Question

I’m trying to get an attribute id (fileID) from my XML document to use as the filename for my XML split. The split works I just need to extract the fileID to use as the name.

[EDITED] I can read the attribute now but it doesn’t create the last xml file. So in my example it create the first 2 files with the correct name but last fileID “000154OP.XML” isn’t created. Can Anyone Help?

This is my xml document

<root>
 <envelope fileID="000152OP.XML">
   <record id="850">
   </record>
</envelope>
<envelope fileID="000153OP.XML">
  <record id="850">
  </record>
</envelope>
<envelope fileID="000154OP.XML">
  <record id="850">
  </record>
</envelope>
</root>

And here’s my Java code

    public static void splitXMLFile (String file) throws Exception {         
    String[] temp;
    String[] temp2;
    String[] temp3;
    String[] temp4;
    String[] temp5;
    String[] temp6;
    File input = new File(file);         
    DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();         
    Document doc = dbf.newDocumentBuilder().parse(input);
    XPath xpath = XPathFactory.newInstance().newXPath();          
    NodeList nodes = (NodeList) xpath.evaluate("//root/envelope", doc, XPathConstants.NODESET);          
    int itemsPerFile = 1;         

    Node staff = doc.getElementsByTagName("envelope").item(0);

    NamedNodeMap attr = staff.getAttributes();
    Node nodeAttr = attr.getNamedItem("fileID");
    String node = nodeAttr.toString();
    temp = node.split("=");
    temp2 = temp[1].split("^\"");
    temp3 = temp2[1].split("\\.");

    Document currentDoc = dbf.newDocumentBuilder().newDocument();         
    Node rootNode = currentDoc.createElement("root");   
    File currentFile = new File("C:\\XMLFiles\\" + temp3[0]+ ".xml"); 

    for (int i=1; i <= nodes.getLength(); i++) {             
        Node imported = currentDoc.importNode(nodes.item(i-1), true);             
        rootNode.appendChild(imported); 

        Node staff2 = doc.getElementsByTagName("envelope").item(i);
        NamedNodeMap attr2 = staff2.getAttributes();
        Node nodeAttr2 = attr2.getNamedItem("fileID");
        String node2 = nodeAttr2.toString();
        temp4 = node2.split("=");
        temp5 = temp4[1].split("^\"");
        temp6 = temp5[1].split("\\.");

        if (i % itemsPerFile == 0) { 

            writeToFile(rootNode, currentFile);                  
            rootNode = currentDoc.createElement("root");    
            currentFile = new File("C:\\XMLFiles\\" + temp6[0]+".xml");


        }         
    }          
    writeToFile(rootNode, currentFile);     
}    

 private static void writeToFile(Node node, File file) throws Exception {         
     Transformer transformer = TransformerFactory.newInstance().newTransformer();         
     transformer.transform(new DOMSource(node), new StreamResult(new FileWriter(file)));     
 } 

Answer 1

There is a lot of duplication in your code but I have a solution that removes a lot of it. I know there are less complex solutions (for example I don’t think the if (i % itemsPerFile == 0) logic is required, but I do not know all of your requirements, so I have left it in.

The main problems you have were overwriting the last file with wrong data but also that your looping logic was duplicated. A good rule of thumb I go by is whenever I think I might have to duplicate code there is something wrong. Your logic was considering the first <envelope> separately to the remaining <envelope> elements, whereas they should be considered as a group of 3. Then your logic need only to apply the same searching, splitting, matching, importing, etc… to each element in turn.

What complicated matters, is that your input XML file had the same <record id="850"> for each <envelope>. I changed mine to 850, 851 and 852. Running your original code, produced 3 files, 000152OP.xml, 000153OP.xml and 000154OP.xml, but the first one contained the 851 record. So I immediately knew the looping logic was incorrect.

A simpler solution is detailed below, which given your input XML file as the argument produces 3 output files in the same directory (I removed the C:\ hard-coding for simplicity), each with the correct <record> element.

import java.io.*;
import java.util.Random;
import org.w3c.dom.*;
import javax.xml.parsers.*;
import javax.xml.xpath.*;
import javax.xml.transform.*;
import javax.xml.transform.dom.*;
import javax.xml.transform.stream.*;

public class SplitXML {
    public static void main(String[] args) throws Exception {
        File input = new File(args[0]);
        DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
        Document doc = dbf.newDocumentBuilder().parse(input);
        XPath xpath = XPathFactory.newInstance().newXPath();
        NodeList nodes = (NodeList) xpath.evaluate("//root/envelope", doc, XPathConstants.NODESET);
        int itemsPerFile = 1;

        Document currentDoc = dbf.newDocumentBuilder().newDocument();

        for (int i=0; i < nodes.getLength(); i++) {
            Node rootNode = currentDoc.createElement("root");

            Node imported = currentDoc.importNode(nodes.item(i), true);
            rootNode.appendChild(imported);

            Node staff = doc.getElementsByTagName("envelope").item(i);
            NamedNodeMap attr = staff.getAttributes();
            Node nodeAttr = attr.getNamedItem("fileID");
            String filename = nodeAttr.getNodeValue();
            String[] fileParts = filename.split("\\.");

            if (i % itemsPerFile == 0) {
                File currentFile = new File(fileParts[0] + "." + fileParts[1].toLowerCase());
                writeToFile(rootNode, currentFile);
            }
        }
    }

    private static void writeToFile(Node node, File file) throws Exception {
        Transformer transformer = TransformerFactory.newInstance().newTransformer();
        transformer.transform(new DOMSource(node), new StreamResult(new FileWriter(file)));
    }
}

You should read up on Node and String::split as there was unnecessary extra code where a native method already exists (for example [Node::getNodeValue()][3]).

Edit: The source for creating 1000 <envelope> elements that I used to test the above code:

import java.io.*;

public class CreateXML {
    public static void main(String[] args) throws Exception {
        FileWriter fstream = new FileWriter(new File("split.xml"));
        BufferedWriter out = new BufferedWriter(fstream);
        out.write("<root>");
        for (int i = 0; i < 1000; i++) {
            out.write("<envelope fileID=\"000" + i +"P.XML\"><record id=\"" + i + "\"></record></envelope>\n");
        }
        out.write("</root>");
        out.close();
    }
}

I ran java CreateXML to create the input file split.xml and then java SplitXML split.xml to create the 1000 files.