[XML4Lib] Problem with sorting and de-duplication in XSLT

[Cary Gordon]

Try <xsl:sort select="orgName"/> for the sort issue. I believe that it is in
the correct location.

If you want to get rid of line breaks in nodes, you will have to do some
form of search and replace. <xsl:strip-space elements="*"/> just removes
nodes that contain only whitespace.

Cary Gordon
The Cherry Hill Company 


-----Original Message-----
From: xml4lib-bounces at webjunction.org
[mailto:xml4lib-bounces at webjunction.org] On Behalf Of Schlosser, Melanie
Brynn
Sent: Wednesday, March 21, 2007 1:01 PM
To: xml4lib at webjunction.org
Subject: [XML4Lib] Problem with sorting and de-duplication in XSLT

Hi,

I'm using XSLT to generate lists of unique values for elements in a
TEI-encoded encyclopedia that can occur more or less anywhere in the
hierarchy. The lists will be used both for decision-making about access
points, and for generating browse lists. Using the Muenchian method, I've
managed to generate some lists, but there are two problems I haven't been
able to resolve:

1. I can't get the sort function to work, so the lists are in document
order. I've tried using the <xsl:sort> element as a child of <xsl:for-each>
and as a child of <xsl:apply-templates>, and I've tried it with and without
the 'select' attribute, with very little result.

2. The lists aren't completely de-duplicated because the processor sees
"John Adams" and "John [linebreak] Adams" as different values. 
normalize-space() seems like the obvious solution, but I've tried it in the
key, i've put it in a variable, i've put it in select="", i've even put it
in the predicate before the count function. At best I can get it to strip
the linebreaks from the result list after the initial de-duplication, which
makes the results prettier, but doesn't solve the problem.

One of my stylesheets (to generate organization names) looks like this:

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

   <xsl:output method="xml" encoding="UTF-8" indent="yes"/>
   <xsl:strip-space elements="*"/>

<!--The first two templates are just to make sure the initial matching is
working-->
   <xsl:template match="/">
       <foundRoot>
           <xsl:apply-templates/>
       </foundRoot>
   </xsl:template>

   <xsl:template match="text">
       <foundText>
           <xsl:apply-templates/>
       </foundText>
   </xsl:template>


   <xsl:key name="orgName" match="orgName" use="." />

   <xsl:template match="//*">


       <xsl:for-each select="orgName[count(. | key('orgName', .)[1]) = 1]" >
           <xsl:sort />
           <Organization>
               <xsl:value-of select="." />
       </Organization>
       </xsl:for-each>


       <xsl:apply-templates/>


   </xsl:template>

   <xsl:template match="@*|*|text()">
       <xsl:apply-templates select="*"/>
   </xsl:template>

</xsl:stylesheet>

The output looks like this:

<?xml version="1.0" encoding="UTF-8"?>
<foundRoot>
  <Organization>Indiana State
                       Library</Organization>
  <Organization>Wabash College Library</Organization>
  <Organization>Albert Lea
                               College</Organization>
  <Organization>Indiana University</Organization>
  <Organization>Whitewater
                                   Presbyterian Academy</Organization>
  <Organization>Indianapolis Public
                               Library</Organization>
  <Organization>Cornell</Organization>
  <Organization>Oxford</Organization>
  <Organization>Wabash College</Organization>
  <Organization>Yale
                               University</Organization>
  <Organization>Dartmouth</Organization>
  <Organization>Harvard</Organization>
  <Organization>Kansas</Organization>
  <Organization>Michigan</Organization>
  <Organization>Yale</Organization>
  <Organization>Burke and
                                   Howe</Organization>
  <Organization>Liber College</Organization>
  <Organization>Union Literary
                               Institute</Organization>
  <Organization>Indiana State Library</Organization>...[and so on ]

Any ideas?

Thanks!
Melanie Schlosser
Indiana University Digital Library Program

_______________________________________________
XML4Lib mailing list
XML4Lib at webjunction.org
http://lists.webjunction.org/mailman/listinfo/xml4lib