Overwriting a specific XML section

[Randy Miller]

Hi guys!

So,I´m quite new here and been working with XML and XSD for just a few weeks. Now I encountered my first problem.Basically,I have 2 XML Files.One of them is an entire XML-File(let´s call it the basic XML file) which you can actually validate against a XSD. The other XML File shows just a section contained in the basic XML file.My goal is,to,somehow overwrite the specific section of the basic XML file with the section given in the additional XML file.So,how can I do this? I´ve searched quite a while on the internet but no result helped me in the end!

Greetings,

Randy

[John Jai]

Randy - welcome to the Ranch

I have not tried it myself - can you check the replaceChild() method replaceChild()

You can also try first removing the node using removeNode() and then adopting the new node from the other document.

To import a Node from other document, you should use adoptNode()(adoptNode or importNode)

[Randy Miller]

Thanks John

So,you´re thinking of the same idea, I thought of before ;)
But in my book, this would be pretty complicated for example when I use a really complex XML and just want to replace a really small part of the XML.
Basically, my first idea(just like yours) was, to load both the entire XML and the replacement XML, read the replacement XML and check out, which nodes would be affected.
I want to keep this programm pretty simple,so how do I then retrieve all the information out of the replacement XML and replace the right nodes in the entire XML?
Or Am I thinking way too complicated and there´s an easier way to solve this problem?

[John Jai]

I can't think of another way Randy than to load both the XMLs and have them as documents. I have used this replace method with importNode() successfully. May be others can help.

You can also show your code and tell us where you think it's failing / making it complex.

[Paul Clapham]

You could quite likely do that with XSLT as well. (But don't ask me to provide an example unless you're going to provide more specific requirements!)

Generally if you want an XSL transform which outputs something only slightly changed from the input, you base your transform on the identity transformation and add in templates which replace specific nodes by something else.

[Randy Miller]

So John,I´ll try your proposal to load both XMLs as Document and try to replace the matching parts. I´ll see how that one works out ;)

Paul,the general XSLT idea doesn´t sound too bad,but that´s over the top for my requirement cause I don´t really use XSLTs.

But yesterday another idea came to my mind.How about,I have both of my XMLs inside two seperate Strings and then search for the corresponding tags trying to replace everything that´s in between?
Just struck my mind, but I don´t know if that´s really a good idea ;)

Maybe John,you could give me a short dummy, a small snippet on the replacing stuff?

I´ll try my best to get this thing to work ;)

[John Jai]

Randy Miller wrote:you could give me a short dummy, a small snippet on the replacing stuff?

Document doc1 = db.parse(new File("C:/test/Person.xml"));
Document doc2 = db.parse(new File("C:/test/Employee.xml"));

Element personElement = doc1.getDocumentElement();
NodeList nl = personElement.getChildNodes();
Node nodeToBeReplaced = null;
for (int i = 0 ; i < nl.getLength(); i++){
	Node childNode = nl.item(i);
	if (childNode.getNodeName().equals("employee")){
		nodeToBeReplaced = childNode;
		break;
	}
}
if (nodeToBeReplaced != null){
	personElement.replaceChild(doc1.importNode(doc2.getDocumentElement(),true),nodeToBeReplaced);
}

Where -
Person.xml
<person> <employee> <name>John</name> <age>27</age> </employee> </person>

Employee.xml
<employee> <name>Paul</name> <age>45</age> </employee>

Result.xml
<person> <employee> <name>Paul</name> <age>45</age> </employee> </person>

[John Jai]

Randy Miller wrote:I have both of my XMLs inside two seperate Strings and then search for the corresponding tags trying to replace everything that´s in between?

Very bad idea to manipulate XML's as Strings... You have loads of API to process XML I believe.

[Randy Miller]

Hi guys!

Thanks for your amazing help so far

That´s just what I thought: Working with Strings to manipulate XMLs is quite a bad idea ;)

Thanks for the code snippet
I´ll try it in just a few minutes
Hopefully it works

But so far,here´s something I tried and I´m not sure why that wouldn´t work.

// Obtain an element in one document; the following method is implemented in
		Document doc1 = parseXmlFile("MyReplacementXML", false);
		NodeList list = doc1.getElementsByTagName("Section To Be Replaced");
		Element element = (Element)list.item(0);

		// Create another document
		Document doc2 = parseXmlFile("XML To Be Altered", false);

		// Make a copy of the element subtree suitable for inserting into doc2
		Node dup = doc2.importNode(element, true);

		// Insert the copy into doc2
		doc2.getDocumentElement().appendChild(dup);

And the parseXML method:

public static Document parseXmlFile(String filename, boolean validating) {
        try {
            // Create a builder factory
            DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
            factory.setValidating(validating);

            // Create the builder and parse the file
            Document doc = factory.newDocumentBuilder().parse(new File(filename));
            return doc;
        } catch (SAXException e) {
            // A parsing error occurred; the xml input is not valid
        } catch (ParserConfigurationException e) {
        } catch (IOException e) {
        }
        return null;
    }

Does anyone see,why this wouldn´t do the trick?Doesn´t look that wrong,right? ;)

Oh yeah,just to give you a short view to the XMLs.

<variableParameters>
      <Load>false</Load>
      <Lib>TRUE</Lib>
      <PARSER>FALSE</PARSER>
    </variableParameters>

That´s the ENTIRE replacement XML,I want to insert in another XML.Note,that no header is defined.(Shouldn´t be a problem!?)

And here´s a snippet of the XML needing to be altered:

 <settings>
    <variableParameters>
        <code>hello</code>
      <LIB>command</LIB>
      <code>...</code>
      <Bearer>...</Bearer>
    </variableParameters>
  </settings>

Here,you have a complete header(not shown here,but believe me,it´s there ;) )
And now,the goal is,just to copy the replacement XML into the big XML.But,the needed section is inside the big XML,not directly at the beginning, which means,you need to search for it ;)
After you found the section,replace the old section with the new one.

Now,I´ll try your provided code,hopefully it works
I think,the problem should be quite clear now.But your provided solution looks quite good ;)
Let´s see,whether or not it does the trick for me ;)

[Randy Miller]

So,just as promised,my update ;)

John,I tried to implement the code snippet you provided,but somehow,it just doesn´t work...
I looked at the XML-files before and after running the application,nothing really changed!?

Maybe I just implemented that bad boy wrong ;)
Take A Look at this:

    try {
	DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
    DocumentBuilder db = dbf.newDocumentBuilder();
    Document doc1;
	try {
		doc1 = db.parse(new File("path to the replacement XML")); 
    Document doc2;
	doc2 = db.parse(new File("path to the XML going to be altered")); 
     
   Element personElement = doc1.getDocumentElement();  
   NodeList nl = personElement.getChildNodes();  
   Node nodeToBeReplaced = null;  
   for (int i = 0 ; i < nl.getLength(); i++){  
	Node childNode = nl.item(i);  
	if (childNode.getNodeName().equals("variableParameters")){  
	    nodeToBeReplaced = childNode;  
	    break;  
	}  
   }  
   if (nodeToBeReplaced != null){  
	personElement.replaceChild(doc1.importNode(doc2.getDocumentElement(),true),nodeToBeReplaced);  
   }
   
	} catch (SAXException e) {
		// TODO Auto-generated catch block
		e.printStackTrace();
	} catch (IOException io) {
		// TODO Auto-generated catch block
		io.printStackTrace();
	}
} catch(ParserConfigurationException pce){
	pce.printStackTrace();
}

Maybe I should try to save the altered file in a new XML?Or does anyone know the trick?

[John Jai]

Did you try saving the altered document to a new XML file.

In case you need the part of code to write document to a file -

Transformer transformer = TransformerFactory.newInstance().newTransformer(); 
transformer.setOutputProperty(OutputKeys.INDENT, "yes"); 
StreamResult result = new StreamResult(new FileWriter("C:\\test\\Output.xml")); 
DOMSource source = new DOMSource(doc1); 
transformer.transform(source, result); 
System.out.println("File saved!"); 

And remember to use print statements to get the flow of the code and check if the document is really being modified...

update - I ran the code you provided and it worked successfully for me

[Randy Miller]

Hi John,

Really??The code worked for you?Now how´s that possible??
Somehow,for me,not...
The application doesn´t even wanna save the new file...
I entered your code sample inside a try-catch-block right after the last if-statement,just like this:

 if (nodeToBeReplaced != null){  
	personElement.replaceChild(doc1.importNode(doc2.getDocumentElement(),true),nodeToBeReplaced); 
	
	try{
	    Transformer transformer = TransformerFactory.newInstance().newTransformer();   
	    transformer.setOutputProperty(OutputKeys.INDENT, "yes");   
	    StreamResult result = new StreamResult(new FileWriter("C:\\Output.xml"));   
	    DOMSource source = new DOMSource(doc1);   
	    transformer.transform(source, result);   
	    System.out.println("File saved!");   
	
	}catch(Exception e){
		e.printStackTrace();

Thanks for the tip,using System.out.print()-commands ;)
Normally,my program is full of those,but I just threw them out to provide a better view of the code ;)
Did you use EXACTLY the same code,I posted?Or did you change one or two things to make it work?And which one?I posted two examples ;)
Or maybe,the replacement-function isn´t handy for larger XML-Files where I want to replace a section deep inside the XML?

[John Jai]

My very first guess would be nodeToBeReplaced might be null. There's where print statements will help you.

[Randy Miller]

Ha,THERE it is ;)
nodeToBeReplaced: nullnullnull...
Good Guess my friend ;)
Now,how´s that possible??I
Looking at the code,I realize,nodeToBeReplaced hasn´t been used,just initialized as null...
Only childNodes is properly use...(I guess ;) )
Therefore,the application correctly terminates.Hm,but how did you get it to work using the code I posted?

[John Jai]

You have to read your code again & again and tell us what is the reason it's stand set as null. Print all childNode's node names prior to this statement.

if (childNode.getNodeName().equals("variableParameters")){    

Again - use print statements to trace the issue.

[Randy Miller]

So,I´ve tried the print-statements to trace the issue and basically,this is how the outprint looks like:

null[#text: 
      ]
null[LoadHexFileAndPutTestKeys_output_full: null]
null[#text: 
      ]
null[TeLibAutoSetC9installParameterDefaultTags: null]

So basically,the application finds the correct nodes(seemingly),but first,there´s a null??And the value is also null??
Now why´s that?Maybe a formatting problem inside the replacement XML??Maybe unknown spaces?
Or perhaps,the XML just lacks a normal XML-Header?

Of course,this is just wild guessing,trying to find an answer ;)

[John Jai]

It's obvious that there is no variableParameters child node which you are looking for. I think you should still ShowSomeEffort to analyze and debug the code.

[Randy Miller]

Of course,there IS a node "variableParameters" John ;)
I tried this 2 different ways,like I posted here and still, I haven´t been able to solve this problem (even though I´m pretty patient) ;)

[James Boswell]

Randy

Might I suggest you refactor lines such as

	personElement.replaceChild(doc1.importNode(doc2.getDocumentElement(),true),nodeToBeReplaced); 

to split method calls over multiple lines.
Reducing chained calls makes your code more readable and easier to debug which may help you out in your current predicament.

[Paul Clapham]

Randy Miller wrote:Of course,there IS a node "variableParameters" John ;)

Just not where you're looking. Your original code shows that you're looking for elements with that name which are children of the document node. So that would mean that the actual node you're looking for is not a child of the document node.

[g tsuji]

I refer to the posting of "Wednesday, February 15, 2012 9:09:21 PM".
This is how the specific block of code should look like to assure general applicability.

   Element personElement = doc1.getDocumentElement();     
   //NodeList nl = personElement.getChildNodes();     
   NodeList nl = personElement.getElementsByTagName("variableParameters");     
   Node nodeToBeReplaced = null;     
   
   /* Replaced by the following
   for (int i = 0 ; i < nl.getLength(); i++){     
    Node childNode = nl.item(i);     
    if (childNode.getNodeName().equals("variableParameters")){     
        nodeToBeReplaced = childNode;     
        break;     
    }     
   }
   */
   if (nl.getLength()!=0) {
    nodeToBeReplaced=nl.item(0);    //or any other findings in document order 
   }    

   if (nodeToBeReplaced != null){     
    //personElement.replaceChild(doc1.importNode(doc2.getDocumentElement(),true),nodeToBeReplaced);     
    nodeToBeReplaced.getParentNode().replaceChild(doc1.importNode(doc2.getDocumentElement(),true),nodeToBeReplaced);
   }  

It conveys at least the precise functioning of replaceChild() and how to find the desired element anywhere as long as we know the NCName (figured as "variableParameters") or the name with a default namespace attached to it.

[John Jai]

This is nice one Tsuji... Thanks

[Randy Miller]

Hi guys!

So,after sooo many hours of debugging, I finally got it to work

Here´s the winner

Document doc1 = parseXmlFile("replacementXML", false);
		NodeList list = doc1.getElementsByTagName("NodeBeingReplaced");
		Element element = (Element)list.item(0);

		// Create another document
		Document doc2 = parseXmlFile("EntireXML", false);

		// Make a copy of the element subtree suitable for inserting into doc2
		Node dup = doc2.importNode(element, true);

		// Insert the copy into doc2
		doc2.getDocumentElement().appendChild(dup);
		
		try{
		    Transformer transformer = TransformerFactory.newInstance().newTransformer();   
		    transformer.setOutputProperty(OutputKeys.INDENT, "yes");   
		    StreamResult result = new StreamResult(new FileWriter("C:\\Output.xml"));   
		    DOMSource source = new DOMSource(doc2);   
		    transformer.transform(source, result);   
		    System.out.println("File saved!");   
		
		}catch(Exception e){
			e.printStackTrace();
		}
		}
	// Parses an XML file and returns a DOM document.
    public static Document parseXmlFile(String filename, boolean validating) {
        try {
            // Create a builder factory
            DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
            factory.setValidating(validating);

            // Create the builder and parse the file
            Document doc = factory.newDocumentBuilder().parse(new File(filename));
            return doc;
        } catch (SAXException e) {
            // A parsing error occurred; the xml input is not valid
        } catch (ParserConfigurationException e) {
        } catch (IOException e) {
        }
        return null;
    }
    
    	// This method visits all the nodes in a DOM tree
    	public static void listMyNodes(Node node, int level) {
    	    // Process node

    	    // If there are any children, visit each one
    	    NodeList list = node.getChildNodes();
    	    for (int i=0; i<list.getLength(); i++) {
    	        // Get child node
    	        Node childNode = list.item(i);

    	        // Visit child node
    	        listMyNodes(childNode, level+1);
    	        System.out.print(list);
    	    }
    }

The trick is pretty easy or no,better: It was a really stupid error...I got a malformed replacement XML and so therefore,the algorithm just wouldn´t work...
So,I opened the XMLs seperately via Firefox and there was the problem,I was looking for for so long

Thanks again for all your help,you´re awesome