Do I have to loop through an IEnumerable return from Dapper even though I only return a single object? - dapper

I'm using Dapper to retrieve employee information when I select that employee from a list. Everything maps correctly, and then the rows are grouped according to employee.id. Just what I want. But Dapper returns an IEnumerable, which makes sense when I query for multiple employees and have to make multiple objects; but it makes less sense when I'm only returning the one. Is there a solution to this, or do I just need to loop through the single item? Here's my code:
public async Task<List<EmployeeModel>> GetSelectedEmployee(int selectedEmployeeID)
{
using (IDbConnection connection = new System.Data.SqlClient.SqlConnection(GlobalConfig.CnnString("WorkDeskDB")))
{
var par = new
{
SelectedEmployeeID = selectedEmployeeID
};
var sql = #"SELECT e.id, e.FirstName, e.LastName, e.Nickname,
em.EmployeeID, em.Address, em.Type,
e.JobTitleID, jt.id, jt.Name,
p.EmployeeID, p.Number, p.Type,
ect.EmployeeID, ect.NameID, ect.InitialDate, ect.ExpirationDate,
ct.id, ct.Name
FROM dbo.Employees e
LEFT JOIN dbo.Emails em ON em.EmployeeID = e.id
LEFT JOIN dbo.JobTitles jt ON e.JobTitleID = jt.id
LEFT JOIN Phones p ON p.EmployeeID = e.id
LEFT JOIN dbo.EmployeeCertificationType ect ON ect.EmployeeID = e.id
LEFT JOIN dbo.CertificationType ct ON ect.NameID = ct.id
WHERE e.id = #SelectedEmployeeID";
var employees = await connection.QueryAsync<EmployeeModel, EmailModel, TitleModel, PhoneModel, CertificationModel, EmployeeModel>(sql, (e, em, t, p, c) =>
{
e.EmailList.Add(em);
e.JobTitle = t;
e.PhoneList.Add(p);
e.CertificationList.Add(c);
return e;
},
par, splitOn: "EmployeeID, JobTitleID, EmployeeID, EmployeeID");
var result = employees.GroupBy(e => e.ID).Select(g =>
{
var groupedEmployee = g.First();
groupedEmployee.EmailList = g.Select(e => e.EmailList.Single()).ToList();
return groupedEmployee;
});
return result.ToList();
}
}

You only have one Employee. The other rows are because an Employee can have multiple Phones, Emails and Certifications.
I suggest you do something like this:
EmployeeModel employee = null;
await connection.QueryAsync<EmployeeModel, EmailModel, TitleModel, PhoneModel, CertificationModel, EmployeeModel>(sql, (e, em, t, p, c) =>
{
if (employee is null)
{
employee = e;
employee.JobTitle = t;
}
employee.EmailList.Add(em);
employee.PhoneList.Add(p);
employee.CertificationList.Add(c);
return employee;
},
par, splitOn: "EmployeeID, JobTitleID, EmployeeID, EmployeeID");
// go on using employee, no need for the employees list ...
Then you skip the afterburning of the list.

My suggestion is that you replace the call connection.QueryAsync<EmployeeModel> by other that returns only one element, as stated here

Related

Does the left join order matter in linq query (EF Core 2)?

I have book table which has a relationship with three tables UserFavorites, UserRates, DiscountItems and last one (DiscountItems) has a relationship with Discounts table so I want to load discount with book if it exist
When create linq left join query with this order (join userfavoirtes then discountitems then discount then userate) generated SQL code is fine as expected
But when change order like below (join userfavoirtes then userate then discountitems then discount) generated query only contain first two join only
var books = (from b in result
join uf in UnitOfWork.Context.UserFavorites
on new { b.Id, RelatedType = RelatedTypeEnum.Book, UserId = userId }
equals new { Id = uf.RelatedId, uf.RelatedType, uf.UserId } into buf
from f in buf.DefaultIfEmpty()
join di in UnitOfWork.Context.DiscountItems
on b.Id equals di.BookId into bdi
from di in bdi.DefaultIfEmpty()
join d in UnitOfWork.Context.Discounts
on (di != null ? di.DiscountId : 0) equals d.Id into bd
from d in bd.DefaultIfEmpty()
join ur in UnitOfWork.Context.UserRates
on new { b.Id, UserId = userId }
equals new { Id = ur.BookId, UserId = ur.CreatedBy } into bur
from r in bur.DefaultIfEmpty()
select new BookViewModel
{
Id = b.Id,
Title = b.Title,
Price = b.Price,
BookImage = SetDefaultImageIfNoImage(b.BookImage),
IsFavorite = f != null ? f.IsFavorite : false,
Rate = bur.Any() ? Math.Round(bur.Average(a => a.Value), 1) : 0,
Discount = d
})
.OrderBy(a => a.CategoryId)
.ThenBy(a => a.Authors)
.Take(10).ToList();
this is generated sql query
SELECT *FROM [Book] AS [a]
LEFT JOIN [UserFavorites] AS [uf] ON (([a].[Id] = [uf].[RelatedId]) AND (3 = [uf].[RelatedType])) AND [uf].[UserId] IS NULL
LEFT JOIN [DiscountItems] AS [di] ON [a].[Id] = [di].[BookId]
LEFT JOIN [Discounts] AS [d] ON CASE
WHEN [di].[Id] IS NOT NULL
THEN [di].[DiscountId] ELSE 0
END = [d].[Id]
LEFT JOIN [UserRates] AS [ur] ON ([a].[Id] = [ur].[BookId]) AND [ur].[CreatedBy] IS NULL
WHERE ((([a].[Active] = 1) AND ([a].[Id] <> 62)) AND ([a].[Status] = 1))
but when write join with this order
var books = (from b in result
join uf in UnitOfWork.Context.UserFavorites
on new { b.Id, RelatedType = RelatedTypeEnum.Book, UserId = userId }
equals new { Id = uf.RelatedId, uf.RelatedType, uf.UserId } into buf
from f in buf.DefaultIfEmpty()
join ur in UnitOfWork.Context.UserRates
on new { b.Id, UserId = userId }
equals new { Id = ur.BookId, UserId = ur.CreatedBy } into bur
from r in bur.DefaultIfEmpty()
join di in UnitOfWork.Context.DiscountItems
on b.Id equals di.BookId into bdi
from di in bdi.DefaultIfEmpty()
join d in UnitOfWork.Context.Discounts
on (di != null ? di.DiscountId : 0) equals d.Id into bd
from d in bd.DefaultIfEmpty()
select new BookViewModel
{
Id = b.Id,
Title = b.Title,
Price = b.Price,
BookImage = SetDefaultImageIfNoImage(b.BookImage),
IsFavorite = f != null ? f.IsFavorite : false,
Rate = bur.Any() ? Math.Round(bur.Average(a => a.Value), 1) : 0,
Discount = d
})
.OrderBy(a => a.CategoryId)
.ThenBy(a => a.Authors)
.Take(10).ToList();
generated SQL query contains only the first two joins
SELECT *
FROM [Book] AS [a]
LEFT JOIN [UserFavorites] AS [uf] ON (([a].[Id] = [uf].[RelatedId]) AND (3 = [uf].[RelatedType])) AND [uf].[UserId] IS NULL
LEFT JOIN [UserRates] AS [ur] ON ([a].[Id] = [ur].[BookId]) AND [ur].[CreatedBy] IS NULL
WHERE ((([a].[Active] = 1) AND ([a].[Id] <> 62)))

Entity Framework Core write MSSQL Server Extended Properties

Is it possible to create the MSSQL Server specific extended properties via Fluent-API or DataAnnotation for a Table / Schema? I would like to include my documentation into the sql server tables to satisfy our DBA.
Kind Regards
I started on an implementation using EntityFrameworkCore.Scaffolding.Handlebars but ran out of time. These are the findings:
Add
public class ScaffoldingDesignTimeServices : IDesignTimeServices
{
public void ConfigureDesignTimeServices(IServiceCollection services)
{
services.AddSingleton<IDatabaseModelFactory, SqlServerDatabaseModelFactory2>();
var options = ReverseEngineerOptions.DbContextAndEntities;
services.AddHandlebarsScaffolding(options);
// https://github.com/TrackableEntities/EntityFrameworkCore.Scaffolding.Handlebars/issues/30
Handlebars.RegisterHelper("f-pn", FormatPropertyName);
}
void FormatPropertyName(TextWriter writer, object context, object[] args)
{
writer.WriteSafeString(args[0].ToString());
}
}
Copy SqlServerDatabaseModelFactory from
SqlServerDatabaseModelFactory and customize it with a function going like this
DbConnection connection,
IReadOnlyList<DatabaseTable> tables,
string tableFilter,
IReadOnlyDictionary<string, (string storeType, string typeName)> typeAliases)
{
using (var command = connection.CreateCommand())
{
var commandText = #"
SELECT u.name AS [table_schema],
t.name AS [table_name],
td.value AS [table_desc],
c.name AS [column_name],
cd.value AS [column_desc]
FROM sysobjects t
INNER JOIN sysusers u
ON u.uid = t.uid
LEFT OUTER JOIN sys.extended_properties td
ON td.major_id = t.id
AND td.minor_id = 0
AND td.name = 'MS_Description'
INNER JOIN syscolumns c
ON c.id = t.id
LEFT OUTER JOIN sys.extended_properties cd
ON cd.major_id = c.id
AND cd.minor_id = c.colid
AND cd.name = 'MS_Description'
WHERE t.type = 'u'
ORDER BY t.name, c.colorder";
command.CommandText = commandText;
using (var reader = command.ExecuteReader())
{
var tableColumnGroups = reader.Cast<DbDataRecord>()
.GroupBy(
ddr => (tableSchema: ddr.GetValueOrDefault<string>("table_schema"),
tableName: ddr.GetValueOrDefault<string>("table_name")));
foreach (var tableColumnGroup in tableColumnGroups)
{
var tableSchema = tableColumnGroup.Key.tableSchema;
var tableName = tableColumnGroup.Key.tableName;
var table = tables.Single(t => t.Schema == tableSchema && t.Name == tableName);
foreach (var dataRecord in tableColumnGroup)
{
var columnName = dataRecord.GetValueOrDefault<string>("column_name");
var tableDesc = dataRecord.GetValueOrDefault<string>("table_desc");
var columnDesc = dataRecord.GetValueOrDefault<string>("column_desc");
//_logger.ColumnFound(
// DisplayName(tableSchema, tableName),
// columnName,
// ordinal,
// DisplayName(dataTypeSchemaName, dataTypeName),
// maxLength,
// precision,
// scale,
// nullable,
// isIdentity,
// defaultValue,
// computedValue);
table.Description = tableDesc; ???
table.Columns.FirstOrDefault(x => x.Name == columnName)?.Description = columnDesc; ???;
}
}
}
}
}
Make a dictionary of the table and column descriptions and use Handlebars Helpers/Transformers EntityFrameworkCore.Scaffolding.Handlebars

SQL query select

I am using MSSQL Server and Nodejs with npm mssql. I need to build a query.
Explanation:
Need to select Users from table that did not receive emails today or never receive.
Query example:
SELECT * FROM dbo.Users u
LEFT JOIN dbo.userEmailHistory ueh
ON u.id = ueh.userId
WHERE IsNull(sDate, '') = ''
OR CONVERT(date,sDate)>=CONVERT(date,DATEADD(day,1,getdate()))
Then:
Need to select the user favorite brand. I am doing async. So the code is looking like this:
function getUserSalesByMalls(params, callback) {
var usersArray = [];
var targetID = params.type;
async.eachLimit(params, 1, function (user, cb) {
sqlRequest("
SELECT TOP 1 bts.BrandCategoryID as title, b.title as brand, b.id
FROM dbo.Users st JOIN SaleView ss ON (st.ID=ss.userID)
JOIN Sales sa ON (sa.ID=ss.SaleID)
JOIN Brands b ON (b.ID=sa.BrandID)
JOIN KEY_BrandcategoryToSale bts ON (bts.SaleID=sa.ID)
WHERE st.ID="+**user.id**+"
GROUP BY bts.BrandCategoryID, b.title, b.id
ORDER BY COUNT(bts.BrandCategoryID) desc",
function (err, result) {
user.favBrand = result;
console.log(user);
usersArray.push(user);
cb();
})
}, function () {
callback(null, usersArray)
})
}
Then:
I need to select top 2 sales by user favorite brand that were most viewed in period 14 days and user never seen it:
function getUserSalesByMalls(params, callback) {
var usersArray = [];
var targetID = params.type;
async.eachLimit(params, 1, function (user, cb) {
sqlRequest("
`SELECT DISTINCT TOP 2 s.id as saleId, s.title, s.description, s.imageUrl, count(sv.saleId) as mostViewsPeriod14Days, s.guid, stm.mallId as mallId, brand.title as brandTitle FROM dbo.Sales s
INNER JOIN dbo.SalesToMall stm ON s.id = stm.saleId
INNER JOIN dbo.SaleView sv ON s.id = sv.saleId
INNER JOIN dbo.Brands brand ON s.brandID = brand.id
WHERE (sv.date <= GETDATE())
AND (sv.date >= GETDATE() - 14) AND s.isActive = 1 AND s.isHotSale = 1 AND b.id = "+user.favBrandId+"
AND s.id NOT IN (SELECT DISTINCT sv2.saleId FROM dbo.SaleView sv2
WHERE sv2.userId = "+user.id+")
GROUP BY s.id, s.title, s.description, s.imageUrl, s.guid, stm.mallId, brand.title
ORDER BY COUNT(sv.saleId) DESC`",
function (err, result) {
user.salesByBrand = result;
console.log(user);
usersArray.push(user);
cb();
})
}, function () {
callback(null, usersArray)
})
}
And then last one: i need to select 3 suggestions for the user by mall id that were most viewed in period 14 days and user never seen it:
function getUserSalesByMalls(params, callback) {
var usersArray = [];
var targetID = params.type;
async.eachLimit(params, requestPerLimit, function (user, cb) {
setTimeout(function () {
sqlRequest("SELECT DISTINCT TOP 3 s.id as saleId, s.title, s.description, s.imageUrl, count(sv.saleId) as mostViewsPeriod14Days, s.guid, stm.mallId as mallId, brand.title as brandTitle
FROM dbo.Sales s
INNER JOIN dbo.SalesToMall stm ON s.id = stm.saleId
INNER JOIN dbo.SaleView sv ON s.id = sv.saleId
INNER JOIN dbo.Brands brand ON s.brandID = brand.id
WHERE (sv.date <= GETDATE())
AND (sv.date >= GETDATE() - 14)
AND s.isActive = 1 AND s.isHotSale = 1
AND stm.mallId = "+user.mallId+"
AND s.id NOT IN (SELECT DISTINCT sv2.saleId
FROM dbo.SaleView sv2
WHERE sv2.userId = "+user.id+")
GROUP BY s.id, s.title, s.description, s.imageUrl, s.guid, stm.mallId, brand.title
ORDER BY COUNT(sv.saleId) DESC", function (err, result) {
if (!result) {
cb();
} else if (result.length == 3) {
sqlInsert("INSERT INTO UserEmailHistory (userID,sDate,targetID) VALUES (" + user.id + ",CONVERT(datetime,DATEADD(hour," + utc_diff + ",getdate()))," + targetID + ")");
console.log("USER WITH 3 SALES BY MALL");
user.sales = result;
console.log(user);
usersArray.push(user);
cb();
}
})
}, requestPerMillisecond)
}, function () {
callback(null, usersArray)
})
}
My question is:
Is there any way to combine and build only one query and get all necessary data regarding my explanation?

Dapper - QueryMultiple - 3 tables

I had this relation:
How to retrieve the information in an order entity and invoice entity with a QueryMultiple entity ?
Thanks
QueryMultiple is used when you are accessing multiple result sets, i.e. multiple select, as in:
select * from Order where Id=#id
select * from Invoice where Id = (...probably some sub-query)
At the moment, there is no inbuilt API to stitch this type of query together; instead you would do something like:
using(var multi = conn.QueryMultiple(...)) {
var order = multi.ReadSingle<Order>();
order.Invoice = multi.ReadSingleOrDefault<Invoice>(); // could be null if 0 rows
return order;
}
I would like to add an improved API for this scenario, but it is very awkward to express "join this to that using this property as the association, where {this}.{SomeMember} equals {that}.{SomeOtherMember}".
However, if you are actually doing a single query, as in:
select o.*, i.*
from Order o
left outer join Link l on ...
left outer join Invoice i on ...
where o.Id = #id
then you can use the various Query<,...,> overloads; for example:
int id = ...
var order = conn.Query<Order, Invoice, Order>(sql,
(x,y) => {x.Invoice = y; return x;}, args: new { id }, splitOn: "NumOrder").Single();
Generic code for three tables:
public static Tuple<IEnumerable<T1>, IEnumerable<T2>, IEnumerable<T3>> ExecuteQueryMultiple<T1, T2, T3>(string sql, object parameters,
Func<GridReader, IEnumerable<T1>> func1,
Func<GridReader, IEnumerable<T2>> func2,
Func<GridReader, IEnumerable<T3>> func3)
{
var objs = getMultiple(sql, parameters, func1, func2, func3);
return Tuple.Create(objs[0] as IEnumerable<T1>, objs[1] as IEnumerable<T2>, objs[2] as IEnumerable<T3>);
}
private static List<object> getMultiple(string procedureName, object param, params Func<GridReader, object>[] readerFuncs)
{
var returnResults = new List<object>();
using (SqlConnection sqlCon = new SqlConnection(connectionString))
{
var gridReader = sqlCon.QueryMultiple(procedureName, param, commandType: CommandType.StoredProcedure);
foreach (var readerFunc in readerFuncs)
{
var obj = readerFunc(gridReader);
returnResults.Add(obj);
}
}
return returnResults;
}
Controller:
[HttpPost]
public ActionResult GetCommodityDetails(int ID)
{
var data = new List<Commodity>();
DynamicParameters param = new DynamicParameters();
param.Add("#ATTRIBUTETYPE", "Your parameter");
param.Add("#CID", Your parameter);
var result = DapperORM.ExecuteQueryMultiple("Store procedure name", param, gr => gr.Read<order>(), gr => gr.Read<Invoice>(), gr => gr.Read<Link>());
return Json(result, JsonRequestBehavior.AllowGet);
}
You can use this concept. It worked for me

LINQ orderby int array index value

Using LINQ I would like to sort by the passed in int arrays index.
So in the code below attribueIds is my int array. I'm using the integers in that array for the where clause but I would like the results in the order that they were in while in the array.
public List BuildTable(int[] attributeIds)
{
using (var dc = new MyDC())
{
var ordering = attributeIds.ToList();
var query = from att in dc.DC.Ecs_TblAttributes
where attributeIds.Contains(att.ID)
orderby(ordering.IndexOf(att.ID))
select new Common.Models.Attribute
{
AttributeId = att.ID,
DisplayName = att.DisplayName,
AttributeName = att.Name
};
return query.ToList();
}
}
I would recommend selecting from the attributeIDs array instead. This will ensure that your items will be correctly ordered without requiring a sort.
The code should go something like this:
var query =
from id in attributeIds
let att = dc.DC.Ecs_TblAttributes.FirstOrDefault(a => a.ID == id)
where att != null
select new Common.Models.Attribute
{
AttributeId = att.ID,
DisplayName = att.DisplayName,
AttributeName = att.Name
};
Why don't you join:
public List BuildTable(int[] attributeIds)
{
using (var dc = new MyDC())
{
var query = from attID in attributeIds
join att in dc.DC.Ecs_TblAttributes
on attID equals att.ID
select new Common.Models.Attribute
{
AttributeId = attID,
DisplayName = att.DisplayName,
AttributeName = att.Name
};
return query.ToList();
}
}

Resources