Edit

Share via

Edit data in datasets using the .NET Framework

  • Article

Note

Datasets and related classes are legacy .NET Framework technologies from the early 2000s that enable applications to work with data in memory while the applications are disconnected from the database. The technologies are especially useful for applications that enable users to modify data and persist the changes back to the database. Although datasets have proven to be a very successful technology, we recommend that new .NET applications use Entity Framework Core. Entity Framework provides a more natural way to work with tabular data as object models, and it has a simpler programming interface.

You edit data in data tables much like you edit the data in a table in any database. The process can include inserting, updating, and deleting records in the table. In a data-bound form, you can specify which fields are user-editable. In those cases, the data-binding infrastructure handles all the change tracking so that the changes can be sent back to the database later. If you programmatically make edits to data, and you intend to send those changes back to the database, you must use the objects and methods that do the change tracking for you.

In addition to changing the actual data, you can also query a DataTable to return specific rows of data. For example, you might query for individual rows, specific versions of rows (original and proposed), rows that have changed, or rows that have errors.

To edit rows in a dataset

To edit an existing row in a DataTable, you need to locate the DataRow you want to edit, and then assign the updated values to the desired columns.

If you don't know the index of the row you want to edit, use the FindBy method to search by the primary key:

NorthwindDataSet.CustomersRow customersRow = 
    northwindDataSet1.Customers.FindByCustomerID("ALFKI");

customersRow.CompanyName = "Updated Company Name";
customersRow.City = "Seattle";

If you know the row index, you can access and edits rows as follows:

northwindDataSet1.Customers[4].CompanyName = "Updated Company Name";
northwindDataSet1.Customers[4].City = "Seattle";

To insert new rows into a dataset

Applications that use data-bound controls typically add new records through the Add New button on a BindingNavigator control.

To manually add new records to a dataset, create a new data row by calling the method on the DataTable. Then, add the row to the DataRow collection (Rows) of the DataTable:

NorthwindDataSet.CustomersRow newCustomersRow = 
    northwindDataSet1.Customers.NewCustomersRow();

newCustomersRow.CustomerID = "ALFKI";
newCustomersRow.CompanyName = "Alfreds Futterkiste";

northwindDataSet1.Customers.Rows.Add(newCustomersRow);

In order to retain the information that the dataset needs to send updates to the data source, use the Delete method to remove rows in a data table. For example, if your application uses a TableAdapter (or DataAdapter), the TableAdapter's Update method deletes rows in the database that have a RowState of Deleted.

If your application does not need to send updates back to a data source, it's possible to remove records by directly accessing the data row collection (Remove).

To delete records from a data table

  • Call the Delete method of a DataRow.

    This method doesn't physically remove the record. Instead, it marks the record for deletion.

    Note

    If you get the count property of a DataRowCollection, the resulting count includes records that have been marked for deletion. To get an accurate count of records that aren't marked for deletion, you can loop through the collection looking at the RowState property of each record. (Records marked for deletion have a RowState of Deleted.) Alternatively, you can create a data view of a dataset that filters based on row state and get the count property from there.

The following example shows how to call the Delete method to mark the first row in the Customers table as deleted:

northwindDataSet1.Customers.Rows[0].Delete();

Determine whether there are changed rows

When changes are made to records in a dataset, information about those changes is stored until you commit them. You commit the changes when you call the AcceptChanges method of a dataset or data table, or when you call the Update method of a TableAdapter or data adapter.

Changes are tracked two ways in each data row:

  • Each data row contains information related to its RowState (for example, Added, Modified, Deleted, or Unchanged).

  • Each changed data row contains multiple versions of that row (DataRowVersion), the original version (before changes) and the current version (after changes). During the period when a change is pending (the time when you can respond to the RowChanging event), a third version — the proposed version— is available as well.

The HasChanges method of a dataset returns true if changes have been made in the dataset. After determining that changed rows exist, you can call the GetChanges method of a DataSet or DataTable to return a set of changed rows.

To determine whether changes have been made to any rows

  • Call the HasChanges method of a dataset to check for changed rows.

The following example shows how to check the return value from the HasChanges method to detect whether there are any changed rows in a dataset named NorthwindDataset1:

if (northwindDataSet1.HasChanges()) 
{
    // Changed rows were detected, add appropriate code.
}
else
{
    // No changed rows were detected, add appropriate code.
}

Determine the type of changes

You can also check to see what type of changes were made in a dataset by passing a value from the DataRowState enumeration to the HasChanges method.

To determine what type of changes have been made to a row

The following example shows how to check a dataset named NorthwindDataset1 to determine whether any new rows have been added to it:

if (northwindDataSet1.HasChanges(DataRowState.Added)) 
{
    // New rows have been added to the dataset, add appropriate code.
}
else
{
    // No new rows have been added to the dataset, add appropriate code.
}

To locate rows that have errors

When working with individual columns and rows of data, you might encounter errors. You can check the HasErrors property to determine whether errors exist in a DataSet, DataTable, or DataRow.

  1. Check the HasErrors property to see whether there are any errors in the dataset.

  2. If the HasErrors property is true, iterate through the collections of tables, and then the through the rows, to find the row with the error.

private void FindErrors() 
{
    if (dataSet1.HasErrors)
    {
        foreach (DataTable table in dataSet1.Tables)
        {
            if (table.HasErrors)
            {
                foreach (DataRow row in table.Rows)
                {
                    if (row.HasErrors)
                    {
                        // Process error here.
                    }
                }
            }
        }
    }
}

Related content