CAPL invalid type cast of uint32 to dword

Question

Please hear me out on this. Let sysA be a system variable (struct member) of type Uint32 used to store a unique hexadecimal ID four bytes long. In my CAPL script, I'd like to perform a check between this UUID and an integer I re-define by composition of four bytes coming from a message. I went with this approach:

variables
{
    dword uuid;
}

on message myMsg
{
    uuid = (this.byte(0)<<24) | (this.byte(1)<<16)| (this.byte(2)<<8) | this.byte(3);
    if (@sysA == uuid)
    {
        // do something
    }
}

this code triggers error, because of incompatible types dword and Uint32. I then tried casting sysA to dword: invalid type cast. I'm not sure why this would be the case in the CAPL language. According to the docs:

byte (unsigned, 1 Byte)

word (unsigned, 2 Byte)

dword (unsigned, 4 Byte)

int (signed, 2 Byte)

long (signed, 4 Byte)

int64 (signed, 8 Byte)

qword (unsigned, 8 Byte)

I cannot type cast Uint32 system variables to int: int i = 32768 is out of range and int i = 32768LL isn't valid. But dword, in this context, is strictly unsigned int of 4 bytes.

What should I correctly use as type for my uuid, in this scenario, and why?

Answer 1

I tried your code snippet too. You should have a try with defining the proper namespace also when you're trying to use system variables.

variables
{
  dword uuid;
}

on message * 
{
  uuid = ((this.byte(0) << 24) | (this.byte(1) << 16) | (this.byte(2) << 8) | (this.byte(3)));

  if ((dword)@namespace::sysA == uuid)
  {
    write ("Hello World!");
  }
}