ef-code-first

问题 I am saving a Cart object to the database that has a nullable datetime. This is the error I get: The conversion of a datetime2 data type to a datetime data type resulted in an out-of-range value. There are quite a few stackoverflow posts documenting fixes to this problem. However, when code first is creating the database it will create the field as a DateTime (allow nulls). But for some reason, code first tries to insert using a DateTime2 field. I am wondering why EF creates the field one way

问题 I am encountering some performance problems with my Entity Framework Code-First queries and I believe that precompilation may be the answer. If I were using "normal" Entity Framework, I would simply use the CompiledQuery.Compile method to precomiple my queries. But since I have a DbContext and not an ObjectContext, I can't get this to work. I do realize that DbContext is an IObjectContextAdapter, which gives me access to the ObjectContext, but I cannot find the method that lets me get an

问题 Consider: Class Client Class Project Class Ticket Class Reply Clients have a sub collection of projects, projects have a sub collection of tickets and tickets have a sub collection of replies. var data = ctx.Set<Ticket>().Include(p => p.Client). Select(p => new { Ticket = p, LastReplyDate = p.Replies.Max(q => q.DateCreated)}); Doesn't work. Neither project nor client are loaded when selecting data this way. I know how to make it work. My question is why doesn't it work like this? 回答1: As

问题 I'd like to run actual integration tests of my EF4.1 repositories against an in-memory database a la ayende's nhibernate version. I have a code first model, against a legacy database (old table and column names need mapping to my entites using code configurations). I'd like to be able to use Sqlite (or other) to: Generate an in-memory database from my model Create a DBContext for my model with this in-memory database I have already in place IoC/DI of a IDBContextFactory which gets constructed

问题 In a legacy app, most string properties can't be null and need to have a default value of string.empty. I know it's possible to do this with migrations, but I'm looking for a way to do this using the fluent configuration interface: protected override void OnModelCreating(DbModelBuilder modelBuilder) { modelBuilder.Properties<string>().Configure(c => { c.HasMaxLength(255); if (!c.ClrPropertyInfo.IsDefined(typeof (NullableAttribute), false)) { c.IsRequired(); // I want to set a default value

问题 I am using CodeFirst approach and struck with an issue where I need to convert DbSet to ObjectQuery. This is what I did for conversion. ObjectContext objectContext = ((IObjectContextAdapter)db).ObjectContext; ObjectSet<Request> objectSet = objectContext.CreateObjectSet<Request>(); where db is the context inheriting from DbContext and Request is class. So, when I try to call the method that expects ObjectQuery as ObjectQueryMethod(objectSet), it throws the following error. "Type of conditional

问题 Following on the heels of my other question about mocking DbContext.Set I've got another question about mocking EF Code First. I now have a method for my update that looks like: if (entity == null) throw new ArgumentNullException("entity"); Context.GetIDbSet<T>().Attach(entity); Context.Entry(entity).State = EntityState.Modified; Context.CommitChanges(); return entity; Context is an interface of my own DbContext. The problem I'm running in to is, how do I handle the Context.Entry(entity)

问题 Following on the heels of my other question about mocking DbContext.Set I've got another question about mocking EF Code First. I now have a method for my update that looks like: if (entity == null) throw new ArgumentNullException("entity"); Context.GetIDbSet<T>().Attach(entity); Context.Entry(entity).State = EntityState.Modified; Context.CommitChanges(); return entity; Context is an interface of my own DbContext. The problem I'm running in to is, how do I handle the Context.Entry(entity)

问题 a.k.a How can we create multiple identity columns in Code First? Because of clustering performance, a common recommendation is to use an autoincremented integer column instead of a GUID created with newid() . In order to declare a column as autoincrement, you have to specify it with the Annotation [DatabaseGenerated(DatabaseGeneratedOption.Identity)] . But, you can only have one identity in a table. So starting with a base model like: public abstract class ModelBase { // the primary key

问题 How do I define a database view using Entity Framework 4 Code-First? I can't find anything about this anywhere! 回答1: That's because you cannot define database view using code-first approach. Database view is database construct which uses SQL Query on top of existing tables / functions. You can't define such constructs using code first. If you want view you must create it manually by executing CREATE VIEW SQL script for example in custom initializer - it will be similar like this answer. Just