Entity Framework: Tracking changes to FK associations - entity-framework

I'm overriding SaveChanges on my DbContext in order to implement an audit log. Working with many-to-many relationships or independent associations is relatively easy as EF creates ObjectStateEntries for any changes to those kinds of relationships.
I am using foreign key associations, and when a relationship between entities changes all you get is an ObjectStateEnty that says for example entity "Title" has "PublisherID" property changed. To a human this is obviously a foreign key in Title entity, but how do I determine this in runtime? Is there a way to translate this change to a "PublisherID" property to let's an EntityKey for the entity that foreign key represents?
I assume I'm dealing with entities that look like this:
public sealed class Publisher
{
public Guid ID { get; set; }
public string Name { get; set; }
public ICollection<Title> Titles { get; set; }
}
public class Title
{
public Guid ID { get; set; }
public string Name { get; set; }
public Guid? PublisherID { get; set; }
public Publisher Publisher { get; set; }
}
There is also EF EntityConfiguration code that defines the relationship and foreign key:
public TitleConfiguration()
{
HasOptional<Publisher>(t => t.Publisher).WithMany(
p => p.Titles).HasForeignKey(t => t.PublisherID);
}
What I'm doing now seems a bit too complicated. I'm hoping there is more elegant way to achieve my goal. For every modified property from ObjectStateEntry I look through all ReferentialConstraints for current entity and see if any of those use it as a foreign key. The code below is called from SaveChanges():
private void HandleProperties(ObjectStateEntry entry,
ObjectContext ctx)
{
string[] changedProperties = entry.GetModifiedProperties().ToArray();
foreach (string propertyName in changedProperties)
{
HandleForeignKey(entry, ctx, propertyName);
}
}
private void HandleForeignKey(ObjectStateEntry entry,
ObjectContext ctx, string propertyName)
{
IEnumerable<IRelatedEnd> relatedEnds =
entry.RelationshipManager.GetAllRelatedEnds();
foreach (IRelatedEnd end in relatedEnds)
{
// find foreign key relationships
AssociationType elementType = end.RelationshipSet.ElementType as
AssociationType;
if (elementType == null || !elementType.IsForeignKey) continue;
foreach (ReferentialConstraint constraint in
elementType.ReferentialConstraints)
{
// Multiplicity many means we are looking at a foreign key in a
// dependent entity
// I assume that ToRole will point to a dependent entity, don't
// know if it can be FromRole
Debug.Assert(constraint.ToRole.RelationshipMultiplicity ==
RelationshipMultiplicity.Many);
// If not 1 then it is a composite key I guess.
// Becomes a lot more difficult to handle.
Debug.Assert(constraint.ToProperties.Count == 1);
EdmProperty prop = constraint.ToProperties[0];
// entity types of current entity and foreign key entity
// must be the same
if (prop.DeclaringType == entry.EntitySet.ElementType
&& propertyName == prop.Name)
{
EntityReference principalEntity = end as EntityReference;
if (principalEntity == null) continue;
EntityKey newEntity = principalEntity.EntityKey;
// if there is more than one, the foreign key is composite
Debug.Assert(newEntity.EntityKeyValues.Length == 1);
// create an EntityKey for the old foreign key value
EntityKey oldEntity = null;
if (entry.OriginalValues[prop.Name] is DBNull)
{
oldEntity = new EntityKey();
oldEntity.EntityKeyValues = new[] {
new EntityKeyMember("ID", "NULL")
};
oldEntity.EntitySetName = newEntity.EntitySetName;
}
else
{
Guid oldGuid = Guid.Parse(
entry.OriginalValues[prop.Name].ToString());
oldEntity = ctx.CreateEntityKey(newEntity.EntitySetName,
new Publisher()
{
ID = oldGuid
});
}
Debug.WriteLine(
"Foreign key {0} changed from [{1}: {2}] to [{3}: {4}]",
prop.Name,
oldEntity.EntitySetName, oldEntity.EntityKeyValues[0],
newEntity.EntitySetName, newEntity.EntityKeyValues[0]);
}
}
}
}
I hope this helps to illustrate better what I am trying to achieve. Any input is welcome.
Thanks!

Looks like my code is the right solution to this problem :/
I did end up using independent associations to avoid this problem altogether.

Related

What is the proper way of updating navigation properties in EF Core?

In my EF Core solution I have the following model:
public class Deal
{
public string Id { get; set; }
public ResponsiblePerson ResponsiblePerson1 { get; set; }
public ResponsiblePerson ResponsiblePerson2 { get; set; }
public ResponsiblePerson ResponsiblePerson3 { get; set; }
}
public class ResponsiblePerson
{
public string Id { get; set; }
public string Name { get; set; }
}
When I am trying to update Deal navigations properties:
private void UpdateResponsiblePersons(string dealId, string person1Id, string person2Id, string person3Id)
{
var existingdeal = _dbContext.Deals
.Include(d => d.ResponsiblePerson1)
.Include(d => d.ResponsiblePerson2)
.Include(d => d.ResponsiblePerson3)
.Single(d => d.Id == dealId);
existingDeal.ResponsiblePerson1 = new ResponsiblePerson { Id = person1Id };
existingDeal.ResponsiblePerson2 = new ResponsiblePerson { Id = person2Id };
existingDeal.ResponsiblePerson3 = new ResponsiblePerson { Id = person3Id };
_dbContext.Entry(deal.ResponsiblePerson1).State = EntityState.Unchanged;
_dbContext.Entry(deal.ResponsiblePerson3).State = EntityState.Unchanged;
_dbContext.Entry(deal.ResponsiblePerson3).State = EntityState.Unchanged;
_dbContext.SaveChanges();
}
EF often fails with
System.InvalidOperationException: The instance of entity type 'ResponsiblePerson' cannot be tracked because another instance with the key value '{Id: 1}' is already being tracked. When attaching existing entities, ensure that only one entity instance with a given key value is attached.
That is because sometimes existingdeal already contains the link to ResponsiblePerson with one of provided IDs in either ResponsiblePerson1 of ResponsiblePerson2 or ResponsiblePerson3 Navigation properties.
I know that one of possible solutions will be first to get ResponsiblePersons used for update from dbContext like
existingDeal.ResponsiblePerson1 = _dbContext.ResponsiblePersons.Find(person1Id)
But that means extra DB roundtrips.
Another solution is to expose foreign keys instead of navigation properties but it would make Deal model quite ugly.
Please advice me what is the best way of updating such references?

"A dependent property in a ReferentialConstraint is mapped to a store-generated column" with Id change

Situation
I have searched for the answer to this extensively (on SO and elsewhere) and I am aware that there are many questions on SO by this same title.
I had a table mapping and model that were working. Then the schema was changed (I do not have direct control of the DB) such that a new Primary Key was introduced and the old Primary Key became the Foreign Key to another table. I believe this is the heart of the problem as no other entities seem to have issues
Mapping
Here is the method that maps my entity (called from OnModelCreating)
private static void MapThing(DbModelBuilder modelBuilder)
{
modelBuilder.Entity<Thing>().ToTable("ThingTable");
modelBuilder.Entity<Thing>().HasKey(p => p.Id);
modelBuilder.Entity<Thing>().Property(p => p.Id).HasColumnName("NewId");
modelBuilder.Entity<Thing>().Property(p => p.Id).HasDatabaseGeneratedOption(DatabaseGeneratedOption.Identity);
modelBuilder.Entity<Thing>().Property(p => p.FileName).HasColumnName("ColumnWhosNameChanged");
modelBuilder.Entity<Thing>().HasRequired(p => p.MetaDataOnThing);
}
The old PK of the table is now defined as a property on the model and it is the same name as the column (the reason it is not defined in the mapping above).
Model
Here is the Model (I have applied names that I hope will make it more clear what has changed):
public class Thing
{
[DatabaseGeneratedAttribute(DatabaseGeneratedOption.Identity)]
public int Id { get; set; }
//This used to be the PK, its names (Property AND Column) have not changed
public int OldId { get; set; }
//The column name for FileName changed to something else
public string FileName { get; set; }
//Unchanged
public byte[] Document { get; set; }
public string ContentType { get; set; }
//Navigation Property
public ThingMetaData MetaDataOnThing { get; set; }
}
Integration test
I removed a lot of structure to, hopefully, make it clear..the test is pretty straight forward
[TestMethod]
public void ThenThingWillBePersisted()
{
var thing = new Thing()
{
OldId = metaDataObject.Id,
Document = new byte[] { 42 },
FileName = "foo.jpg",
ContentType = "image/jpeg"
};
context.Things.Add(thing);
context.SaveChanges();
}
This test produces the error "A dependent property in a ReferentialConstraint is mapped to a store-generated column. Column:'NewId'" and the inner exception points to the NewId as being the issue. It does so on the SaveChanges() call.
Admittedly, I have a lot more experience with nHibernate than I do with Entity Framework but I am pretty sure my mappings and model are setup properly.
Has anyone seen this issue and how did you solve it?

How to not assign an id when id is a fk with a custom id generator in ef

I have a project where I'm using EF5, I made a custom Guid Generator and I have an override of the SaveChanges method to assign the ids of my entities.
Everything is working fine except in one case: when the ID of one entity is a FK to another ID of another entity.
A little bit of code to explain the problem:
I have two entities I cannot change:
public class FixedEntityA
{
public Guid Id { get; set;}
public string SomeText { get; set; }
}
public class FixedEntityB
{
public Guid Id { get; set;}
public int OneInt { get; set; }
}
In my project I have an entity defined like this:
public class ComposedEntity
{
public Guid Id { get; set;}
public FixedEntityA FixedA { get; set; }
public FixedEntityB FixedB { get; set; }
public double OneDouble { get; set; }
}
The relationships are:
ComposedEntity may have 0 or 1 FixedEntityA
ComposedEntity may have 0 or 1 FixedEntityB
The constraints on the id are:
The Id of FixedEntityA is a FK pointing to the Id of ComposedEntity
The Id of FixedEntityB is a FK pointing to the Id of ComposedEntity
The mapping class are:
public ComposedEntity(): EntityTypeConfiguration<ComposedEntity>
{
HasOptional(fea => fea.FixedA).WithRequired();
HasOptional(feb => feb.FixedB).WithRequired();
}
Here is my SaveChanges override:
foreach (var entry in ChangeTracker.Entries<IEntity>().Where(e => e.State == EntityState.Added))
{
Type t = entry.Entity.GetType();
List<DatabaseGeneratedAttribute> info = t.GetProperty("Id")
.GetCustomAttributes(typeof (DatabaseGeneratedAttribute), true)
.Cast<DatabaseGeneratedAttribute>().ToList();
if (!info.Any() || info.Single().DatabaseGeneratedOption != DatabaseGeneratedOption.Identity)
{
if (entry.Entity.Id == Guid.Empty)
entry.Entity.Id = (Guid) _idGenerator.Generate();
}
}
return base.SaveChanges();
This code works fine everywhere for all kind of relationships except in this case, I am missing a test to make sure I'am not setting an id on id that are foreign keys, and I have no clue on how to check if an Id is a FK...
Here is a sample object where this code fails:
var fea = new FixedEntityA();
var feb = new FixedEntityB();
var composedEntity = new ComposedEntity();
composedEntity.FixedA = fea;
composedEntity.FixedB = feb;
If you insert the whole graph, all three objects are marked as Added and all Ids are default.
The problem is, with the current SaveChanges method, I will go through all object with the Added state in the change tracker and I will assign an Id to all entity with a default Guid and break my FK constraints.
Thanks in advance guys!
Here is some code that will get the FK properties for a given type (it's horrible I know). Should be simple enough to plug this into your code.
var typeName = "Category";
var fkProperties = ((IObjectContextAdapter)db)
.ObjectContext
.MetadataWorkspace
.GetItems<AssociationType>(DataSpace.CSpace)
.Where(a => a.IsForeignKey)
.Select(a => a.ReferentialConstraints.Single())
.Where(c => c.FromRole.GetEntityType().Name == typeName)
.SelectMany(c => c.FromProperties)
.Select(p => p.Name);

How to use Entity Framework to map results of a stored procedure to entity with differently named parameters

I am trying to create a basic example using Entity Framework to do the mapping of the output of a SQL Server Stored procedure to an entity in C#, but the entity has differently (friendly) names parameters as opposed to the more cryptic names. I am also trying to do this with the Fluent (i.e. non edmx) syntax.
What works ....
The stored procedure returns values called: UT_ID, UT_LONG_NM, UT_STR_AD, UT_CITY_AD, UT_ST_AD, UT_ZIP_CD_AD, UT_CT
If I create an object like this ...
public class DBUnitEntity
{
public Int16 UT_ID { get; set; }
public string UT_LONG_NM { get; set; }
public string UT_STR_AD { get; set; }
public string UT_CITY_AD { get; set; }
public string UT_ST_AD { get; set; }
public Int32 UT_ZIP_CD_AD { get; set; }
public string UT_CT { get; set; }
}
and an EntityTypeConfiguration like this ...
public class DbUnitMapping: EntityTypeConfiguration<DBUnitEntity>
{
public DbUnitMapping()
{
HasKey(t => t.UT_ID);
}
}
... which I add in the OnModelCreating of the DbContext, then I can get the entities just fine out of the database, which is nice, using this ....
var allUnits = _context.Database.SqlQuery<DBUnitEntity>(StoredProcedureHelper.GetAllUnitsProc);
BUT, What Doesn't Work
If I want an entity like this, with friendlier names ....
public class UnitEntity : IUnit
{
public Int16 UnitId { get; set; }
public string Name { get; set; }
public string Address { get; set; }
public string City { get; set; }
public string State { get; set; }
public Int32 Zip { get; set; }
public string Category { get; set; }
}
and an EntityTypeConfiguration like this ...
public UnitMapping()
{
HasKey(t => t.UnitId);
Property(t => t.UnitId).HasColumnName("UT_ID");
Property(t => t.Name).HasColumnName("UT_LONG_NM");
Property(t => t.Address).HasColumnName("UT_STR_AD");
Property(t => t.City).HasColumnName("UT_CITY_AD");
Property(t => t.State).HasColumnName("UT_ST_AD");
Property(t => t.Zip).HasColumnName("UT_ZIP_CD_AD");
Property(t => t.Category).HasColumnName("UT_CT");
}
When I try to get the data I get a System.Data.EntityCommandExecutionException with the message ....
"The data reader is incompatible with the specified 'DataAccess.EFCodeFirstSample.UnitEntity'. A member of the type, 'UnitId', does not have a corresponding column in the data reader with the same name."
If I add the "stored procedure named" property to the entity, it goes and complains about the next "unknown" property.
Does "HasColumnName" not work as I expect/want it to in this code-first stored procedure fluent style of EF?
Update:
Tried using DataAnnotations (Key from ComponentModel, and Column from EntityFramework) ... ala
public class UnitEntity : IUnit
{
[Key]
[Column("UT_ID")]
public Int16 UnitId { get; set; }
public string Name { get; set; }
That did remove the need for any EntityTypeConfiguration at all for the DBUnitEntity with the database-identical naming (i.e. just adding the [Key] Attribute), but did nothing for the entity with the property names that don't match the database (same error as before).
I don't mind using the ComponentModel Annotations in the Model, but I really don't want to use the EntityFramework Annotations in the model if I can help it (don't want to tie the Model to any specific data access framework)
From Entity Framework Code First book (page 155):
The SQLQuery method always attempts the column-to-property matching based on property name...
None that the column-to-property name matching does not take any mapping into account. For example, if you had mapped the DestinationId property to a column called Id in the Destination table, the SqlQuery method would not use this mapping.
So you cannot use mappings when calling stored procedure. One workaround is to modify your stored procedure to return result with aliases for each column that will match your object properties' names.
Select UT_STR_AD as Address From SomeTable etc
This isn't using Entity Framework but it is stemming from dbcontext. I have spent hours upon hours scouring the internet and using dot peek all for nothing. I read some where that the ColumnAttribute is ignored for SqlQueryRaw. But I have crafted up something with reflection, generics, sql datareader, and Activator. I am going to be testing it on a few other procs. If there is any other error checking that should go in, comment.
public static List<T> SqlQuery<T>( DbContext db, string sql, params object[] parameters)
{
List<T> Rows = new List<T>();
using (SqlConnection con = new SqlConnection(db.Database.Connection.ConnectionString))
{
using (SqlCommand cmd = new SqlCommand(sql, con))
{
cmd.CommandType = CommandType.StoredProcedure;
foreach (var param in parameters)
cmd.Parameters.Add(param);
con.Open();
using (SqlDataReader dr = cmd.ExecuteReader())
{
if (dr.HasRows)
{
var dictionary = typeof(T).GetProperties().ToDictionary(
field => CamelCaseToUnderscore(field.Name), field => field.Name);
while (dr.Read())
{
T tempObj = (T)Activator.CreateInstance(typeof(T));
foreach (var key in dictionary.Keys)
{
PropertyInfo propertyInfo = tempObj.GetType().GetProperty(dictionary[key], BindingFlags.Public | BindingFlags.Instance);
if (null != propertyInfo && propertyInfo.CanWrite)
propertyInfo.SetValue(tempObj, Convert.ChangeType(dr[key], propertyInfo.PropertyType), null);
}
Rows.Add(tempObj);
}
}
dr.Close();
}
}
}
return Rows;
}
private static string CamelCaseToUnderscore(string str)
{
return Regex.Replace(str, #"(?<!_)([A-Z])", "_$1").TrimStart('_').ToLower();
}
Also something to know is that all of our stored procs return lowercase underscore delimited. The CamelCaseToUnderscore is built specifically for it.
Now BigDeal can map to big_deal
You should be able to call it like so
Namespace.SqlQuery<YourObj>(db, "name_of_stored_proc", new SqlParameter("#param",value),,,,,,,);
The example posted by "DeadlyChambers" is great but I would like to extend the example to include the ColumnAttribute that you can use with EF to add to a properties to map a SQL field to a Class property.
Ex.
[Column("sqlFieldName")]
public string AdjustedName { get; set; }
Here is the modified code.
This code also include a parameter to allow for custom mappings if needed by passing a dictionary.
You will need a Type Converter other than Convert.ChangeType for things like nullable types.
Ex. If you have a field that is bit in the database and nullable boolean in .NET you will get a type convert issue.
/// <summary>
/// WARNING: EF does not use the ColumnAttribute when mapping from SqlQuery. So this is a "fix" that uses "lots" of REFLECTION
/// </summary>
/// <typeparam name="T"></typeparam>
/// <param name="database"></param>
/// <param name="sqlCommandString"></param>
/// <param name="modelPropertyName_sqlPropertyName">Model Property Name and SQL Property Name</param>
/// <param name="sqlParameters">SQL Parameters</param>
/// <returns></returns>
public static List<T> SqlQueryMapped<T>(this System.Data.Entity.Database database,
string sqlCommandString,
Dictionary<string,string> modelPropertyName_sqlPropertyName,
params System.Data.SqlClient.SqlParameter[] sqlParameters)
{
List<T> listOfT = new List<T>();
using (var cmd = database.Connection.CreateCommand())
{
cmd.CommandText = sqlCommandString;
if (cmd.Connection.State != System.Data.ConnectionState.Open)
{
cmd.Connection.Open();
}
cmd.Parameters.AddRange(sqlParameters);
using (var dataReader = cmd.ExecuteReader())
{
if (dataReader.HasRows)
{
// HACK: you can't use extension methods without a type at design time. So this is a way to call an extension method through reflection.
var convertTo = typeof(GenericExtensions).GetMethods(BindingFlags.Static | BindingFlags.Public).Where(mi => mi.Name == "ConvertTo").Where(m => m.GetParameters().Count() == 1).FirstOrDefault();
// now build a new list of the SQL properties to map
// NOTE: this method is used because GetOrdinal can throw an exception if column is not found by name
Dictionary<string, int> sqlPropertiesAttributes = new Dictionary<string, int>();
for (int index = 0; index < dataReader.FieldCount; index++)
{
sqlPropertiesAttributes.Add(dataReader.GetName(index), index);
}
while (dataReader.Read())
{
// create a new instance of T
T newT = (T)Activator.CreateInstance(typeof(T));
// get a list of the model properties
var modelProperties = newT.GetType().GetProperties();
// now map the SQL property to the EF property
foreach (var propertyInfo in modelProperties)
{
if (propertyInfo != null && propertyInfo.CanWrite)
{
// determine if the given model property has a different map then the one based on the column attribute
string sqlPropertyToMap = (propertyInfo.GetCustomAttribute<ColumnAttribute>()?.Name ?? propertyInfo.Name);
string sqlPropertyName;
if (modelPropertyName_sqlPropertyName!= null && modelPropertyName_sqlPropertyName.TryGetValue(propertyInfo.Name, out sqlPropertyName))
{
sqlPropertyToMap = sqlPropertyName;
}
// find the SQL value based on the column name or the property name
int columnIndex;
if (sqlPropertiesAttributes.TryGetValue(sqlPropertyToMap, out columnIndex))
{
var sqlValue = dataReader.GetValue(columnIndex);
// ignore this property if it is DBNull
if (Convert.IsDBNull(sqlValue))
{
continue;
}
// HACK: you can't use extension methods without a type at design time. So this is a way to call an extension method through reflection.
var newValue = convertTo.MakeGenericMethod(propertyInfo.PropertyType).Invoke(null, new object[] { sqlValue });
propertyInfo.SetValue(newT, newValue);
}
}
}
listOfT.Add(newT);
}
}
}
}
return listOfT;
}

Entity Framework 4.1+ many-to-many relationships change tracking

How can I detect changes of ICollection<> properties (many-to-many relationships)?
public class Company
{
...
public virtual ICollection<Employee> Employees { get; set; }
}
using (DataContext context = new DataContext(Properties.Settings.Default.ConnectionString))
{
Company company = context.Companies.First();
company.Employees.Add(context.Employees.First());
context.SaveChanges();
}
public class DataContext : DbContext
{
public override int SaveChanges()
{
return base.SaveChanges();
// Company's entity state is "Unchanged" in this.ChangeTracker
}
}
Here is how to find all the changed many-to-many relationships. I've implemented the code as extension methods:
public static class IaExtensions
{
public static IEnumerable<Tuple<object, object>> GetAddedRelationships(
this DbContext context)
{
return GetRelationships(context, EntityState.Added, (e, i) => e.CurrentValues[i]);
}
public static IEnumerable<Tuple<object, object>> GetDeletedRelationships(
this DbContext context)
{
return GetRelationships(context, EntityState.Deleted, (e, i) => e.OriginalValues[i]);
}
private static IEnumerable<Tuple<object, object>> GetRelationships(
this DbContext context,
EntityState relationshipState,
Func<ObjectStateEntry, int, object> getValue)
{
context.ChangeTracker.DetectChanges();
var objectContext = ((IObjectContextAdapter)context).ObjectContext;
return objectContext
.ObjectStateManager
.GetObjectStateEntries(relationshipState)
.Where(e => e.IsRelationship)
.Select(
e => Tuple.Create(
objectContext.GetObjectByKey((EntityKey)getValue(e, 0)),
objectContext.GetObjectByKey((EntityKey)getValue(e, 1))));
}
}
Some explanation. Many-to-many relationships are represented in EF as Independent Associations, or IAs. This is because the foreign keys for the relationship are not exposed anywhere in the object model. In the database the FKs are in a join table, and this join table is hidden from the object model.
IAs are tracked in EF using "relationship entries". These are similar to the DbEntityEntry objects you get from the DbContext.Entry except that they represent a relationship between two entities rather than an entity itself. Relationship entries are not exposed in the DbContext API, so you need to drop down to ObjectContext to access them.
A new relationship entry is created when a new relationship between two entities is created, for example by adding an Employee to the Company.Employees collection. This relationship is in the Added state.
Likewise, when a relationship between two entities is removed, then the relationship entry is put into the Deleted state.
This means that to find changed many-to-many relationships (or actually any changed IA) we need to find added and deleted relationship entries. This is what the GetAddedRelationships and GetDeletedRelationships do.
Once we have relationship entries, we need to make sense of them. For this you need to know a piece of insider knowledge. The CurrentValues property of an Added (or Unchanged) relationship entry contains two values which are the EntityKey objects of the entities at either end of the relationship. Likewise, but annoyingly slightly different, the OriginalValues property of a Deleted relationship entry contains the EntityKey objects for the entities at either end of the deleted relationship.
(And, yes, this is horrible. Please don’t blame me—it is from well before my time.)
The CurrentValues/OriginalValues difference is why we pass a delegate into the GetRelationships private method.
Once we have the EntityKey objects we can use GetObjectByKey to get the actual entity instances. We return these as tuples and there you have it.
Here’s some entities, a context, and an initializer, I used to test this. (Note—testing was not extensive.)
public class Company
{
public int Id { get; set; }
public string Name { get; set; }
public virtual ICollection<Employee> Employees { get; set; }
public override string ToString()
{
return "Company " + Name;
}
}
public class Employee
{
public int Id { get; set; }
public string Name { get; set; }
public virtual ICollection<Company> Companies { get; set; }
public override string ToString()
{
return "Employee " + Name;
}
}
public class DataContext : DbContext
{
static DataContext()
{
Database.SetInitializer(new DataContextInitializer());
}
public DbSet<Company> Companies { get; set; }
public DbSet<Employee> Employees { get; set; }
public override int SaveChanges()
{
foreach (var relationship in this.GetAddedRelationships())
{
Console.WriteLine(
"Relationship added between {0} and {1}",
relationship.Item1,
relationship.Item2);
}
foreach (var relationship in this.GetDeletedRelationships())
{
Console.WriteLine(
"Relationship removed between {0} and {1}",
relationship.Item1,
relationship.Item2);
}
return base.SaveChanges();
}
}
public class DataContextInitializer : DropCreateDatabaseAlways<DataContext>
{
protected override void Seed(DataContext context)
{
var newMonics = new Company { Name = "NewMonics", Employees = new List<Employee>() };
var microsoft = new Company { Name = "Microsoft", Employees = new List<Employee>() };
var jim = new Employee { Name = "Jim" };
var arthur = new Employee { Name = "Arthur" };
var rowan = new Employee { Name = "Rowan" };
newMonics.Employees.Add(jim);
newMonics.Employees.Add(arthur);
microsoft.Employees.Add(arthur);
microsoft.Employees.Add(rowan);
context.Companies.Add(newMonics);
context.Companies.Add(microsoft);
}
}
Here’s an example of using it:
using (var context = new DataContext())
{
var microsoft = context.Companies.Single(c => c.Name == "Microsoft");
microsoft.Employees.Add(context.Employees.Single(e => e.Name == "Jim"));
var newMonics = context.Companies.Single(c => c.Name == "NewMonics");
newMonics.Employees.Remove(context.Employees.Single(e => e.Name == "Arthur"));
context.SaveChanges();
}
I cant give you the exact code for your situation, but I can tell you your situation will be simplified ten fold by having a joiner table inbetween Employees and Company just to break up the many to many relationship.