Adds support for namespaces and iteration through the collection of attribute nodes.

System_CAPS_ICON_note.jpg Note

You can use books.xml to run this sample code.

var xmlDoc = new ActiveXObject("Msxml2.DOMDocument.6.0");  
var nodeBook, namedNodeMap;  
xmlDoc.async = false;  
xmlDoc.setProperty("SelectionLanguage", "XPath");  
if (xmlDoc.parseError.errorCode != 0) {  
   var myErr = xmlDoc.parseError;  
   WScript.Echo("You have error " + myErr.reason);  
} else {  
   nodeBook = xmlDoc.selectSingleNode("//book");  
   namedNodeMap = nodeBook.attributes;  
   WScript.Echo (namedNodeMap.length);  



The following C/C++ example creates and appends a new attribute to the root document element.

#include “msxml6.h”  
inline void TESTHR( HRESULT _hr )   
   { if FAILED(_hr) throw(_hr); }  
void XMLDOMNamedNodeMap()  
    try {  
        IXMLDOMDocumentPtr docPtr;  
        IXMLDOMNodePtr DOMNodePtr;  
        IXMLDOMNamedNodeMapPtr DOMNamedNodeMapPtr;  
        VARIANT vtTemp;  
        vtTemp.iVal = MSXML2::NODE_ATTRIBUTE;  
        // load a document  
        _variant_t varXml("C:\\book.xml");  
        _variant_t varOut((bool)TRUE);  
        varOut = docPtr->load(varXml);  
        if ((bool)varOut == FALSE)  
        DOMNodePtr = docPtr->createNode(vtTemp, "Sci-Fi", "");  
        DOMNamedNodeMapPtr = docPtr->documentElement->attributes;  
        MessageBox(NULL, _bstr_t(docPtr->xml), _T("New Document"), MB_OK);  
    } catch(...)  
        MessageBox(NULL, _T("Exception occurred"), _T("Error"), MB_OK);  

IXMLDOMNamedNodeMap is a node collection that allows access by name as well as by index. This collection is typically used for attributes.

The World Wide Web Consortium (W3C) Document Object Model (DOM) definition does not require NamedNodeMap objects to be maintained in any particular order. Objects contained in a NamedNodeMap can also be accessed by an ordinal index, but this is to allow convenient enumeration and does not imply that the DOM specifies an order for these nodes. DOM implementations are not required to preserve the node order.

The Microsoft® implementation preserves the attributes in the order in which they appear in the source. Additional attributes are added to the end of the list. A namespaceattribute, xmlns, is inserted as the first in the list if the tag name requires the namespace, or immediately before the attribute that uses an undeclared namespace.

Like the node list, however, a named node map collection is live; that is, the addition and removal of nodes, and changes within nodes, are immediately reflected in the collection. This means that two successive requests for items using the same index can return two different items, depending on changes to the collection. This also means that changes to the node objects are immediately available in the nodes obtained from the list.

When a default value is specified in the schema for an attribute and that attribute is removed, it is automatically regenerated with the default value, as specified in the W3C standard. The attribute reappears immediately because the collection is live, with a corresponding change to the count of items in the collection. When removing an attribute with such a default setting, you must again retrieve the number of items in the collection to obtain an accurate count. You should also reset any index into the collection.


msxml3.dll, msxml2.lib (MSXML 3.0)

msxml6.dll, msxml6.lib (MSXML 6.0)

Header and IDL files:

msxml2.h, msxml2.idl, msxml6.h, msxml6.idl


Implemented in: MSXML 3.0, MSXML 6.0

IXMLDOMNamedNodeMap Members