SELECT * WHERE {
 	?id <http://www.w3.org/2008/05/skos#prefLabel> ?preferredLabel .
}
ORDER BY DESC( ?preferredLabel  )
limit 100

TADA! The results of that query are now ordered by the ?preferredLabel placeholder. Except that they’re actually being ordered by the ASCII code of that text in the preferredLabel – so you end up with a case-sensitive ORDER BY. That means that all the uppercase letters will appear before lowercase letters (‘Foo’ comes before ‘bar’, even though you would expect the order to be ‘bar’, then ‘Foo’).

Luckily, the fine contributors to the OpenRDF project have a solution. In the sesame-ext directory of their subversion repository lives a little project called “xpath-functions”. You can get a little bit of the background from this mailing list thread – but suffice it to say that this project has several classes that implement the Function interface allowing you to perform xpath functions like:

• Concat


• Date


• DateTime


• LowerCase


• UpperCase

I wasn’t able to find a JAR file already built and ready for download but it’s easy enough to build the JAR from scratch yourself.

First, check out the source code for sesame-ext xpath-functions project (http://repo.aduna-software.org/svn/org.openrdf/sesame-ext/xpath-functions/trunk/)

Make a few updates to the pom.xml in the checked out code – update the sesame version placeholder to the current version you’re using (in this case we’re working with version 2.3.1):

 
<properties>
	<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
	<project.reporting.outputEncoding>UTF-8</project.reporting.outputEncoding>
 
	<sesame.version>2.3.1</sesame.version>
	<slf4j.version>1.5.6</slf4j.version>
</properties>

Add the aduna maven repository to the pom.xml:

 
 <repositories>
   <repository>
     <id>aduna</id>
     <name>aduna</name>
     <url>http://repo.aduna-software.org/maven2/releases</url>
   </repository>
</repositories>

and then run:



mvn package



from the command line to build the jar file.

If you’re working with a local Sesame repository, you can just add the JAR to your classpath. If you’re using the openrdf-sesame WAR file running on an application server to host your repository, then you’ll need to add the JAR to the classpath of that WAR file on the server.

Once you have the JAR file in the appropriate spot, we can leverage the xpath functions in the ORDER BY clause, like this:

 
PREFIX fn: <http://www.w3.org/2005/xpath-functions#>
SELECT * WHERE {
 	?id <http://www.w3.org/2008/05/skos#prefLabel> ?preferredLabel .
}
ORDER BY DESC( fn:lower-case(?preferredLabel)  )
limit 100

Notice the “PREFIX: fn” line and the “fn:lower-case(?preferredLabel)” in the ORDER BY clause. Now the ORDER BY will be applied to the results in a case-insensitive manner!