请教几个TXMLDocument问题，请指教

[lshfong]

<?xml version="1.0" encoding="utf-8"?>
<ROOT>
<user>
<record>
<key>1</key>
<username>pangmen</username>
<sex>男</sex>
<tel>139999999</tel>
<address>北京</address>
</record>
<record>
<key>2</key>
<username>pangmen</username>
<sex>男</sex>
<tel>139999999</tel>
<address>北京</address>
</record>

<record>
<key>3</key>
<username>pangmen</username>
<sex>男</sex>
<tel>139999999</tel>
<address>北京</address>
</record>
</user>
<Role>
<row>
<key>1</key>
<pid>1,2,3</pid>
</row>
<row>
<key>2</key>
<pid>2,3</pid>
</row>
<row>
<key>3</key>
<pid>1</pid>
</row>
</Role>

</ROOT>

=========================================
1、怎么取指定节点下的子节点数据，比如我只遍历取Role节点下的数据，而不需要从根开始遍历
2、不遍历的情况下怎么查询子节点，在c#可以用XmlObj.SelectSingleNode实现，bcb中有对应的函数吗

[zzbinfo]

可以用这两个,具体你自己琢磨一下,今天没有时间帮着写代码了

      NodeList = NewDataSetNode->GetChildNodes();
     _di_IXMLNode  Node = NodeList->FindNode(WideString("PackgType"))

[lshfong]

支持xpath语法吗

[lshfong]

for (int n = 0; n < nodes->Count; n++) {
node = nodes->Nodes[n]->ChildNodes->First();
for (int i = 0; i < nodes->Nodes[n]->ChildNodes->Count; i++) {

Memo1->Lines->Append(nodes->Nodes[n]->ChildNodes->Nodes[i]->GetText));
}
}
用遍历的方法，当数据多或层很深时太耗资源了

[lshfong]

在Delphi区中找到了一个解决方案，哪个老师翻译成bcb的

[zzbinfo]

现在贴过来,我给你弄

[zzbinfo]

还有,你上面贴得xml都不全,你看看是不是贴少了

[lshfong]

http://topic.csdn.net/t/20030329/09/1591778.html

[lshfong]

怎么才能支持xpath语法呢

[lshfong]

noderef:=XMLDocument1.DocumentElement.DOMNode as IXMLDOMNodeRef;
root:=noderef.GetXMLDOMNode;
NodeList:=root.selectNodes( 'node1/node2/node3/node4 ');

象delphi这样

[ccrun.com]

[Quote=引用 13 楼 lshfong 的回复:]
noderef:=XMLDocument1.DocumentElement.DOMNode as IXMLDOMNodeRef;
root:=noderef.GetXMLDOMNode;
NodeList:=root.selectNodes( 'node1/node2/node3/node4 ');

象delphi这样
[/Quote]

XMLDocument1->LoadFromFile(L"D:\\ccrun\\123.xml");
_di_IXMLDOMNodeRef refNode = XMLDocument1->DocumentElement->DOMNode;
_di_IXMLDOMNode root = refNode->GetXMLDOMNode();

_di_IXMLDOMNodeList NodeList;
root->selectNodes("node1/node2/node3", NodeList);

[lshfong]

妖哥，能再详细一点吗，兄弟初学bcb

[lshfong]

[Quote=引用 6 楼 zzbinfo 的回复:]
还有,你上面贴得xml都不全,你看看是不是贴少了
[/Quote]
没有啊，我在c#和php里可以正确解析的，浏览器也可以解析

[zzbinfo]

不用遍历,这样弄,但是你上面得文件内容要改一下,由于有中文编码要用GB2312,改成<?xml version="1.0" encoding="GB2312" ?>

void __fastcall TForm1::btn1Click(TObject *Sender)
{
   _di_IXMLDocument XMLDoc = NewXMLDocument();
   XMLDoc->LoadFromFile("3.xml");
   try{
      XMLDoc->Active = true;
      }
    catch(...)
    {
     ShowMessage("文件格式出错");
     return ;
    }
     _di_IXMLNode   rootdataset   =  XMLDoc->DocumentElement;
     _di_IXMLNodeList NewDataSetNodeList = rootdataset->GetChildNodes();
     _di_IXMLNode NewDataSetNode =  NewDataSetNodeList->FindNode("Role");
     _di_IXMLNodeList NodeList = NewDataSetNode->GetChildNodes();
     for(int i=0;i<NodeList->GetCount() ;i++)
     {
      _di_IXMLNode RowNode = NodeList->Get(i);
      _di_IXMLNodeList NodeList1 = RowNode->GetChildNodes();
      _di_IXMLNode Node = NodeList1->FindNode("key");
      ShowMessage(Node->GetNodeValue());
      Node = NodeList1->FindNode("pid");
      ShowMessage(Node->GetNodeValue());
     }

}

[zzbinfo]

必要得头文件

#include <msxmldom.hpp>
#include <XMLDoc.hpp>
#include <xmldom.hpp>
#include <XMLIntf.hpp>

[ccrun.com]

如果没有往窗体上拖过TXMLDocument控件，那么编译时还需要添加一个lib文件。

[ahaozi]

        #include <XMLDoc.hpp>
        #include <msxmldom.hpp>
        TXMLDocument *xmlDoc = new TXMLDocument(this);
	xmlDoc->LoadFromFile("a.xml");

	_di_IXMLDOMNodeRef refNode = xmlDoc->DocumentElement->DOMNode;
	_di_IXMLDOMNode root = refNode->GetXMLDOMNode();

	if (root!=NULL)
	{
		_di_IXMLDOMNode n;
		root->selectSingleNode("//Config/Base/Ckdp",n);
		if (n!=NULL)
		{
			ShowMessage(n->text);
		}
	}

	delete xmlDoc;