When you declare a variable in Entity Framework model, migration translates the C# type to SQL Server type. Here is a table describing how each C# type is implemented in SQL.
| C# primitive type | C# type in System namespace | SQL Server type |
| byte | Byte | tinyint |
| short | Int16 | smallint |
| int | Int32 | int |
| long | Int64 | bigint |
| float | Single | real |
| double | Double | float |
| decimal | Decimal | decimal(18,2) |
| bool | Boolean | bit |
| DateTime | datetime | |
| char | Char | |
| string | String | nvarchar(max) null |
Just two interesting things:
- char type is not translated to any SQL type;
- string type is the only type that will be nullable, the others are not.
And some tips:
- to make a field NULL, for example int, you should use a nullable type int? instead;
- to make a string field NOT NULL, place a [Required] attribute before declaration in EF model;
- if your string field won’t have an infinite length, you can (and I recommend to) place a [StringLength(n)] attribute where n can be up to 4000.
[UPDATE 2016-11-17]
Unsigned integer types and sbyte are not translated too. See here Not All .NET Types Can Be Used In Entity Framework