Can someone give me an example of the following

The following comment was made by one of the amazon dot com engineers on my schools internal bbs system.

Quote:

---

The return value of malloc compares equal to zero upon failure.

The ctype functions return nonzero (true)
if and only if the value of the argument conforms to that
in the description of the function.

If you want the return value
to be able to indicate which mode of failure occured,
then that's when zero is the value to return for success.

---

Which ctype functions is he talking about?

Perhaps isalpha() and company?

He is talking about the family of functions that are declared in the ctype.h header file.

Ohh..... okay. Thanks.

For example if malloc returns 0 (or better, NULL) it failed. If isdigit returns 0, it failed (that means the argument is not a digit).

> malloc returns 0
Only in C++ :-).

Quote:

---

Originally Posted by ISO/IEC 9899 - 7.17

The macros are NULL ... which expands to an implementation-defined null pointer constant

---

Quote:

---

Originally Posted by zacs7

Only in C++

---

No, the behaviour is the same for C as well, though it would be more accurate to talk about a null pointer being returned.

EDIT:
Oh, you were trying to be pedantic. Unfortunately for you, 0 is a null pointer constant in both C and C++. However, what is returned by malloc() would be a null pointer of type void*, if it is not a pointer to allocated memory.

My point was, in C, NULL != 0 (well, it doesn't have to be).

I've heard of such a system where NULL was not at "address 0". Myth perhaps, but I was told it made testing rather hard.

[edit]
Hmm, it seems
ISO/IEC 9899 - 6.3.2.3 Pointers says: "An integer constant expression with the value 0, or such an expression cast to type void *, is called a null pointer constant."

So perhaps I was wrong, sorry :-). I guess you shouldn't listen to other people without first reading the standard for yourself ;)
[/edit]

NULL = 0 (with an implicit cast in C). Often it is 0.

Quote:

---

Originally Posted by zacs7

My point was, in C, NULL != 0 (well, it doesn't have to be).

---

Well, according to the Standard it does:

Quote:

---

Originally Posted by ISO/IEC 9899

6.3.2.3
3 An integer constant expression with the value 0, or such an expression cast to type void *, is called a null pointer constant. If a null pointer constant is converted to a pointer type, the resulting pointer, called a null pointer, is guaranteed to compare unequal to a pointer to any object or function.

---

Since NULL is a macro that expands to the null pointer constant, NULL must be 0. Contra zacs7 and brafil, it is not implementation dependent.

[edit: laserlight caught my failure to read zac7's entire post. At least we are obviously on "the same page" with this, ha ha.]

Quote:

---

Originally Posted by MK27

Well, according to the Standard it does:

---

To be fair, zacs7 quoted the same part of the C standard in the edit to the post that you quoted.

Quote:

---

Originally Posted by MK27

Since NULL is a macro that expands to the null pointer constant, NULL must be 0. Contra zacs7 and brafil, it is not implementation dependent.

---

It is implementation dependent as to exactly which null pointer constant NULL expands to (i.e., the type can differ). zacs7 cited the correct part of the C standard that states as such.

The standard guarantees that NULL doesn't point to any object or function and is not a legal pointer. That means undefined behavior when deferencing NULL.