Getting Started
Visual Basic.NET provides integrated tools for using XML, schemas, XSL, and other XML-related tools, but there's no need to wait for VB.NET in order to get started. Microsoft XML Core Services (MSXML) version 4.0 gives you the tools you need to load and save XML documents from your Visual Basic 6 application.
Download the most recent version of MSXML at msdn.microsoft.com/xml/default.asp, and install it on your computer. To use MSXML in a Visual Basic 6 program, select the Project menu's References command. Select Microsoft XML, v 4.0 (or whatever the version du jour is), and click OK. Now you're ready to add XML objects to your program.
DOMDocument Class
The DOM (document object model) describes an XML file's hierarchical nature using a corresponding set of programming objects. DOMDocument is the MSXML class that represents an XML document's DOM structure.
The DOMDocument class provides only a few really useful properties and methods. The load method makes the object load an XML file. The loadXML method makes it load XML data from a string. For example, the following code loads a small XML file into the document named xml_document.
Dim xml_document As New DOMDocument xml_document.loadXML _ "<Person>" & vbCrLf & _ " <FirstName>Rod</FirstName>" & vbCrLf & _ " <LastName>Stephens</LastName>" & vbCrLf & _ "</Person>"
The DOMDocument's xml property returns the document's XML representation. You can display this value to see what the document looks like. You could also save it into a file, but there's really no need to do that because the DOMDocument object's save method does it automatically.
The object's documentElement property is a node that represents the document's topmost data node. All the other nodes lie within that one. Usually, when you manipulate an XML document's nodes, you will start at this node.
DOMDocument provides several methods for creating new nodes. The createElement method makes a new element node for the document. This is the kind of node you usually want to create. Other node creation methods include createAttribute, createProcessingInstruction, and createTextNode, although I won't say anything about those methods in this article.
IXMLDOMNodeClass
The class IXMLDOMNode represents a node. This class provides a number of properties and methods you can use to search and manipulate the XML document.
The selectSingleNode method searches the node's descendants for a specific node. The language you use to specify the node you want is called XPath, and it's too involved to cover here. Two particularly simple and useful XPath specifications search the node's immediate children and all of its descendants.
If you pass selectSingleNode a child node's name, the method searches the node's children for an exact match. If you pass selectSingleNode the string ".//" followed by a node's name, the method searches the node's descendants for the name.
' Search for a child node named "LastName." Set last_name_node = address_node.selectSingleNode("LastName") ' Search for any descendant named "LastName." Set last_name_node = address_node.selectSingleNode(".//LastName")
The following list shows a few of the IXMLDOMNode object's more useful properties:
attributes. A collection of the node's attributes.
nodeName. Gives the node's tag name.
nodeTypeString. Tells you the node's type.
ownerDocument. Returns the DOMDocument object containing the node.
text. Gives the text contained in the node. If the node contains other nodes, this is the combined text of them all.
xml. Gives the node's XML content, as in "<FirstName>Rod</FirstName>".
The childNodes collection contains references to the node's children. To add a new node to this one, you must first create it using one of the DOMDocument object's node-creation methods. Then you add the new node to the parent's childNodes collection. The following code shows a CreateNode subroutine that creates a new node and adds it to a parent node's children using the parent's appendChild method.
' Add a new node to the indicated parent node. Private Sub CreateNode(ByVal indent As Integer, _ ByVal parent As IXMLDOMNode, ByVal node_name As String, _ ByVal node_value As String) Dim new_node As IXMLDOMNode ' Create the new node. Set new_node = parent.ownerDocument.createElement(node_name) ' Set the node's text value. new_node.Text = node_value ' Add the node to the parent. parent.appendChild new_node End Sub
SaveValues Program
With these tools, you can build a simple program that uses XML. Program SaveValues, shown in Figure 1, loads and saves values in an XML file. When it starts, the program loads its text boxes using the values in the file Values.xml. When it stops, the program saves its current values into that file.
Figure 1 Program SaveValues loads and saves values in an XML file.
The following code shows a typical Values.xml file.
<Values> <FirstName>Rod</FirstName> <LastName>Stephens</LastName> <Street>1234 Programmer Place</Street> <City>Bugsville</City> <State>CO</State> <Zip>80276</Zip> </Values>
Listing 1 shows how program SaveValues works. When the program's form loads, the Form_Load event handler calls subroutine LoadValues.
LoadValues creates a DOMDocument object named xml_document, loads the XML file, and uses its selectSingleNode method to find the node named Values. It then uses the GetNodeValue helper function to fetch the values it needs from the descendants of the Values node.
GetNodeValue uses the Values node's selectSingleNode method to find the target node. If the node isn't there, the function returns a default value. If it finds the node, GetNodeValue returns the node's text value. For the data nodes in Values.xml, the text value is simply the text contained in the nodes.
When the program's form unloads, the Form_Unload event handler calls subroutine SaveValues. That routine creates a new DOMDocument object. It creates a new node named Values, and uses the document's appendChild method to add the new node to the document.
SaveValues then calls the CreateNode helper subroutine to save each of the values in the program's text boxes. SaveValues passes a reference to the Values node to CreateNode so that the routine can place the new node in the document's Values section.
After it has created all of the new nodes, SaveValues uses the DOMDocument's save method to save the new XML file.
Note that the new file overwrites the old one. There is no way to change just part of an XML file using the DOMDocument object. You can load the file, make a few small changes, and save the file, but the file is completely overwritten. That is usually not a problem unless some other program may have changed the file in the meantime. In that case, any changes the other program made will be lost when you overwrite the file.
The last piece of the program is the CreateNode subroutine. CreateNode adds a new node with a given value to a parent node. First, the routine uses the parent node's ownerDocument method to get a reference to the DOMDocument object. It uses that object's createElement method to create a new node.
CreateNode then sets the new node's Text property to its desired value, and adds the node to the parent node's children.
Listing 1. Program SaveValues Uses this Code to Load and Save Values in an XML File.
Option Explicit Private m_AppPath As String Private Sub Form_Load() ' Get the application's startup path. m_AppPath = App.Path If Right$(m_AppPath, 1) <> "\" Then m_AppPath = m_AppPath & "\" ' Load the values. LoadValues End Sub Private Sub Form_Unload(Cancel As Integer) ' Save the current values. SaveValues End Sub ' Load saved values from XML. Private Sub LoadValues() Dim xml_document As DOMDocument Dim values_node As IXMLDOMNode ' Load the document. Set xml_document = New DOMDocument xml_document.Load m_AppPath & "Values.xml" ' If the file doesn't exist, then ' xml_document.documentElement is Nothing. If xml_document.documentElement Is Nothing Then ' The file doesn't exist. Do nothing. Exit Sub End If ' Find the Values section. Set values_node = xml_document.selectSingleNode("Values") ' Read the saved values. txtFirstName.Text = GetNodeValue(values_node, "FirstName", "???") txtLastName.Text = GetNodeValue(values_node, "LastName", "???") txtStreet.Text = GetNodeValue(values_node, "Street", "???") txtCity.Text = GetNodeValue(values_node, "City", "???") txtState.Text = GetNodeValue(values_node, "State", "???") txtZip.Text = GetNodeValue(values_node, "Zip", "???") End Sub ' Return the node's value. Private Function GetNodeValue(ByVal start_at_node As IXMLDOMNode, _ ByVal node_name As String, _ Optional ByVal default_value As String = "") As String Dim value_node As IXMLDOMNode Set value_node = start_at_node.selectSingleNode(".//" & node_name) If value_node Is Nothing Then GetNodeValue = default_value Else GetNodeValue = value_node.Text End If End Function ' Save the current values. Private Sub SaveValues() Dim xml_document As DOMDocument Dim values_node As IXMLDOMNode ' Create the XML document. Set xml_document = New DOMDocument ' Create the Values section node. Set values_node = xml_document.createElement("Values") ' Add the Values section node to the document. xml_document.appendChild values_node ' Create nodes for the values inside the ' Values section node. CreateNode values_node, "FirstName", txtFirstName.Text CreateNode values_node, "LastName", txtLastName.Text CreateNode values_node, "Street", txtStreet.Text CreateNode values_node, "City", txtCity.Text CreateNode values_node, "State", txtState.Text CreateNode values_node, "Zip", txtZip.Text ' Save the XML document. xml_document.save m_AppPath & "Values.xml" End Sub ' Add a new node to the indicated parent node. Private Sub CreateNode(ByVal parent As IXMLDOMNode, _ ByVal node_name As String, ByVal node_value As String) Dim new_node As IXMLDOMNode ' Create the new node. Set new_node = parent.ownerDocument.createElement(node_name) ' Set the node's text value. new_node.Text = node_value ' Add the node to the parent. parent.appendChild new_node End Sub
Download the SaveValues source code here:
SaveValuesIndented Program
Although everyone makes a big deal about how easy XML files are to read, the tools that manipulate them typically ignore the whitespace that makes their structure obvious. The XML parser ignores the indentation and spacing you add to make the file easier to understand.
Unfortunately, the routines that write XML files tend to omit this whitespace, too. Program SaveValues actually creates an XML file that looks like the following with all the data run together on a single line.
<Values><FirstName>Rod</FirstName><LastName>Stephens</LastNa me><Street>1234 Programmer Place</Street><City>Bugsville</Ci ty><State>CO</State><Zip>80276</Zip></Values>
VB.NET includes text writer classes that can format XML documents. They aren't trivial to use, but they do work. MSXML doesn't include those text writers, so if you want to save an XML document with a nicely indented format, you need to add the formatting yourself.
In addition to named elements such as the ones shown here, an XML file can contain unnamed text nodes. In fact, the values inside the nodes in this example are already contained in text nodes. For example, if you pick the Values.xml file's structure apart carefully, you will find that the FirstName element contains a text node that holds the text Rod.
You can add your own text nodes between named elements. To give this file a nicely indented structure, you need to add text nodes containing spaces before each node to indent it, and new line characters after each node to make the next node start on the next line. This may seem difficult, but it's really not too hard.
Program SaveValuesIndented uses the code in Listing 2 to save its values. The SaveValues subroutine is almost the same as the previous version. After it creates the Values element, however, this routine adds a text node containing a carriage return and line feed. Putting this text node inside the Values node makes the resulting XML document contain a new line right after the <Values> tag.
SaveValues then calls CreateNode to create the new data nodes. It passes CreateNode a new parameter telling it how far to indent the new nodes.
CreateNode starts by adding a text node containing spaces to the parent node to indent the new node. It then adds the new node to the parent's children, much as the previous version did. It finishes by adding a text node containing a carriage return and line feed to the parent, so whatever follows the new node starts on the next line.
Listing 2. This Code Adds Text Nodes to Indent the Resulting XML Document.
' Save the current values. Private Sub SaveValues() Dim xml_document As DOMDocument Dim values_node As IXMLDOMNode ' Create the XML document. Set xml_document = New DOMDocument ' Create the Values section node. Set values_node = xml_document.createElement("Values") ' Add a new line. values_node.appendChild xml_document.createTextNode(vbCrLf) ' Add the Values section node to the document. xml_document.appendChild values_node ' Create nodes for the values inside the ' Values section node. CreateNode 4, values_node, "FirstName", txtFirstName.Text CreateNode 4, values_node, "LastName", txtLastName.Text CreateNode 4, values_node, "Street", txtStreet.Text CreateNode 4, values_node, "City", txtCity.Text CreateNode 4, values_node, "State", txtState.Text CreateNode 4, values_node, "Zip", txtZip.Text ' Save the XML document. xml_document.save m_AppPath & "Values.xml" End Sub ' Add a new node to the indicated parent node. Private Sub CreateNode(ByVal indent As Integer, _ ByVal parent As IXMLDOMNode, ByVal node_name As String, _ ByVal node_value As String) Dim new_node As IXMLDOMNode ' Indent. parent.appendChild _ parent.ownerDocument.createTextNode(Space$(indent)) ' Create the new node. Set new_node = parent.ownerDocument.createElement(node_name) ' Set the node's text value. new_node.Text = node_value ' Add the node to the parent. parent.appendChild new_node ' Add a new line. parent.appendChild parent.ownerDocument.createTextNode(vbCrLf) End Sub
Download the SaveValuesIndented source code here: