XSLT - Sorting, Limiting Nodes

Hi,

I have a long XML document that I'm using to write and sort news for my personal website that I transform with XSLT and pre-process with PHP/Sablotron to produce valid XHTML.

But I don't want to display *every* node in the tree - only the first five full entries, and then just the headlines for the rest.

Is there a way to use XSL to transform this page thusly? Then of course I'd need to have visitors click on the headlines to traverse to just *that* node (also pre-processed with PHP/Sablot) on a results page.

My XML document is here:
http://64.191.4.184/home.xml

and the XSL transformation that I'm using for it is here:

http://64.191.4.184/home.xsl

And if you simply go to:

http://64.191.4.184/index.php

you can see them processed with Sablot, CSS, etc.

-- Jough

XPATH will help you accomplish this..

here's an introduction:
http://www.devshed.com/Server_Side/XML/XPath/page1.html


you're links are not loading for me right now, so this is all i can say at the moment, but this should get you started

Thanks for your reply. I had read that article, but it doesn't really go in depth enough for my needs (like many articles on that site - great introductions if you've never heard of a technology, but they generally don't give you enough info to be useful for a real-world application).

Anyway, traversing the nodes isn't really a problem - I just can't seem to find a way to limit node retreival in XSLT (using XPath or otherwise) as I would in say, a SQL db call:

SELECT field FROM table WHERE var='value' LIMIT 0,5;

Is there way to limit node traversal in XSL/XPath/XLink/XPointer/XWhatever?

-- Jough

i'm not an expert, but
how about something like this:


<?xml version="1.0"?>
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:template match="/">


<xsl:for-each select="//adjust-your-position">

<xsl:if test="position() < 6">
<xsl:value-of select="@value"/>
</xsl:if>

</xsl:for-each>
</xsl:template>
</xsl:stylesheet>

Well, I was doing something similar, Jason:



But there's a bug in Sablotron that parses child nodes of the root element incorrectly - it parses whitespace text as a child node, regardless of whether there's an <xml:strip-space elements="*" /> transormation or not.

So the above works with the MS XML 4 parser but not with Sablotron. I've e-mailed them to see if they have a fix or workaround. I'll post it here when I get a reply.

-- Jough

P.S. And yes, I suppose I could just double my position() test number to account for the whitespace text nodes, but that would be wrong, especially if I want to code for standards compliance across multiple platforms - it's much better to get the Ginger Alliance people (those wacky Czechs who make Sablotron) to get their parser up to spec.

ahhhhh,

good to know thanks for the heads-up.

let me know ..

sorry, i couldn't help you out!!


================
justin_dago

That's okay. Actually, I wound up doing it a slightly different way, which selects the child nodes of the root more exactly:



That way, I specifically only select the "entry" elements whose positions are tested in the brackets. Sablotron wrongly implements the parsing of white space (they say that's how it *should* be done, but the W3C spec, as well as every other parser on the market that I've tested, seems to disagree with them) but either way, you can always test using element[child/test/condition].

Good to know.

Cheers,

-- Jough

Cool, Thanks jough!!

Have fun!




justin.dago