DevCity.NET :: Searching an XML Document Using a DataSet and DataView

Sometimes you need the basic features of a database but don't want the hassle, and possibly cost, of creating one for a small application. With the .NET DataSet object and a simple XML document we can emulate the basic features of a database. In this example, we use an XML document which stores our product information. We want to look up a product by SKU and return the price and description to a Web page.

We have three products in our list. The following is a basic XML document, productlist.xml, containing our data:

<?xml version="1.0" encoding="utf-8" ?>
<ProductList>
<Products>
    <SKU>1</SKU>
    <Price>100.00</Price>
    <Description>Widget #1</Description>
</Products>
<Products>
    <SKU>2</SKU>
    <Price>10.00</Price>
    <Description>Widget #2</Description>
</Products>
<Products>
    <SKU>3</SKU>
    <Price>30.00</Price>
    <Description>Widget #3</Description>
</Products>
</ProductList>

If you think of this document in terms of a database, the data within the ProductList tags are our database. The data within the Products tags are a table in the database. SKU, Price, and Descriptions are columns in the Products table. Just like a database, we want to return a row of values, given a column value.

First, we need to read our XML document into something that .NET can use. This is very easy to do with the DataSet object's ReadXml method. By using a DataSet, we can treat our data like a database. Also, compared to using the .NET XML objects, the DataSet results in less code and complexity.

All we need to do is create a new DataSet and tell it to read our XML document into it:

DataSet dsProducts = new DataSet();
dsProducts.ReadXml(Server.MapPath("productlist.xml"));

The ReadXml method is passed the file name of the XML file to read. This example assumes that the file is named productlist.xml and is located in the same directory as the page using it.

Now that we have our data in a DataSet, we need to be able to query it for an SKU and return the Price and Description for the corresponding product. To do this, we will use a DataView object:

DataView dvProducts = new DataView(dsProducts.Tables["Products"]);
dvProducts.Sort = "SKU";

The DataView is created by passing the constructor a DataTable, "Products", in our DataSet. After we create the DataView, we need to sort it. Since we are using the SKU as the lookup value, we need to sort on it. This is similar to an index or primary key in a database. If we don't sort the DataView, we will get a runtime error when we try to use it's Find method.

With our DataView sorted, use the Find method to return only the row(s) which match the SKU we give it. In this example, we want to return the Price and Description for a product with a SKU equal to 1:

int rowIndex = dvProducts.Find("1");
string Price, Description;

if (rowIndex == -1)
{
    // The SKU was not found in our data
    Response.Write("SKU not found");
}
else
{
    Price = dvProducts[rowIndex]["Price"].ToString();
    Response.Write("Price: " + Price + "<br>";
    // Price: 100.00

    Description = dvProducts[rowIndex]["Description"].ToString();
    Response.Write("Description: " + Description);
    // Description: Widget #1
}

The Find method of a DataView returns and int which is the index of the row in the DataView containing the sort key value passed to it. If there is no match, it returns -1.

That's all there is to it. For simple uses, this method of storing and retrieving information may be much easier than using a database, however, there are a few things you should keep in mind. First, our simple XML document stores all data as text strings. This isn't too much of a problem since we can use the .NET data types to parse our string data to other types like integer and datetime, but remember when you pass the key value to the Find method that it needs to be a string. Second, if we only want one "row" of data returned for a lookup key, we need to ensure that we don't duplicate the key in our XML document--in a database we could easily enforce this by setting the column to be an identity column.

If you are using this method in a busy ASP.NET application where you have a relatively small amount of data and your data doesn't change too often, you may want to store the DataSet object in cache so the XML document doesn't need to be read and parsed for each request. Use a CacheDependency on the XML file to ensure that the cache is expired when the document is updated.

Download the following for complete code in a console application:

using System;
using System.Data;

namespace productsxml
{
    /// <summary>
    /// Summary description for Class1.
    /// </summary>
    class main
    {
        /// <summary>
        /// The main entry point for the application.
        /// </summary>
        [STAThread]
        static void Main(string[] args)
        {
            //
            // TODO: Add code to start application here
            //

            DataSet dsProducts = new DataSet();
            dsProducts.ReadXml("productlist.xml");

            DataView dvProducts = new
            DataView(dsProducts.Tables["Products"]);
            dvProducts.Sort = "SKU";

            int rowIndex = dvProducts.Find("1");
            string Price, Description;

            if (rowIndex == -1)
            {
                // The SKU was not found in our data
                Console.WriteLine("Product Not found");
            }
            else
            {
                Price = dvProducts[rowIndex]["Price"].ToString();
                Console.WriteLine("Price: " + Price);
                // Price: 100.00

                Description = dvProducts[rowIndex]
                ["Description"].ToString();
                Console.WriteLine("Description: " + Description);
                // Description: Widget #1
            }
        }
    }
}