Question

我正在尝试将两个不同的XML文件编组/解组到POJOS。第一个XML文件如下所示：

--Network.xml--
<Network>
  <Nodes>
    <Node id="ROD" />
    <Node id="KFI" />
    <Node id="JND" />
  </Nodes>
  <Arcs>
    <Arc fromNode="ROD" />
    <Arc fromNode="JND" />
  </Arcs>
</Network>
---------

使用@XmlID和@XmlIDREF注释，我可以成功填充Arc类以指向它引用的正确节点。

但是，我还必须解析这个XML：

--NetworkInputs.xml--
<NetworkInputs>
  <Flows>
    <Flow toNode="JND" />
    <Flow toNode="ROD" />
  </Flows>
</NetworkInputs>
------

目前，我的程序成功解组了Network对象，但是Network和NetworkInput之间没有连接，允许JAXB“看到”Network中存在的节点。我希望我的Flow对象指向Network类中的正确Node。

我基本上想要这样做： http://old.nabble.com/JAXB-Unmarshalling-and-XmlIDREF-using-different-stores-td14035248.html

我尝试过这样做： http://weblogs.java.net/blog/kohsuke/archive/2005/08/pluggable_ididr.html 它只是不起作用，因为我无法从静态上下文中获取填充网络的节点数据。

甚至可以做这样的事情吗？

Answer 1

这可以使用XmlAdapter完成。诀窍是需要使用Network.xml中的所有节点初始化XmlAdapter并将其传递给与NetworkInputs.xml一起使用的Unmarshaller：

import java.io.File;
import javax.xml.bind.JAXBContext;
import javax.xml.bind.Marshaller;
import javax.xml.bind.Unmarshaller;

public class Demo {

    public static void main(String[] args) throws Exception {
        JAXBContext jc = JAXBContext.newInstance(Network.class, NetworkInputs.class);

        File networkXML = new File("Network.xml");
        Unmarshaller unmarshaller = jc.createUnmarshaller();
        Network network = (Network) unmarshaller.unmarshal(networkXML);

        File networkInputsXML = new File("NetworkInputs.xml");
        Unmarshaller unmarshaller2 = jc.createUnmarshaller();
        NodeAdapter nodeAdapter = new NodeAdapter();
        for(Node node : network.getNodes()) {
            nodeAdapter.getNodes().put(node.getId(), node);
        }
        unmarshaller2.setAdapter(nodeAdapter);
        NetworkInputs networkInputs = (NetworkInputs) unmarshaller2.unmarshal(networkInputsXML);

        Marshaller marshaller = jc.createMarshaller();
        marshaller.setProperty(Marshaller.JAXB_FORMATTED_OUTPUT, true);
        marshaller.marshal(networkInputs, System.out);
    }
}

诀窍是使用XmlAdapter映射Flow上的toNode属性：

import javax.xml.bind.annotation.XmlAttribute;
import javax.xml.bind.annotation.adapters.XmlJavaTypeAdapter;

public class Flow {

    private Node toNode;

    @XmlAttribute
    @XmlJavaTypeAdapter(NodeAdapter.class)
    public Node getToNode() {
        return toNode;
    }

    public void setToNode(Node toNode) {
        this.toNode = toNode;
    }

}

适配器将如下所示。诀窍是我们将一个知道所有节点的已配置的XmlAdapter传递给unmarshaller：

import java.util.HashMap;
import java.util.Map;

import javax.xml.bind.annotation.adapters.XmlAdapter;

public class NodeAdapter extends XmlAdapter<String, Node>{

    private Map<String, Node> nodes = new HashMap<String, Node>();

    public Map<String, Node> getNodes() {
        return nodes;
    }

    @Override
    public Node unmarshal(String v) throws Exception {
        return nodes.get(v);
    }

    @Override
    public String marshal(Node v) throws Exception {
        return v.getId();
    }

}

Answer 2

我的解决方案： ID解析由（不幸的）内部类（com.sun.xml.internal.bind.IDResolver）处理，可以从外部设置。

final Unmarshaller unmarshaller = context.createUnmarshaller();
unmarshaller.setProperty(IDResolver.class.getName(), resolver);

旋转变压器可用于许多unmarshaller的情况。但重点是解析器不会在startDocument上清除它自己作为com.sun.xml.internal.bind.v2.runtime.unmarshaller.DefaultIDResolver的默认实现：

import java.text.MessageFormat;
import java.util.HashMap;
import java.util.Map;
import java.util.concurrent.Callable;

import org.xml.sax.SAXException;

import com.sun.xml.internal.bind.IDResolver;

public final class IDResolverExtension extends IDResolver {
    public static final class CallableImplementation implements Callable<Object> {
        private final Object value;

        private CallableImplementation(final Object value) {
            this.value = value;
        }

        @Override
        public Object call() {
            return value;
        }
    }

    private final Map<KeyAndClass, Object> m = new HashMap<KeyAndClass, Object>();

    @SuppressWarnings("rawtypes")
    @Override
    public synchronized CallableImplementation resolve(final String key0, final Class clazz) throws SAXException {
        assert clazz != null;
        assert key0 != null;
        final KeyAndClass key = new KeyAndClass(clazz, key0);
        final Object value = m.get(key);
        return new CallableImplementation(value);
    }

    static class KeyAndClass {
        public final Class<?> clazz;
        public final String key;

        public KeyAndClass(final Class<?> clazz, final String key) {
            this.clazz = clazz;
            this.key = key;
        }

        @Override
        public int hashCode() {
            final int prime = 31;
            int result = 1;
            result = prime * result + clazz.hashCode();
            result = prime * result + key.hashCode();
            return result;
        }

        @Override
        public boolean equals(Object obj) {
            if (this == obj) {
                return true;
            }
            if (obj == null) {
                return false;
            }
            if (getClass() != obj.getClass()) {
                return false;
            }
            final KeyAndClass other = (KeyAndClass) obj;
            if (!clazz.equals(other.clazz)) {
                return false;
            }
            if (!key.equals(other.key)) {
                return false;
            }
            return true;
        }

    }

    @Override
    public synchronized void bind(final String key0, final Object value) throws SAXException {
        assert key0 != null;
        assert value != null;
        Class<? extends Object> clazz = value.getClass();
        assert clazz != null;
        final KeyAndClass key = new KeyAndClass(clazz, key0);
        final Object oldValue = m.put(key, value);
        if (oldValue != null) {
            final String message = MessageFormat.format("duplicated key ''{0}'' => ''{1}'' - old: ''{2}''", key, value,
                    oldValue);
            throw new AssertionError(message);
        }
    }
}

使用JAXB从两个XML文件交叉引用XmlID

2 个答案: