Java xml解析器循环

Question

I have this xml file:

<root>
    <application>
        <interface />
        <interface />
    </application>

    <datatransmit>
        <interface />
    </datatransmit>
</root>

What I am trying to do is first do a loop through the interfaces within the <application> tags and then another loop through the interfaces withing the <datatransmit> tags.

I tried this with this Java code:

NodeList application = doc.getElementsByTagName("application");
for (int i = 0; i < application.getLength(); i++) {
    NodeList interfaces = doc.getElementsByTagName("interface");
    for (int j = 0; j < interfaces.getLength(); j++) {
        do some stuff...
    }
}

I noticed that with this loop, it goes through a loop of all the interface elements. Not just the interfaces withing the application tags but also the interfaces within datatransmit . What would be a way to solve this?

Answer 1

See the javadoc here

getElementsByTagName() returns all descendants , and since you call doc.getElementsByTagName() you get all descendants of doc which match the element name, rather than all descendants of your application element

Answer 2

Nearly there.

Your problem is using doc as your root again in:

NodeList interfaces = doc.getElementsByTagName("interface");

Meaning it will search the whole document. Instead, you should use the getElementsByTagName method on the application Element, to limit the range of your search:

NodeList application = doc.getElementsByTagName("application");
for (int i = 0; i < application.getLength(); i++) {
  Element applicationElement = (Element) application.item(i);
  NodeList interfaces = applicationElement.getElementsByTagName("interface");
  for (int j = 0; j < interfaces.getLength(); j++) {
    do some stuff...
  }
}

Answer 3

You need to get the list of application nodes from here:

for (int i = 0; i < application.getLength(); i++) {
   Node appNode = application.item(i);
   ....
}

and inspect the name/tag of the node via getNodeName() for the value interface .

That applies for getting interface as a child of application . If interface only occurs under application , then you can skip the first step and simply do

NodeList interfaces = doc.getElementsByTagName("interface");

Perhaps a more concise/flexible solution would be to use XPath , with a path such as /root/application/interface ?

Answer 4

As you are lopping through all elements from document you are getting all elements.

doc.getElementsByTagName("interface");

You should get elements from each application tag object.