DbSet.Find() in Entity Framework

DbSet.Find method in Entity framework provides some cool and awesome features.

Unlike other it doesn’t except lambda expression whereas it accept only primary key(s) value.

var cat = dbContext.Categories.Find(1);

Above will retrieve the category where category id is 1. “categoryid” is defined as primary key in the underlying table. Incase composite primary key we should pass values in the same sequence of keys.

var data = dbContext.Products.Find(1, 10);

So far, behavior is simple but nothing special. Wait ….following features will make it incredible.

Find() will always not make a database trip to fetch the data rather it supports following principles.

     1.  Find the data in memory and returns it.

var cat = dbContext.Categories.Find(1);

cat.Description = "Modified";

//Modified data in the memory will be returned

cat = dbContext.Categories.Find(1);

     2.  Look at the newly added data in memory.

var cats = new Category

{

  CategoryID = 1001, CategoryName = "New Category",

  Description = "New Category Description"

};

dbContext.Categories.Add(cats);

// newly added data in the memory will be returned

cats = dbContext.Categories.Find(1001);

3   3.  It makes a database trip only when 1 and 2 fails to return any data.

Wonderful isn’t it. I really liked it very much.

LINQ and equivalent LAMDA expression

This discussion I will try to provide some common LINQ statement and the equivalent LAMDA expression. This is purely personal preference to choose LINQ or Lamda. I am not going to describe which one is the best way considering performance as parameter. What I noticed both emit the same IL code after compilation. I used Entity framework 5.0 and Northwind database for this sample.

Retrieve all records

LINQ

from c in dbContext.Categories

       select c;

LAMDA

dbContext.Categories.Select(c => c);

Specific fields

LINQ

from c in dbContext.Categories

       select new { c.CategoryName, c.CategoryID };

LAMDA

dbContext.Categories.Select(c => new { c.CategoryName, c.CategoryID })

Orderby asc

LINQ

from c in dbContext.Categories

       orderby c.CategoryName

       select c;

LAMDA

dbContext.Categories.OrderBy(c => c.CategoryName);

Orderby dsc

LINQ

from c in dbContext.Categories

       orderby c.CategoryName descending

       select c;

LAMDA

      dbContext.Categories.OrderByDescending(c => c.CategoryName);

Multiple orderby

LINQ

from c in dbContext.Categories

       orderby c.CategoryName , c.CategoryID

       select c;

LAMDA

      dbContext.Categories.OrderBy(c => c.CategoryName).ThenBy(c => c.CategoryID);

Orderby asc and dec

LINQ

from c in dbContext.Categories

       orderby c.CategoryName, c.CategoryID descending

       select c;

LAMDA

dbContext.Categories.OrderBy(c => c.CategoryName)
.ThenByDescending(c => c.CategoryID);

Filter

LINQ

from c in dbContext.Categories

       where c.CategoryID.Equals(1)

       select c;

LAMDA

dbContext.Categories.Where(c => c.CategoryID.Equals(1));

Filter with multiple criteria

LINQ

from c in dbContext.Categories

       where c.CategoryID.Equals(1) && c.CategoryName.StartsWith("A")

       select c;

LAMDA

dbContext.Categories.Where(c => c.CategoryID.Equals(1) &&
       c.CategoryName.StartsWith("A"));

Group By

LINQ

from c in dbContext.Categories

       group c by c.CategoryName into g

       select g;

LAMDA

dbContext.Categories.GroupBy(c => c.CategoryName)

Multiple group By

LINQ

from c in dbContext.Categories

       group c by new { c.CategoryID, c.CategoryName } into g

       select g;

LAMDA

dbContext.Categories.GroupBy(c => new { c.CategoryID, c.CategoryName })

Record Count

LINQ

(from c in dbContext.Categories

        where c.CategoryName.StartsWith("B")

       select c).Count();

LAMDA

dbContext.Categories.Count(c => c.CategoryName.StartsWith("B"));

Join

LINQ

from c in dbContext.Categories

       join p in dbContext.Products1 on c.CategoryID equals p.CategoryID

       select new

              {

                 CategoryName,

                 p.ProductID,

                 p.ProductName

              };

LAMDA

dbContext.Categories.Join(dbContext.Products1, c => c.CategoryID, P =>
       P.CategoryID,((c,p)=>new {c.CategoryName,p.ProductID,p.ProductName}));

Entity framework and data iteration

We can write following piece of code to retrieve all records using DBSet. If we look at the database pro-filer we could see “SELECT * FROM Category” statement issued in the database.

var dbContext = new NorthwindEntities()

foreach (Category cat in dbContext.Categories)

{

Console.WriteLine(cat.CategoryName);

}

Some time we may need to iterate the same set of result set multiple times in a logical unit like below.

private void LogicalUnit()

{

          var dbContext = new NorthwindEntities()

foreach (Category cat in dbContext.Categories)

{

Console.WriteLine(cat.CategoryName);

}

foreach (Category cat in dbContext.Categories)

{

Console.WriteLine(cat.CategoryName);

}

}

Apparently everything looks fine but I wondered when I looked into the pro-filer and saw above SQL statement issued twice.

To avoid this, we can perform LINQ operation such as ToList() to copy the result set into a list and iterate the list whenever required.

var categories = dbContext.Categories.ToList();

foreach (Category cat in categories)

{

Console.WriteLine(cat.CategoryName);

}

foreach (Category cat in categories)

{

Console.WriteLine(cat.CategoryName);

}

This will hit the database only once to execute the query. We can refer the local result set for further reference. This way we can bring down the repeated database hit.