Question

Can you please suggest the best approach for my requirement(below) in Java. I have tried using XPath but I had to recreate the whole XML. I am hoping if there is a simple way of doing this.

I want to remove nodes whose xnuy != 1

Sample input XML:

<object-stream>
  <com.vo.ShapeEventsVO>   
    <mtx>198.0</mtx>
    <mty>79.0</mty>
    <xnuy>1</xnuy>
  </com.vo.ShapeEventsVO>
 <com.vo.ShapeEventsVO>
    <mtx>198.0</mtx>
    <mty>79.0</mty>
    <xnuy>1</xnuy>
  </com.vo.ShapeEventsVO>
 <com.vo.ShapeEventsVO>  
    <mtx>198.0</mtx>
    <mty>79.0</mty>
    <xnuy>2</xnuy>
  </com.vo.ShapeEventsVO>

</object-stream>

output XML:

<object-stream>
  <com.vo.ShapeEventsVO>   
    <mtx>198.0</mtx>
    <mty>79.0</mty>
    <xnuy>1</xnuy>
  </com.vo.ShapeEventsVO>
 <com.vo.ShapeEventsVO>
    <mtx>198.0</mtx>
    <mty>79.0</mty>
    <xnuy>1</xnuy>
  </com.vo.ShapeEventsVO>
</object-stream>   

Answer 1

Write a trivial SAX document handler — on every END ELEMENT event for com.vo.ShapeEventsVO, look at the xnuy, if it fits, write that ShapeEventsVO into a new document.

You will need to track only the most recent mty, mtd and xnuy to have all the data you need to produce the new, xnuy==1 document.