question about casting

Code:

---


char c = 0xff;
unsigned int a;

a = (unsigned int) c;

---

How come in the above cast "a" becomes ULONG_MAX or 0xffffffff? Since I am
casting to unsigned, shouldn't it cast to 0x000000ff, or would I have to declare
c as unsigned char in order for that to happen?

thanks :)

>How come in the above cast "a" becomes ULONG_MAX or 0xffffffff?
Sign extension. Your char type is signed by default, so assigning it 0xff makes it negative. Casting to unsigned int then sign extends the value to 0xffffffff.

>or would I have to declare c as unsigned char in order for that to happen?
Try it and see.

don't use unsigned, it just creates trobules.

in some cases

Code:

---

unsigned int a = 5;
int b = 5;

---

a==b is not true.

That's just weird advice that has nothing to do with the question, and moreover, it is indeed safe compare signed stuff with signed stuff, and compare unsigned stuff with unsigned stuff without changing the logical answer: using the same signedness on both sides of a comparison makes the bit pattern deterministic.

Quote:

---

Originally Posted by Govalant

a==b is not true.

---

I'd like to know more about that, or a link or something.

Quote:

---

Originally Posted by robwhit

I'd like to know more about that, or a link or something.

---

http://en.wikipedia.org/wiki/Signed_...epresentations

Quote:

---

Originally Posted by Govalant

don't use unsigned, it just creates trobules.

in some cases

Code:

---

unsigned int a = 5;
int b = 5;

---

a==b is not true.

---

Your compiler is broken. It's time to get a new one.

Quote:

---

Originally Posted by swoopy

Your compiler is broken. It's time to get a new one.

---

Yep. What he says is in direct contradiction to ISO/IEC 9899 section 6.3 (2) and section 6.3.1.3 (1).

Quote:

---

Originally Posted by brewbuck

Yep. What he says is in direct contradiction to ISO/IEC 9899 section 6.3 (2) and section 6.3.1.3 (1).

---

I think I'll read through that section tomorrow just for fun.