A data source is represented by the NDataSource class, which derives from NAttribute. A data source aggregates a NDataTable and exposes methods and properties identical to the properties and methods of NDataTable (see Data Tables for a complete discussion of the API).

The purpose of the NDataSource is thus to aggregate a data table and make it possible to incorporate data tables inside a DOM hierarchy. An instance of the NDataSource class is thus often exposed as property of elements that need to have data binding abilities. The aggregated data table is accessible from the DataTable readonly property. A data source is associated with a data table at data source creation time:

// create a data table with two fields
NMemoryDataTable dataTable = new NMemoryDataTable(
    new NFieldInfo("Name", typeof(string)),
    new NFieldInfo("Birthday", typeof(DateTime)));
           
// fill the data table
dataTable.AddRow("George", new DateTime(1976, 12, 19));
dataTable.AddRow("Andrew", new DateTime(1982, 9, 27));
// create a data source
NDataSource dataSource = new NDataSource(dataTable);

When a data source is deep cloned it copies the aggregated data table by reference. This means that deep cloning a data source produces two data sources that simultaneously reference the same data table:

NMemoryDataTable dataTable = new NMemoryDataTable(
    new NFieldInfo("Name", typeof(string)),
    new NFieldInfo("Birthday", typeof(DateTime)));
NDataSource dataSource1 = new NDataSource(dataTable);
NDataSource dataSource2 = (NDataSource)dataSource1.DeepClone();
Debug.Assert(Object.ReferenceEquals(dataSource1.DataTable, dataSource2.DataTable));

A data table is designed to notify parties interested in changes that occur in the data table, by subscribing for changes via the INDataTableObserver interface. The NDataSource class implements this interface and exposes the following events that let you monitor for changes in the underlying data table: