4.6. Working with Data
The entities in the model definition contain no data. The easiest way to to load data is to call the Load
method. For example,
private void LoadCustomersData()
{
dbContext.CUSTOMERS.Load();
var customers = dbContext.CUSTOMERS.Local;
bindingSource.DataSource = customers.ToBindingList();
}
private void CustomerForm_Load(object sender, EventArgs e)
{
LoadCustomersData();
dataGridView.DataSource = bindingSource;
dataGridView.Columns["CUSTOMER_ID"].Visible = false;
}
However, this approach has a few drawbacks:
The
Load
method loads all data from theCUSTOMER
table to memory at onceAlthough lazy properties (
INVOICES
) are not loaded immediately, but only once they are accessed, they will be loaded anyway when the records are shown in the grid and it will happen each time a group of records is shownRecord ordering is not defined
To get around these drawbacks, we will use a feature of the LINQ (Language Integrated Query) technology, LINQ to Entities. LINQ to Entities offers a simple and intuitive approach to getting data using C# statements that are syntactically similar to SQL query statements. You can read about the LINQ syntax in LINQ to Entities.
4.6.1. LINQ Extension Methods
The LINQ extension methods can return two objects: IEnumerable
and IQueryable
. The IQueryable
interface is inherited from IEnumerable
so, theoretically, an IQueryable
object is also an IEnumerable
. In reality, they are distinctly different.
The IEnumerable
interface is in the System.Collections
namespace. An IEnumerable
object is a collection of data in memory that can be addressed only in a forward direction. During the query execution, IEnumerable
loads all data. Filtering, if required, is done on the client side.
The IQueryable
interface is in the System.Linq
namespace. It provides remote access to the database and movement through the data can be bi-directional. During the process of creating a query that returns an IQueryable
object, the query is optimized to minimise memory usage and network bandwidth.
The Local
property returns the IEnumerable interface, through which we can create LINQ queries.
private void LoadCustomersData()
{
var dbContext = AppVariables.getDbContext();
dbContext.CUSTOMERS.Load();
var customers =
from customer in dbContext.CUSTOMERS.Local
orderby customer.NAME
select new customer;
bindingSource.DataSource = customers.ToBindingList();
}
However, as this query will be executed on the data in memory, it is really useful only for small tables that do not need to be filtered beforehand.
For a LINQ query to be converted into SQL and executed on the server, we need to access the dbContext.CUSTOMERS
directly instead of accessing the dbContext.CUSTOMERS.Local
property in the LINQ query. The prior call to dbContext.CUSTOMERS.Load();
to load the collection to memory is not required.
IQueryable and BindingList
IQueryable
objects present a small problem: they cannot return BindingList. BindingList
is a base class for creating a two-way data-binding mechanism. We can use the IQueryable
interface to get a regular list by calling ToList
but, this way, we lose handy features such as sorting in the grid and several more. The deficiency was fixed in .NET Framework 5 by creating a special extension. To do the same thing in FW4, we will create our own solution.
public static class DbExtensions
{
// Internal class for map generator values to it
private class IdResult
{
public int Id { get; set; }
}
// Cast IQueryable to BindingList
public static BindingList<T> ToBindingList<T>
(this IQueryable<T> source) where T : class
{
return (new ObservableCollection<T>(source)).ToBindingList();
}
// Get the next value of the sequence
public static int NextValueFor(this DbModel dbContext, string genName)
{
string sql = String.Format(
"SELECT NEXT VALUE FOR {0} AS Id FROM RDB$DATABASE", genName);
return dbContext.Database.SqlQuery<IdResult>(sql).First().Id;
}
// Disconnect all objects from the DbSet collection from the context
// Useful for updating the cache
public static void DetachAll<T>(this DbModel dbContext, DbSet<T> dbSet)
where T : class
{
foreach (var obj in dbSet.Local.ToList())
{
dbContext.Entry(obj).State = EntityState.Detached;
}
}
// Update all changed objects in the collection
public static void Refresh(this DbModel dbContext, RefreshMode mode,
IEnumerable collection)
{
var objectContext = ((IObjectContextAdapter)dbContext).ObjectContext;
objectContext.Refresh(mode, collection);
}
// Update the object
public static void Refresh(this DbModel dbContext, RefreshMode mode,
object entity)
{
var objectContext = ((IObjectContextAdapter)dbContext).ObjectContext;
objectContext.Refresh(mode, entity);
}
}
Other Extensions
There are several more extensions in the iQueryable
interface:
NextValueFor
is used to get the next value from the generator.
dbContext.Database.SqlQuery
allows SQL queries to be executed directly and their results to be displayed on some entity (projection).
DetachAll
is used to detach all objects of the DBSet collection from the context. It is necessary to update the internal cache, because all retrieved data are cached and are not retrieved from the database again. However, that is not always useful because it makes it more difficult to get the latest version of records that were modified in another context.
In web applications, a context usually exists for a very short period. A new context has an empty cache. |
Refresh
is used to update the properties of an entity object. It is useful for updating the properties of an object after it has been edited or added.
Code for Loading the Data
Our code for loading data will look like this:
private void LoadCustomersData()
{
var dbContext = AppVariables.getDbContext();
// disconnect all loaded objects
// this is necessary to update the internal cache
// for the second and subsequent calls of this method
dbContext.DetachAll(dbContext.CUSTOMERS);
var customers =
from customer in dbContext.CUSTOMERS
orderby customer.NAME
select customer;
bindingSource.DataSource = customers.ToBindingList();
}
private void CustomerForm_Load(object sender, EventArgs e)
{
LoadCustomersData();
dataGridView.DataSource = bindingSource;
dataGridView.Columns["INVOICES"].Visible = false;
dataGridView.Columns["CUSTOMER_ID"].Visible = false;
dataGridView.Columns["NAME"].HeaderText = "Name";
dataGridView.Columns["ADDRESS"].HeaderText = "Address";
dataGridView.Columns["ZIPCODE"].HeaderText = "ZipCode";
dataGridView.Columns["PHONE"].HeaderText = "Phone";
}
Adding a Customer
This is the code of the event handler for clicking the Add button:
private void btnAdd_Click(object sender, EventArgs e) {
var dbContext = AppVariables.getDbContext();
// creating a new entity instance
var customer = (CUSTOMER)bindingSource.AddNew();
// create an editing form
using (CustomerEditorForm editor = new CustomerEditorForm()) {
editor.Text = "Add customer";
editor.Customer = customer;
// Form Close Handler
editor.FormClosing += delegate (object fSender,
FormClosingEventArgs fe) {
if (editor.DialogResult == DialogResult.OK) {
try {
// get next sequence value
// and assign it
customer.CUSTOMER_ID = dbContext.NextValueFor("GEN_CUSTOMER_ID");
// add a new customer
dbContext.CUSTOMERS.Add(customer);
// trying to save the changes
dbContext.SaveChanges();
// and update the current record
dbContext.Refresh(RefreshMode.StoreWins, customer);
}
catch (Exception ex) {
// display error
MessageBox.Show(ex.Message, "Error");
// Do not close the form to correct the error
fe.Cancel = true;
}
}
else
bindingSource.CancelEdit();
};
// show the modal form
editor.ShowDialog(this);
}
}
While adding the new record, we used the generator to get the value of the next identifier. We could have done it without applying the value of the identifier, leaving the BEFORE INSERT
trigger to fetch the next value of the generator and apply it. However, that would leave us unable to update the added record.
Editing a Customer
The code of the event handler for clicking the Edit button is as follows:
private void btnEdit_Click(object sender, EventArgs e) {
var dbContext = AppVariables.getDbContext();
// get instance
var customer = (CUSTOMER)bindingSource.Current;
// create an editing form
using (CustomerEditorForm editor = new CustomerEditorForm()) {
editor.Text = "Edit customer";
editor.Customer = customer;
// Form Close Handler
editor.FormClosing += delegate (object fSender, FormClosingEventArgs fe) {
if (editor.DialogResult == DialogResult.OK) {
try {
// trying to save the changes
dbContext.SaveChanges();
dbContext.Refresh(RefreshMode.StoreWins, customer);
// update all related controls
bindingSource.ResetCurrentItem();
}
catch (Exception ex) {
// display error
MessageBox.Show(ex.Message, "Error");
// Do not close the form to correct the error
fe.Cancel = true;
}
}
else
bindingSource.CancelEdit();
};
// show the modal form
editor.ShowDialog(this);
}
}
The form for editing the customer looks like this:
Figure 25. Customer edit form
The code for binding to data is very simple.
public CUSTOMER Customer { get; set; }
private void CustomerEditorForm_Load(object sender, EventArgs e)
{
edtName.DataBindings.Add("Text", this.Customer, "NAME");
edtAddress.DataBindings.Add("Text", this.Customer, "ADDRESS");
edtZipCode.DataBindings.Add("Text", this.Customer, "ZIPCODE");
edtPhone.DataBindings.Add("Text", this.Customer, "PHONE");
}
Deleting a Customer
The code of the event handler for clicking the Delete button is as follows:
private void btnDelete_Click(object sender, EventArgs e) {
var dbContext = AppVariables.getDbContext();
var result = MessageBox.Show("Are you sure you want to delete the customer?",
"Confirmation",
MessageBoxButtons.YesNo,
MessageBoxIcon.Question);
if (result == DialogResult.Yes) {
// get the entity
var customer = (CUSTOMER)bindingSource.Current;
try {
dbContext.CUSTOMERS.Remove(customer);
// trying to save the changes
dbContext.SaveChanges();
// remove from the linked list
bindingSource.RemoveCurrent();
}
catch (Exception ex) {
// display error
MessageBox.Show(ex.Message, "Error");
}
}
}