I have looked all through this site and have been banging my head to resolve a seemingly simple question with EF and Fluent API. I've tried to fit some of the answers to my problem to no avail so my apologies in advance.
I am using VS2013 and C#. I have the following model defined:
[Table("party")]
public class Party
{
[Key, DatabaseGenerated(DatabaseGeneratedOption.None)]
[Column(Order = 0)]
[ScaffoldColumn(false)]
public long party_id { get; set; }
[Key]
[Column(Order = 1)]
[ScaffoldColumn(false)]
public string party_type { get; set; }
}
I have another model like this:
[Table("person")]
public class Person
{
[ScaffoldColumn(false)]
public long person_id { get; set; }
public virtual Party party { get; set; }
}
The PERSON model needs to be foreign key linked to the PARTY. The "party_type" column holds a constant string of the base-type of party being added with incrementing values on "party_id". Meaning I could have:
Party_ID Party_Type
1 person
2 person
1 company
1 org
2 company
3 person
How do I in FLUENT API connect the PERSON model to the PARTY model based on the "person_id" column and a constant of "person"?
Thanks and help!
John
Related
I'm looking for a way to map several POCO objects into single table in the ServiceStack.
Is it possible to do this in a clean way, without "hacking" table creation process?
As a general rule, In OrmLite: 1 Class = 1 Table.
But I'm not clear what you mean my "map several POCO objects into single table", it sounds like using Auto Mapping to populate a table with multiple POCO instances, e.g:
var row = db.SingleById<Table>(id);
row.PopulateWithNonDefaultValues(instance1);
row.PopulateWithNonDefaultValues(instance2);
db.Update(row);
If you need to maintain a single table and have other "sub" classes that maintain different table in the universal table you can use [Alias] so all Update/Select/Insert's reference the same table, e.g:
public class Poco
{
public int Id { get; set; }
public string Name { get; set; }
public int Age { get; set; }
}
[Alias(nameof(Poco))]
public class PocoName
{
public int Id { get; set; }
public string Name { get; set; }
}
[Alias(nameof(Poco))]
public class PocoAge
{
public int Id { get; set; }
public int Age { get; set; }
}
Although I don't really see the benefit over having a single table that you use AutoMapping to map your other classes to before using that in OrmLite.
I have a product class that is stored in Cosmos DB together with its variants.
[DataContract]
public class Product
{
[DataMember]
public virtual string Name { get; set; }
[DataMember]
public virtual IList<Variant> Variants { get; set; }
}
[DataContract]
public class Variant
{
[DataMember]
public virtual string Name { get; set; }
}
I would like to query a projection of the variants that includes the Product.
[DataContract]
public class VariantProjection
{
[DataMember]
public virtual Product Product { get; set; }
[DataMember]
public virtual string Name { get; set; }
}
I'm using the DocumentDB Linq api, but if it's not possible with this api any other API would be ok.
Sounds like you are looking for JOINs and Projections (might want to try the Cosmos DB Query Playground, it has scenarios for both).
It would be great to have a simple dataset to test but I believe something like this might help:
SELECT p as product, variant.name
FROM p
JOIN variant IN p.variants
Keep in mind though that you are retrieving the entire Product for each variant. That's what you are trying to achieve in your C# code.
We have a DTO - Employee - with many (> 20) related DTOs and DTO collections. For "size of returned JSON" reasons, we have marked those relationships as [Ignore]. It is then up to the client to populate any related DTOs that they would like using other REST calls.
We have tried a couple of things to satisfy clients' desire to have some related Employee info but not all:
We created a new DTO - EmployeeLite - which has the most-requested fields defined with "RelatedTableNameRelatedFieldName" approach and used the QueryBase overload and that has worked well.
We've also tried adding a property to a request DTO - "References" - which is a comma-separated list of related DTOs that the client would like populated. We then iterate the response and populate each Employee with the related DTO or List. The concern there is performance when iterating a large List.
We're wondering if there a suggested approach to what we're trying to do?
Thanks for any suggestions you may have.
UPDATE:
Here is a portion of our request DTO:
[Route("/employees", "GET")]
public class FindEmployeesRequest : QueryDb<Employee> {
public int? ID { get; set; }
public int[] IDs { get; set; }
public string UserID { get; set; }
public string LastNameStartsWith { get; set; }
public DateTime[] DateOfBirthBetween { get; set; }
public DateTime[] HireDateBetween { get; set; }
public bool? IsActive { get; set; }
}
There is no code for the service (automagical with QueryDb), so I added some to try the "merge" approach:
public object Get(FindEmployeesRequest request) {
var query = AutoQuery.CreateQuery(request, Request.GetRequestParams());
QueryResponse<Employee> response = AutoQuery.Execute(request, query);
if (response.Total > 0) {
List<Clerkship> clerkships = Db.Select<Clerkship>();
response.Results.Merge(clerkships);
}
return response;
}
This fails with Could not find Child Reference for 'Clerkship' on Parent 'Employee'
because in Employee we have:
[Ignore]
public List<Clerkship> Clerkships { get; set; }
which we did because we don't want "Clerkships" with every request. If I change [Ignore] to [Reference] I don't need the code above in the service - the List comes automatically. So it seems that .Merge only works with [Reference] which we don't want to do.
I'm not sure how I would use the "Custom Load References" approach in an AutoQuery service. And, AFAIKT, the "Custom Fields" approach can't be use for related DTOs, only for fields in the base table.
UPDATE 2:
The LoadSelect with include[] is working well for us. We are now trying to cover the case where ?fields= is used in the query string but the client does not request the ID field of the related DTO:
public partial class Employee {
[PrimaryKey]
[AutoIncrement]
public int ID { get; set; }
.
.
.
[References(typeof(Department))]
public int DepartmentID { get; set; }
.
.
.
public class Department {
[PrimaryKey]
public int ID { get; set; }
public string Name { get; set; }
.
.
.
}
So, for the request
/employees?fields=id,departmentid
we will get the Department in the response. But for the request
/employees?fields=id
we won't get the Department in the response.
We're trying to "quietly fix" this for the requester by modifying the query.SelectExpression and adding , "Employee"."DepartmentID" to the SELECT before doing the Db.LoadSelect. Debugging shows that query.SelectExpression is being modified, but according to SQL Profiler, "Employee"."DepartmentID" is not being selected.
Is there something else we should be doing to get "Employee"."DepartmentID" added to the SELECT?
Thanks.
UPDATE 3:
The Employee table has three 1:1 relationships - EmployeeType, Department and Title:
public partial class Employee {
[PrimaryKey]
[AutoIncrement]
public int ID { get; set; }
[References(typeof(EmployeeType))]
public int EmployeeTypeID { get; set; }
[References(typeof(Department))]
public int DepartmentID { get; set; }
[References(typeof(Title))]
public int TitleID { get; set; }
.
.
.
}
public class EmployeeType {
[PrimaryKey]
public int ID { get; set; }
public string Name { get; set; }
}
public class Department {
[PrimaryKey]
public int ID { get; set; }
public string Name { get; set; }
[Reference]
public List<Title> Titles { get; set; }
}
public class Title {
[PrimaryKey]
public int ID { get; set; }
[References(typeof(Department))]
public int DepartmentID { get; set; }
public string Name { get; set; }
}
The latest update to 4.0.55 allows this:
/employees?fields=employeetype,department,title
I get back all the Employee table fields plus the three related DTOs - with one strange thing - the Employee's ID field is populated with the Employee's TitleID values (I think we saw this before?).
This request fixes that anomaly:
/employees?fields=id,employeetypeid,employeetype,departmentid,department,titleid,title
but I lose all of the other Employee fields.
This sounds like a "have your cake and eat it too" request, but is there a way that I can get all of the Employee fields and selective related DTOs? Something like:
/employees?fields=*,employeetype,department,title
AutoQuery Customizable Fields
Not sure if this is Relevant but AutoQuery has built-in support for Customizing which fields to return with the ?fields=Field1,Field2 option.
Merge disconnected POCO Results
As you've not provided any source code it's not clear what you're trying to achieve or where the inefficiency with the existing solution lies, but you don't want to be doing any N+1 SELECT queries. If you are, have a look at how you can merge disconnected POCO results together which will let you merge results from separate queries based on the relationships defined using OrmLite references, e.g the example below uses 2 distinct queries to join Customers with their orders:
//Select Customers who've had orders with Quantities of 10 or more
List<Customer> customers = db.Select<Customer>(q =>
q.Join<Order>()
.Where<Order>(o => o.Qty >= 10)
.SelectDistinct());
//Select Orders with Quantities of 10 or more
List<Order> orders = db.Select<Order>(o => o.Qty >= 10);
customers.Merge(orders); // Merge disconnected Orders with their related Customers
Custom Load References
You can selectively control which references OrmLite should load by specifying them when you call OrmLite's Load* API's, e.g:
var customerWithAddress = db.LoadSingleById<Customer>(customer.Id,
include: new[] { "PrimaryAddress" });
Using Custom Load References in AutoQuery
You can customize an AutoQuery Request to not return any references by using Db.Select instead of Db.LoadSelect in your custom AutoQuery implementation, e.g:
public object Get(FindEmployeesRequest request)
{
var q = AutoQuery.CreateQuery(request, Request);
var response = new QueryResponse<Employee>
{
Offset = q.Offset.GetValueOrDefault(0),
Results = Db.Select(q),
Total = (int)Db.Count(q),
};
return response;
}
Likewise if you only want to selectively load 1 or more references you can change LoadSelect to pass in an include: array with only the reference fields you want included, e.g:
public object Get(FindEmployeesRequest request)
{
var q = AutoQuery.CreateQuery(request, Request);
var response = new QueryResponse<Employee>
{
Offset = q.Offset.GetValueOrDefault(0),
Results = Db.LoadSelect(q, include:new []{ "Clerkships" }),
Total = (int)Db.Count(q),
};
return response;
}
I see these types of model is many samples online.
public class User
{
public long Id { get; set; }
public string Name{ get; set; }
public virtual ICollection<Product> Products { get; set; }
}
Is it considered a good practice to instantiate a collection in the constructor like the code below? If so what are the reasons? How about objects in the model?
public class User
{
public User()
{
Products = new List<Product>();
}
public long Id { get; set; }
public string Name{ get; set; }
public virtual ICollection<Product> Products { get; set; }
}
Well, I would say it depends on the situation, but Products in this case would be filled from the database, via a repository, so most probably ORM of some sort, so no initialization to new List would be needed in the constructor. The meaning of null for Products is indicative that the list isn't loaded yet. On the other hand, let's say that your object must have this collection initialized. For simple objects DDD says constructors are perfectly fine to to these things, but in case of complex objects, move the construction to the Factory.
I have this domain model (simplified), that represents a product that has a basic price and attached has numerous suppliers that provide a specific discount percentage against the basic price:
public class CarDerivative
{
public string Name { get; set; } e.g. StackOverflow Supercar 3.0L Petrol
public double BasicPrice { get; set; } e.g. 10,000
public double Delivery { get; set; } e.g. 500
public IList<SupplierDiscount> { get; set; } // has 3 various suppliers
}
public class SupplierDiscount
{
public string SupplierName; { get; set; }
//public SupplierInformation SupplierDetails { get; set; } // maybe later
public double BasicDiscount { get; set; }
public double DeliveryDiscount { get; set; } // e.g. 0.10 = 10%
}
Now, I'm thinking about where stuff should sit that does stuff with this:
For example, where best does BasicDiscountedPrice (stable formula) sit, should it ideally sit on SupplierDiscount which is furnished with a reference to the parent CarDerivative via constructor injection?
Where in your opinion should an unstable formula sit? Such as SupplierPriceForDerivative (basic + delivery + tax +++) ?
My knee-jerk reaction to this would be that Discounts should be Policies. Each DiscountPolicy might look like this:
public interface IDiscountPolicy
{
decimal GetDiscountedPrice(CarDerivative car);
}
Your SupplierDiscount might be a simple implementation of such a Policy, while your unstable formulas might be implemented in a more complex class.
I think it would be safest to keep the Entities and the Policies separate, so that you can vary them independently from each other.