How to use unsigned int / long types with Entity Framework?
Update Entity Framework Core Feb 2021
EF Core 3.1: EF Core now supports long and ulong types. Using code first, the long or ulong is mapped to EF Core's new 'Decimal Type'
public ulong MyULong{ get; set; } //==> decimal(20, 0)
A ulong results in a decimal being defined in the database with 20 digits and 0 digits to the right of the decimal point, which is sufficient to store a 64 bit ulong.
EF 5: Thankyou to @Tomasz for noting that in EF 5 and 6 the ulong is mapped to a BigInt, rather than the Decimal type as per my original answer, now under the heading "EF Core 3.1" above
Update Feb 2021
Apparently EF Core now supports ulong
-- see @JimbobTheSailor's answer below.
Older Entity Framework versions:
Turns out that Entity Framework does not support unsigned
data types. For uint
columns, one could just store the value in a signed data type with a larger range (that is, a long
). What about ulong
columns? The common solution couldn't work for me because there is no EF-supported signed data type that can hold a ulong
without overflowing.
After a bit of thinking, I figured out a simple solution to this problem: just store the data in the supported long
type and cast it to ulong
when accessed. You might be thinking: "But wait, ulong's max value > long's max value!" You can still store the bytes of a ulong in a long and then cast it back to ulong when you need it, since both have 8 bytes. This will allow you to save a ulong variable to a database through EF.
// Avoid modifying the following directly.
// Used as a database column only.
public long __MyVariable { get; set; }
// Access/modify this variable instead.
// Tell EF not to map this field to a Db table
[NotMapped]
public ulong MyVariable
{
get
{
unchecked
{
return (ulong)__MyVariable;
}
}
set
{
unchecked
{
__MyVariable = (long)value;
}
}
}
The casting is unchecked
to prevent overflow exceptions.
Hope this helps someone.