Differences in XML outputs when transforming XML using both XMLSpy and Java

[John Paul Hoey]

I have the following XML:

<Catalog> <CD> <TITLE>Empire Burlesque</TITLE> <ARTIST>Bob Dylan</ARTIST> <COUNTRY>USA</COUNTRY> <COMPANY>Columbia</COMPANY> <PRICE>10.90</PRICE> <YEAR>1985</YEAR> </CD> <CD> <TITLE>Empire Burlesque</TITLE> <ARTIST>Bob Dylan</ARTIST> <COUNTRY>USA</COUNTRY> <COMPANY>Columbia</COMPANY> <PRICE>10.90</PRICE> <YEAR>1985</YEAR> </CD> <CD> <TITLE>Hide your heart</TITLE> <ARTIST>Bonnie Tyler</ARTIST> <COUNTRY>UK</COUNTRY> <COMPANY>CBS Records</COMPANY> <PRICE>9.90</PRICE> <YEAR>1988</YEAR> </CD> <CD> <TITLE>Empire Burlesque</TITLE> <ARTIST>Bonnie Tyler</ARTIST> <COUNTRY>USA</COUNTRY> <COMPANY>Columbia</COMPANY> <PRICE>10.90</PRICE> <YEAR>1985</YEAR> </CD> <CD> <TITLE>Greatest Hits</TITLE> <ARTIST>Dolly Parton</ARTIST> <COUNTRY>USA</COUNTRY> <COMPANY>RCA</COMPANY> <PRICE>9.90</PRICE> <YEAR>1982</YEAR> </CD> </Catalog>

And I want to remove the duplicate records from the XML that contain the same <TITLE> and <ARTIST> elements so that I end up with the following XML output:

<Catalog> <CD> <TITLE>Empire Burlesque</TITLE> <ARTIST>Bob Dylan</ARTIST> <COUNTRY>USA</COUNTRY> <COMPANY>Columbia</COMPANY> <PRICE>10.90</PRICE> <YEAR>1985</YEAR> </CD> <CD> <TITLE>Hide your heart</TITLE> <ARTIST>Bonnie Tyler</ARTIST> <COUNTRY>UK</COUNTRY> <COMPANY>CBS Records</COMPANY> <PRICE>9.90</PRICE> <YEAR>1988</YEAR> </CD> <CD> <TITLE>Empire Burlesque</TITLE> <ARTIST>Bonnie Tyler</ARTIST> <COUNTRY>USA</COUNTRY> <COMPANY>Columbia</COMPANY> <PRICE>10.90</PRICE> <YEAR>1985</YEAR> </CD> <CD> <TITLE>Greatest Hits</TITLE> <ARTIST>Dolly Parton</ARTIST> <COUNTRY>USA</COUNTRY> <COMPANY>RCA</COMPANY> <PRICE>9.90</PRICE> <YEAR>1982</YEAR> </CD> </Catalog>

The expected output works when I use the following stylesheet on the XML and transform it using XML Spy:

<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"> <xsl:output method="xml" version="1.0" encoding="UTF-8" indent="yes"/> <xsl:strip-space elements="*"/> <xsl:template match="@*|node()"> <xsl:copy> <xsl:apply-templates select="@*|node()"/> </xsl:copy> </xsl:template> <xsl:template match="CD[TITLE = following::CD/TITLE and ARTIST = following::CD/ARTIST]"/> </xsl:stylesheet>

However when I pass this transformation to my Java code, the duplicates are removed however the last record within my XML is also removed. I have tried this with different XML files and it always removes the duplicates along with the last record - I have no idea why the last record is also being removed.

My Java code is as follows:

import java.io.File;
import java.io.FileWriter;
import java.io.IOException;
import javax.xml.transform.Transformer;
import javax.xml.transform.TransformerConfigurationException;
import javax.xml.transform.TransformerException;
import javax.xml.transform.TransformerFactory;
import javax.xml.transform.stream.StreamSource;
import org.dom4j.Document;
import org.dom4j.DocumentException;
import org.dom4j.io.DocumentResult;
import org.dom4j.io.DocumentSource;
import org.dom4j.io.SAXReader;
import org.dom4j.io.XMLWriter;


public class CreateXml 
{

    private Document doc, newDoc;
    private final String stylesheetPath = "C:\\transformation.xsl";
    private final String inputFile = "C:\\Xml_File.xml";
    private final String outputFile = "C:\\Transformed_File.xml";
    

    public static void main(String[] args) 
    {
        
        CreateXml xml = new CreateXml();
        xml.RemoveDuplicates();
        
    }
    
    public void RemoveDuplicates()
    {   
        
        try 
        {
            // Get XML File
            File xmlFile = new File(inputFile);
            
            // Parse XML Document
            doc = ParseDocument(xmlFile);
            
            // Get newly styled XML File
            newDoc = StyleDocument(doc);
            
            // Write XML to file
            XMLWriter writer = new XMLWriter(new FileWriter(outputFile));
            writer.write(newDoc);
            writer.close();
            
        } 
        catch (IOException ex) 
        {
            ex.printStackTrace();
        }
    } 
    
    
    public Document ParseDocument(File fileName)
    { 
        
        Document document = null;
        
        try 
        {
            SAXReader reader = new SAXReader();
            document = reader.read(fileName);
        } 
        catch (DocumentException ex) 
        {
            ex.printStackTrace();
        }
        
        return document;
  
    }
    
    public Document StyleDocument(Document document)
    {

        Document transformedDoc = null;
        
        
        try 
        {
            // load the transformer using JAXP
            TransformerFactory factory = TransformerFactory.newInstance();
            Transformer transformer = factory.newTransformer(new StreamSource(stylesheetPath));
            
            // now lets style the given document
            DocumentSource source = new DocumentSource(document);
            DocumentResult result = new DocumentResult();
            transformer.transform(source, result);
            
            // return the transformed document
            transformedDoc = result.getDocument();
            
        } 
        catch (TransformerConfigurationException ex) 
        {
            ex.printStackTrace();
        } 
        catch (TransformerException ex) 
        {
            ex.printStackTrace();
        }
        
        return transformedDoc;
   
    }
    
}

Has anyone got any idea as to why when I transform the XML using XMLSpy it works but when I transform the XML using my Java code the last record is being removed?

Thanks in advance

[g tsuji]

[0] Bear in mind that dom4j is very much lacking continuous support and dated, unfortunately for those who like it as it has certainly its own merit.

[1] The problem certainly stamps from the outdated/incomplete/not-fully-debugged implement of xpath/xslt engine. The xmlspy rendering based on your description is certainly a better or a more to-date interpretation of the recommendation. A hack to remedy the dated implementation is to add a node-test condition on the existence of following::CD. Like this.
<!-- <xsl:template match="CD[TITLE = following::CD/TITLE and ARTIST = following::CD/ARTIST]"/> --> <!-- not really recommended to do, just a hack... --> <xsl:template match="CD[following::CD and (TITLE = following::CD/TITLE and ARTIST = following::CD/ARTIST)]"/>
[1.1] Though it should make it work, this is not what I would recommend anyone doing because that would induce one to learn it the wrong way.

[2] What you could do is simply download or make yourself available the new xalan-j, say from Apache such as 2.7.x+, and make its jars in the classpath setup. The main needed jars are xml-api.jar, xalan.jar and xercesImpl.jar if you want to be more specific. After that, the xslt should function correctly without the hack shown in [1].

[3] Just a side-note, the xslt would not complete deliver what you seem to suggest it doing because the identified following::CD may not be the same in the node-test(s). But that's another story not of the  immediate concern.

[John Paul Hoey]

Hi Tsuji,

It was indeed what you suggested. I downloaded and added the most recent .jars as you suggested and it worked as expected.

Thanks again for taking the time out to help me which is really appreciated.