2

I am converting old VB6 code and am having an issue with these constants:

VB6 code:

Private Const GENERIC_WRITE = &H40000000
Private Const GENERIC_READ = &H80000000

C# code:

private const int GENERIC_WRITE = 0x40000000;
private const uint GENERIC_READ = 0x80000000;

Why does the second line have to be uint? If I do it as int, it gives me an error?

StayOnTarget
  • 11,743
  • 10
  • 52
  • 81
KHerron
  • 23
  • 2

3 Answers3

4

The correct C# translation is:

const int GENERIC_READ = unchecked((int)0x80000000);

In VB, 'GENERIC_READ' is implicitly an Integer with a negative value. The only way to reproduce this in C# is to cast the unsigned integer literal to an int in an unchecked context.

The source of the confusion is a historical difference between VB and C# regarding hex literals: 0x80000000 == 2147483648 in C# but not in VB.NET

In the absence of a type suffix on the literal, hex literals in VB.NET are either of type Integer or Long, never unsigned types.

Dave Doknjas
  • 6,394
  • 1
  • 15
  • 28
  • Dave, why is unchecked needed? – StayOnTarget May 31 '19 at 20:58
  • @DaveInCaz, the documentation is at your fingertips, no need to ask ;-) https://learn.microsoft.com/en-us/dotnet/csharp/language-reference/keywords/unchecked –  May 31 '19 at 21:00
  • @DaveInCaz: Since 0x80000000 is an unsigned integer literal in C#, the only way it casts to an int is by using unchecked. VB is doing this behind the scenes. The value being assigned to the constant must be -2147483648. – Dave Doknjas May 31 '19 at 21:02
  • And that will result in the same binary value being stored in the constant? Meaning for instance, an equality test would be true between the C# and VB6 constants? – StayOnTarget May 31 '19 at 21:03
  • I know that when I examine the 2 values in VB.NET and C#, they are both assigning -2147483648 to the constant. VB6 is too far in the past for me to look at that. – Dave Doknjas May 31 '19 at 21:09
2

The maximum value of Int32 is 2,147,483,647, and the value of 0x80000000 is 2,147,483,648, or Int32.MaxValue + 1.

Using a uint (or long) gives you a type that's large enough to hold the value you are assigning.

Jonathon Chase
  • 9,396
  • 21
  • 39
0

The primitive data types prefixed with "u" are unsigned versions with the same bit sizes. Effectively, this means they cannot store negative numbers, but on the other hand they can store positive numbers twice as large as their signed counterparts.

Daniil T.
  • 1,145
  • 2
  • 13
  • 33