenient for the second part of your problem -- just
copy all nodes, whose position() in the original node-set is not one of
the positions, contained in the result of the partial sort.

Note: Please, note that your problem only has meaning if it is
guaranteed that there are nore duplicate node values -- in case there
are, it is difficult to define what is "the first k nodes".



=====
Cheers,

Dimitre Novatchev.
http://fxsl.sourceforge.net/ -- the home of FXSL

__________________________________________________
Do You Yahoo!?
HotJobs - Search Thousands of New Jobs
http://www.hotjobs.com

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list