Question

I have following LINQ to SQL query expression

from msg in TblUserMessages 
join user in Aspnet_Users on msg.FromUserID equals user.UserId
select new {
       msg.FromUserID, 
       msg.ToUserID, 
       msg.MessageLocationID, 
       msg.MessageID, 
       user.UserName
       }

And following LINQ method expression:

TblUserMessages
.Join (
  Aspnet_Users, 
  msg => msg.FromUserID, 
  user => user.UserId, 
  (msg, user) => 
     new  
     {
        FromUserID = msg.FromUserID, 
        ToUserID = msg.ToUserID, 
        MessageLocationID = msg.MessageLocationID, 
        MessageID = msg.MessageID, 
        UserName = user.UserName
     }
)

Both of these return same result set. Which is e.g. :

82522f05-2650-466a-a430-72e6c9fb68b7
6b2a174a-8141-43d2-b3ad-5b199bcbfcae
1
1
waheed

Which one is better to use. The FIRST one or the SECOND one.

Thanks

Answer 1

They are equivalent. They don’t just return the same result set – they compile to the same code.

Use query expressions or dot notation for individual cases depending on readability. For joins, I find the dot notation quite cumbersome – but I use it for situations where I only have one or two clauses (usually where/select). Even with two clauses (usually where and select) I find dot notation nice if you then need to use it anyway. For example, I like:

var query = people.Where(person => person.Age > 18)
                  .Select(person => person.Name)
                  .Skip(100)
                  .Take(10);

over

var query = (from person in people
             where person.Age > 18
             select person.Name)
            .Skip(100)
            .Take(10);

For more complicated queries (e.g. joins) I’d probably just separate the two:

var baseQuery = from person in people
                where person.Age > 18
                join company on person.CompanyId equals company.CompanyId
                select new { person.Name, company.Name };

var fullQuery = baseQuery.Skip(100)
                         .Take(10);

I just find this separation makes it easier to read.

I believe it’s really useful for developers to understand at least the basics of what query expressions do – the fact that they’re basically translations into dot notation, and that the language itself doesn’t know anything about LINQ to Objects, LINQ to SQL etc; it’s just a case of following an appropriate pattern. It’s a great bit of design, which means query expressions only affect one little bit of the language specification.