Need to implement Streambased Stax parser

[yatin baraiya]

Hy everyone

Right now i have implement the Staxbased xml parsing , which take as input as inputstream and OutputStream. and generate the file.
But now i wan to implementation it in another stream-based concept.,

Current Implementation:
===============
Right now i had create the xml using the following code

 Document testdata =DocumentBuilderFactory.newInstance().newDocumentBuilder().newDocument();
Element root = testdata .createElement("Workbook");

		Element shet = testdata .createElement("Fan");
		shet .setAttribute("title", "Test information");
		root.appendChild(shet );
		testdata .appendChild(root);
                ......
                .......

Above code generate element for the xml using the Document object
and i will generate the xml file using the testdata object , and make the fileinputstrem object and pass it to my staxparser method , which give me the appropriate file.

But now i do not want to not generate the xml file , and giving whatever write in the dom element pass it to staxbased parser ,
so it will increase the my performance, but i don't know any idea how do i start it?How to start that implementation for it? if any body having know this concept/approach /reference link/solution , please share it with me.


thanks & regards
Yatin Baraiya

[Paul Clapham]

In the example you posted, you are creating a DOM element in your code. There is no parsing being done at all. So there wouldn't be any point in using a StAX parser, because you aren't doing any parsing.

Perhaps you could explain your actual requirement in a bit more detail?

[yatin baraiya]

Hi
Paul Clapham

First of thanks for the reply

As per i had making the Xml file ,check below code

	public File createXMLFile(Document docReport) throws Exception
	{
	File file = File.createTempFile("xmlTempData", ".xml");
	OutputStreamWriter sout = new OutputStreamWriter(new FileOutputStream(file), "UTF-8");
	StringWriter swXmlStr = new StringWriter();
	Transformer tFormer = TransformerFactory.newInstance().newTransformer();
	tFormer.setOutputProperty(OutputKeys.ENCODING, "UTF-8");
	tFormer.transform(new DOMSource(docReport), new StreamResult(swXmlStr));
	sout.write(swXmlStr.toString());
	try
	{
	sout.close();
	}
	catch (Exception e)
	{
	}
	return file;
	}
	

While above code making the xml file
Now i have created one StaxbasedParser java file which have one method ,which take fileinputstream and outputstream, and this method give me all the attribute value of the given xml based fileinputstream , but now i want to not making any xml file and give the dom object directly to my method
i want solution like below

        Element root = testdata .createElement("Workbook");  
        Element shet = testdata .createElement("Fan");  
        shet .setAttribute("title", "Test information");  
       root.appendChild(shet );
       testdata .appendChild(root);

at this stage my dom element have making the below part
<?xml version="1.0" encoding="UTF-8"?> <Workbook> <fan title="Test information">

at this stage i want to parse the above partial xml part using staxbased parsing

Now dom is still running for the next element generating

               colwidth= testdata .createElement("colwidth");
		shet .appendChild(colwidth);
		colwidth.setAttribute("default", "23");

		width = docReport.createElement("Width");
		colwidth.appendChild(width);
		width.setAttribute("col", "0");
		width.appendChild(docReport.createTextNode("2000"));

		width = docReport.createElement("Width");
		colwidth.appendChild(width);
		width.setAttribute("col", "1");
		width.appendChild(docReport.createTextNode("10000"));

now above code again making the below xml portion

<colwidth default="23"> <width col="0">2000</width> <width col="1">10000</width> </colwidth>

now again i need to parse the above portion .

.......
........
.........
..........element added in xml is still going on..

Now you getting my requirement or not

Still let me know if you want to other information?

Thanks & Regards
Yatin Baraiya

[Paul Clapham]

If I understand all of that, what you're saying is that you have a DOM and you want to "parse" that DOM using StAX.

And as I already said, that doesn't make sense. That's because parsing converts the external form -- the actual XML document which is text -- into an internal form, which could be a DOM document or a sequence of StAX events. But here you already have an internal form, so parsing is unnecessary. And not even possible.

If you really wanted to parse that DOM using StAX, then you would have to convert it to the external form (that's called "serialization" by the way, it helps in these discussions if you use the right names for things) and then parse that external form using StAX.

[yatin baraiya]

Can you guide me the approach of serialization in my case?
Will you provide me the Any reference link,example demo etc...

[Paul Clapham]

Serialization? That's the code you posted right after

As per i had making the Xml file ,check below code

[g tsuji]

@op
The StAX input factory has method(s) basically taking on InputStream in general. FileInputStream is just a subclass of it. You are free to use some memory stream such as ByteArrayInputStream to feed to the input factory to create your XMLStreamReader or XMLEventReader. The latter depends on your existing implementation of StAX parsing class.

This is how you can do it without passing the step of making a physical file --- which, if there is no need of it, it would ridiculous to do that and reload it with that kind of power support of the platform available. I make it as an inline integration to you existing code. You should obviously better isolate it in order not to disrupt too much the flow of logic and for the repetitive use of the same functionality.

    //etc etc
    colwidth= testdata .createElement("colwidth");   
    shet.appendChild(colwidth);   
    colwidth.setAttribute("default", "23");   
  
    width = docReport.createElement("Width");   
    colwidth.appendChild(width);   
    width.setAttribute("col", "0");   
    width.appendChild(docReport.createTextNode("2000"));   

    width = docReport.createElement("Width");   
    colwidth.appendChild(width);   
    width.setAttribute("col", "1");   
    width.appendChild(docReport.createTextNode("10000"));  
    colwidth= testdata .createElement("colwidth");
    shet .appendChild(colwidth);
    colwidth.setAttribute("default", "23");

    width = docReport.createElement("Width");
    colwidth.appendChild(width);
    width.setAttribute("col", "0");
    width.appendChild(docReport.createTextNode("2000"));

    width = docReport.createElement("Width");
    colwidth.appendChild(width);
    width.setAttribute("col", "1");
    width.appendChild(docReport.createTextNode("10000")); 

    //I suppose here the element node colwidth is done and you're intended to parse this subtree
    /*
        You can make it to a separate method 
        input: Element elem
        output: ByteArrayInputStream bais
    */
    ByteArrayOutputStream baos=new ByteArrayOutputStream();
    Transformer t=TransformerFactory.newInstance().newTransformer();
    t.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION,"yes");
    t.transform(new DOMSource(doc), new StreamResult(baos));
    ByteArrayInputStream bais=new ByteArrayInputStream(baos.toByteArray()); 

    //you can then feed the bais to your method of createXMLStreamReader() or createXMLEventReader

[g tsuji]

[amendment-typo]
I have the chance to re-read my above post, there is a typo which should be obvious in the context. Sorry!

t.transform(new DOMSource(doc), new StreamResult(baos));

should be read:

t.transform(new DOMSource(colwidth), new StreamResult(baos));

That's the idea that I wanted to convey: feed the subtree with the root colwidth for StAX parsing.