getChildNodes 給出意外結果

Question

我的 XML 看起來像這樣-

<collected_objects>
        <object flag="complete" id="objId" version="1">
          <variable_value variable_id="varId">ValueGoesHere</variable_value>
          <reference item_ref="2"/>
        </object>
        <object comment="objComment" flag="complete" id="objId" version="1">
          <reference item_ref="1"/>
        </object>
</collected_objects>

我正在使用以下代碼處理它-

Document dom = parser.getDocument();
    NodeList collected_objects = dom.getElementsByTagName("object");
    System.out.println("Number of collected objects are " + collected_objects.getLength());

        for (int i = 0; i < collected_objects.getLength(); i++) {

            Node aNode = collected_objects.item(i);
            //get children of "objects"         
            NodeList refNodes = aNode.getChildNodes();

            System.out.println("# of chidren are " + refNodes.getLength());

            //print attributes of "objects"

            NamedNodeMap attributes = aNode.getAttributes();
            for (int a = 0; a < attributes.getLength(); a++) {
             Node theAttribute = attributes.item(a);
             System.out.println(theAttribute.getNodeName() + "=" + theAttribute.getNodeValue());

        }

}

它輸出為-

Number of collected objects are 2
# of chidren are 5
flag=complete
id=objId
version=1
# of chidren are 3
comment=objComment
flag=complete
id=objId
version=1

我的問題是為什么“孩子的數量”分別是 5 和 3？ 我不應該分別期待 2 和 1 嗎？ 因為第一個對象有“ variable_value ”和“ reference ”，第二個對象只有“ reference ”

從本質上講，我的目的是處理“對象”的孩子。

Answer 1

確保 <object> 節點子節點之間沒有空格。 空格被視為子節點並按原樣返回。

測試是否

childNode.getNodeType() == Node.ELEMENT_NODE

應該夠了。

Answer 2

那是因為每個子節點之間有 2 個TEXT_NODE ( #text )。

以下包括文本節點及其相應的值。

<object flag="complete" id="objId" version="1">
    <TEXT_NODE />
    <variable_value variable_id="varId">ValueGoesHere</variable_value>
    <reference item_ref="2"/>
    <TEXT_NODE />
</object>

這可以通過修改您的代碼來驗證：

DocumentBuilderFactory dbFactory = DocumentBuilderFactory.newInstance();
        DocumentBuilder dBuilder = dbFactory.newDocumentBuilder();
        Document dom = dBuilder.parse(new ByteArrayInputStream(S.getBytes()));
        NodeList collected_objects = dom.getElementsByTagName("object");
        System.out.println("Number of collected objects are "
                + collected_objects.getLength());

        for (int i = 0; i < collected_objects.getLength(); i++) {

            Node aNode = collected_objects.item(i);
            // get children of "objects"
            NodeList refNodes = aNode.getChildNodes();

            System.out.println("# of chidren are " + refNodes.getLength());

            //
            for (int x = 0; x < refNodes.getLength(); x++) {
                Node n = refNodes.item(x);
                System.out.println(n.getNodeType() + " = " + n.getNodeName() + "/" + n.getNodeValue());
            }

            // print attributes of "objects"

            NamedNodeMap attributes = aNode.getAttributes();
            for (int a = 0; a < attributes.getLength(); a++) {
                Node theAttribute = attributes.item(a);
                System.out.println(theAttribute.getNodeName() + "="
                        + theAttribute.getNodeValue());

            }

        }

輸出：

Number of collected objects are 2
# of chidren are 5
3 = #text/          
1 = variable_value/null
3 = #text/          
1 = reference/null
3 = #text/        
flag=complete
id=objId
version=1
# of chidren are 3
3 = #text/          
1 = reference/null
3 = #text/        
comment=objComment
flag=complete
id=objId
version=1

其中， 3 = TEXT_NODE和 1 = ELEMENT_NODE 。

Answer 3

您只計算 ELEMENT 節點類型。 如果您只對子元素感興趣，您可以更改代碼以包含以下檢查

 if (aNode.getNodeType() == Node.ELEMENT_NODE) 
{
...
}