We Keep Coding

C# LINQ to SQL - one to many relationship

I have troubles creating this query in LINQ:
USE Northwind
GO
SELECT emp.FirstName, emp.LastName, tr.TerritoryDescription, reg.RegionDescription
FROM Employees emp
INNER JOIN EmployeeTerritories empt ON empt.EmployeeID = emp.EmployeeID
INNER JOIN Territories tr ON tr.TerritoryID = empt.TerritoryID
INNER JOIN Region reg ON reg.RegionID = tr.RegionID
This is my current creation:
var query = await context
.Employees
.Select(x => new
{
x.FirstName,
x.LastName,
TerritoryId = x.EmployeeTerritories. //cannot access properties
})
.ToListAsync();
But i can't easily access EmployeeTerritories properties, since it's not 1:1 relationship. I accept both clues and full solution to this problem.
Edit
So this is what i currently have:
var query = await context
.Employees
.Select(x => new
{
x.FirstName,
x.LastName,
TerritoryDescription = x.EmployeeTerritories
.Select(et => et.Territory.TerritoryDescription)
.ToList(),
RegionDesicription = x.EmployeeTerritories
.Select(et => et.Territory.Region.RegionDescription)
.ToList()
})
.ToListAsync();
Is there a way to optimize it? RegionDescription is now a list that contains one element, but i don't know how to do it the better way.

Try something like this (assuming you have corresponding relations):
var query = await context
.Employees
.Select(x => new
{
x.Employee.FirstName,
x.Employee.LastName,
TerritoryDescription = x.EmployeeTerritories
.Select(et => et.Territory.TerritoryDescription)
.ToList(),
})
.ToListAsync();
UPD
To flatten in your particular case you can use solution posted by #dhrumil shah(it is more generic one) or try something like that, if you have EmployeeTerritories set up in your context :
var query = await context
.EmployeeTerritories
.Select(et => new
{
et.Employee.FirstName,
et.Employee.LastName,
et.Territory.TerritoryDescription,
et.Territory.Region.RegionDescription
})
.ToListAsync();

(from emp in context.Employees
join empt in context.EmployeeTerritories
on emp.EmployeeID equals empt.EmployeeID
join tr in context.EmployeeTerritories
on empt.TerritoryID equals tr.EmployeeID
join reg in context.Region
on reg.RegionID equals tr.RegionID
select new {
emp.FirstName,
emp.LastName,
tr.TerritoryDescription,
reg.RegionDescription
}).ToList();

Converting SQL to LINQ with Multiple Tables and Group

There seem to be lots of questions about SQL to LINQ, but I can't seem to find examples with joined tables and grouping; specifically with a need to get data from multiple tables.
Take this simple SQL:
SELECT
s.showId, s.showName, v.venueName, Min(dateTime) startDate
FROM
shows s
INNER JOIN venues v ON s.venueId = v.venueId
INNER JOIN showDates d ON s.showId = d.showId
GROUP BY
s.showId
The best I can come up with is the following
var ungrouped = (
from s in db.Shows
join v in db.Venues on s.VenueId equals v.VenueId
join d in db.ShowDates on s.ShowId equals d.ShowId
select new { s, v, d }
).ToList();
var grouped = (
from s in ungrouped
group s by s.s.ShowId into grp
select new
{
showId = grp.Key,
name = (from g in grp select g.s.showName).FirstOrDefault(),
venue = (from g in grp select g.v.VenueName).FirstOrDefault(),
startDate = grp.Max(g => g.d.DateTime)
}
);
This works but it feels messy. I don't like:
It being split into two statements
Having to repeatedly write (from g in grp select ...).FirstOrDefault()
Bits like s.s.ShowId
How its vastly more lines of code than the SQL
This example is a simple one, it only gets worse when I have 5+ tables to join and 10+ columns to select.
Question: Is this the best way to do this, and I should just accept it; or is there a better way to write this query?

I am not sure if you are looking for something like this but it's a bit cleaner, it's not split in 2 statements and you might find it helpful. I couldn't use a dbcontext so I used lists to make sure the syntax is correct.
var res = Shows.Join(Venues,
show => show.VenueID,
venue => venue.VenueID,
(show, venue) => new { show, venue })
.Join(ShowDates,
val => val.show.ShowID,
showdate => showdate.ShowID,
(val, showDate) => new { val.show, val.venue, showDates = showDate })
.GroupBy(u => u.show.ShowID)
.Select(grp => new
{
showId = grp.Key,
name = grp.FirstOrDefault()?.show.showName,
venue = grp.FirstOrDefault()?.venue.VenueName,
startDate = grp.Max(g => g.showDates.DateTime)
});

we need to now realation beetwen them one to one or one to many , but not too far from this answer.
var GrouppedResult = Shows.Include(x=>x.Veneu).Include(x=>x.ShowDates)
.Where(x=>x.Veneu.Any()&&x.ShowDates.Any())
.GroupBy(x=>x.ShowId)
.Select(x=>///anything you want);
or
from show in Shows
join veneu in Veneu on veneu.VeneuId equals show.VeneuId
join showDates in ShowDates on showDates.ShowId=show.ShowID
group show by show.Id into grouppedShows
select new { ///what you want };

Linq method based left join with multiple criteria

How i can obtain the similar Linq query as this SQL query using Linq method based?
SELECT * FROM F_ARTICLE A
LEFT JOIN F_ARTFOURNISS AF ON AF.AR_Ref = A.AR_Ref AND AF.AF_Principal = 1
ORDER BY A.AR_Design DESC
OFFSET 500 ROWS FETCH NEXT 100 ROWS ONLY
I'm using method based due to System.Linq.Dynamic requirements.
The mapping is like this:
I started by this but i don't know how to limit to AF_Principal = 1:
var query = context.F_ARTICLE
.Join(
context.F_ARTFOURNISS,
a => a.AR_Ref,
t => t.AR_Ref,
(a, t) => new { a, t })
.Select(
z => new ArticleItem()
{
Article = z.a,
Tarif = z.t.FirstOrDefault()
})
.OrderBy($"{field} {direction}");
return query.Skip(startIndex).Take(count).ToList();

The code is from my head but the point is using DefaultIfEmpty for doing LEFT JOIN:
var query = from farticle in context.F_ARTICLE
join b in context.F_ARTFOURNISS
on new {farticle.AR_Ref, AF_Principal = 1} equals new {b.AR_Ref, b.AF_Principal} into gj
from subres in gj.DefaultIfEmpty()
select farticle;
return query.Skip(startIndex).Take(count).ToList();
Also you can limit by AF_Principal using Where clause (I think DBMS engine will optimize the query anyway)

see if this works
var query = from F1 in context.F_ARTICLE join F2 in context.F_ARTFOURNISS
on F1.AR_Ref equals F2.AR_Ref into newF
from F3 in newF.where(f=>f.AF_Principal == 1).DefaultIfEmpty()
select f;
return query.Skip(startIndex).Take(count).ToList();

I think i found the query:
var query = context.F_ARTICLE
.GroupJoin(
context.F_ARTFOURNISS,
a => a.AR_Ref,
t => t.AR_Ref,
(a, t) => new { a, t })
.Select(
z => new ArticleItem()
{
Article = z.a,
Tarif = z.t.Where(t=>t.AF_Principal == 1).FirstOrDefault()
})
.OrderBy($"{field} {direction}");
return query.Skip(startIndex).Take(count).ToList();

Can make left join in Entity Framework without dataset

I have this query
return _ctx.TestPackages.Where(s => s.Id == TestPackageId).
Join(_ctx.TestPackageReportDetails, s => s.Id, d => d.TestPackageId, (s, d) => new { reportDetail = d, testpack = s }).
Join(_ctx.TestPackageReports, p => p.reportDetail.TestPackageReportId, o => o.Id, (p, o) => new { combined = p, report = o })
.ToList()
As you can see my query makes join between 3 tables TestPackages TestPackageReportDetails and TestPackageReports. When I have more than one record in TestPackageReportDetails with same testpackageid, the result is repeated 3 times in the output. How can I avoid the repetition?
Should I make a left join between TestPackageReportDetails and TestPackages? If yes how can I do that?

If your intention is to Eager load the report details then you should be using .Include :
_ctx.TestPackages.Include(t=>t.TestPackageReportDetails.TestPackageReports).Where(s => s.Id == TestPackageId);

Since you're only selecting three fields (As shown before you edit your answer again and remove the select) then you can do this:
(from s in _ctx.TestPackages
join d in _ctx.TestPackageReportDetails,
on s.Id equals d.TestPackageId
join r in _ctx.TestPackageReports
on s.Id equals r.reportDetail.TestPackageReportId
where s.Id == TestPackageId
select new
{
s.Id,
s.packageNumber,
s.Size,
s.TestPackageOrder
}).Distinct().ToList().Select(m=> new ..) // continue your normal selection

How to perform Join between multiple tables in LINQ lambda

I am trying to perform a Join between multiple tables in LINQ. I have the following classes:
Product {Id, ProdName, ProdQty}
Category {Id, CatName}
ProductCategory{ProdId, CatId} //association table
And I use the following code (where product, category and productcategory are instances of the above classes):
var query = product.Join(productcategory, p => p.Id, pc => pc.ProdID, (p, pc) => new {product = p, productcategory = pc})
.Join(category, ppc => ppc.productcategory.CatId, c => c.Id, (ppc, c) => new { productproductcategory = ppc, category = c});
With this code I obtain an object from the following class:
QueryClass { productproductcategory, category}
Where producproductcategory is of type:
ProductProductCategoryClass {product, productcategory}
I do not understand where the joined "table" is, I was expecting a single class that contains all the properties from the involved classes.
My aim is to populate another object with some properties resulting from the query:
CategorizedProducts catProducts = query.Select(m => new { m.ProdId = ???, m.CatId = ???, //other assignments });
how can I achieve this goal?

For joins, I strongly prefer query-syntax for all the details that are happily hidden (not the least of which are the transparent identifiers involved with the intermediate projections along the way that are apparent in the dot-syntax equivalent). However, you asked regarding Lambdas which I think you have everything you need - you just need to put it all together.
var categorizedProducts = product
.Join(productcategory, p => p.Id, pc => pc.ProdId, (p, pc) => new { p, pc })
.Join(category, ppc => ppc.pc.CatId, c => c.Id, (ppc, c) => new { ppc, c })
.Select(m => new {
ProdId = m.ppc.p.Id, // or m.ppc.pc.ProdId
CatId = m.c.CatId
// other assignments
});
If you need to, you can save the join into a local variable and reuse it later, however lacking other details to the contrary, I see no reason to introduce the local variable.
Also, you could throw the Select into the last lambda of the second Join (again, provided there are no other operations that depend on the join results) which would give:
var categorizedProducts = product
.Join(productcategory, p => p.Id, pc => pc.ProdId, (p, pc) => new { p, pc })
.Join(category, ppc => ppc.pc.CatId, c => c.Id, (ppc, c) => new {
ProdId = ppc.p.Id, // or ppc.pc.ProdId
CatId = c.CatId
// other assignments
});
...and making a last attempt to sell you on query syntax, this would look like this:
var categorizedProducts =
from p in product
join pc in productcategory on p.Id equals pc.ProdId
join c in category on pc.CatId equals c.Id
select new {
ProdId = p.Id, // or pc.ProdId
CatId = c.CatId
// other assignments
};
Your hands may be tied on whether query-syntax is available. I know some shops have such mandates - often based on the notion that query-syntax is somewhat more limited than dot-syntax. There are other reasons, like "why should I learn a second syntax if I can do everything and more in dot-syntax?" As this last part shows - there are details that query-syntax hides that can make it well worth embracing with the improvement to readability it brings: all those intermediate projections and identifiers you have to cook-up are happily not front-and-center-stage in the query-syntax version - they are background fluff. Off my soap-box now - anyhow, thanks for the question. :)

What you've seen is what you get - and it's exactly what you asked for, here:
(ppc, c) => new { productproductcategory = ppc, category = c}
That's a lambda expression returning an anonymous type with those two properties.
In your CategorizedProducts, you just need to go via those properties:
CategorizedProducts catProducts = query.Select(
m => new {
ProdId = m.productproductcategory.product.Id,
CatId = m.category.CatId,
// other assignments
});

take look at this sample code from my project
public static IList<Letter> GetDepartmentLettersLinq(int departmentId)
{
IEnumerable<Letter> allDepartmentLetters =
from allLetter in LetterService.GetAllLetters()
join allUser in UserService.GetAllUsers() on allLetter.EmployeeID equals allUser.ID into usersGroup
from user in usersGroup.DefaultIfEmpty()// here is the tricky part
join allDepartment in DepartmentService.GetAllDepartments() on user.DepartmentID equals allDepartment.ID
where allDepartment.ID == departmentId
select allLetter;
return allDepartmentLetters.ToArray();
}
in this code I joined 3 tables and I spited join condition from where clause
note: the Services classes are just warped(encapsulate) the database operations

public ActionResult Index()
{
List<CustomerOrder_Result> obj = new List<CustomerOrder_Result>();
var orderlist = (from a in db.OrderMasters
join b in db.Customers on a.CustomerId equals b.Id
join c in db.CustomerAddresses on b.Id equals c.CustomerId
where a.Status == "Pending"
select new
{
Customername = b.Customername,
Phone = b.Phone,
OrderId = a.OrderId,
OrderDate = a.OrderDate,
NoOfItems = a.NoOfItems,
Order_amt = a.Order_amt,
dis_amt = a.Dis_amt,
net_amt = a.Net_amt,
status=a.Status,
address = c.address,
City = c.City,
State = c.State,
Pin = c.Pin
}) ;
foreach (var item in orderlist)
{
CustomerOrder_Result clr = new CustomerOrder_Result();
clr.Customername=item.Customername;
clr.Phone = item.Phone;
clr.OrderId = item.OrderId;
clr.OrderDate = item.OrderDate;
clr.NoOfItems = item.NoOfItems;
clr.Order_amt = item.Order_amt;
clr.net_amt = item.net_amt;
clr.address = item.address;
clr.City = item.City;
clr.State = item.State;
clr.Pin = item.Pin;
clr.status = item.status;
obj.Add(clr);
}

var query = from a in d.tbl_Usuarios
from b in d.tblComidaPreferidas
from c in d.tblLugarNacimientoes
select new
{
_nombre = a.Nombre,
_comida = b.ComidaPreferida,
_lNacimiento = c.Ciudad
};
foreach (var i in query)
{
Console.WriteLine($"{i._nombre } le gusta {i._comida} y nació en {i._lNacimiento}");
}

it has been a while but my answer may help someone:
if you already defined the relation properly you can use this:
var res = query.Products.Select(m => new
{
productID = product.Id,
categoryID = m.ProductCategory.Select(s => s.Category.ID).ToList(),
}).ToList();