sizeof operator

[JS]

I read in K&R page 204 that sizeof use on a char returns 1. But when I write
the following I get 4!

int main(void){
printf("%d\n",s izeof('g'));

}

Was it not supposed to return 1?

[Neil Kurzman]

JS wrote:

"Ben Pfaff" <bl*@cs.stanfor d.edu> skrev i en meddelelse
news:87******** ****@benpfaff.o rg...

"JS" <[e-mail] m> writes:

I read in K&R page 204 that sizeof use on a char returns 1. But when I write the following I get 4!

printf("%d\n",s izeof('g'));

In C, character constants have type `int', so sizeof 'g' is equal
to sizeof(int).

Why do they write char if they mean int?

lookup Integer Promotion

[Steven K. Mariner]

Perhaps.

Manual examination noted that a '-' got translated to '--' somehow
during my transfering of it to a text file. When I switched from MinGW
to an old bash shell version of gcc, I got a more intelligent error
about trying to modify an lvalue. That plus a comment below by Jens
helped me find the problem, and when I fixed it so my text file more
closely matched your source, it compiled just fine.

I thought it would simply be a cute display message, but I was dying to
see the message without having to completely reverse-engineer it.

And I was right. It's a cute little display from a fairly basic cipher
Nice.

[Steven K. Mariner]

Strange character was inserted (as best I can deduce) by cut-n-paste
processes. In the final analysis, a single minus sign translated
somehow to a double (probably some kind of auto-adjust courtesy of the
Microsoft products I'm using in this environment [*gag*]).

Corrected it by hand and the program now compiles and functions
correctly.

Thanks for the note.

[Old Wolf]

Neil Kurzman wrote:

JS wrote:

"Ben Pfaff" <bl*@cs.stanfor d.edu> skrev:

"JS" <[e-mail] m> writes:

I read in K&R page 204 that sizeof use on a char returns 1.
But when I write the following I get 4!

printf("%d\n",s izeof('g'));

In C, character constants have type `int', so sizeof 'g' is equal
to sizeof(int).

Why do they write char if they mean int?

lookup Integer Promotion

This has nothing to do with integer promotion. The argument
to 'sizeof' does not undergo any promotions or conversions.

[Old Wolf]

Jack Klein wrote:

Let's just assume ASCII for a moment.

Consider that the only differences between 'A', 65, 0x41,
and 081 exist syntactically in the way they are parsed

081 ?

[Keith Thompson]

"Old Wolf" <ol*****@inspir e.net.nz> writes:

Neil Kurzman wrote:

JS wrote:

"Ben Pfaff" <bl*@cs.stanfor d.edu> skrev:
"JS" <[e-mail] m> writes:

> I read in K&R page 204 that sizeof use on a char returns 1.
> But when I write the following I get 4!
>
> printf("%d\n",s izeof('g'));

In C, character constants have type `int', so sizeof 'g' is equal
to sizeof(int).

Why do they write char if they mean int?

lookup Integer Promotion

This has nothing to do with integer promotion. The argument
to 'sizeof' does not undergo any promotions or conversions.

It's relevant, but only very indirectly. Since expressions of type
char are almost always promoted to int, there would be little point
(in C) in making character constants be of type char rather than of
type int.

This is part of the rationale for why the standard is the way it is;
it's not something from which you can logically infer that character
constants are of type int. The standard could just as easily have
said that character constants are of type char; they would then be
immediately promoted to int, so it would make a difference only when
the character constant is the operand of sizeof.

(Note that if plain char were unsigned, and sizeof(int)==1 (which
requires CHAR_BIT>=16), character constants of type char would promote
to unsigned int rather than int, but that doesn't apply to most
implementations .)

The real answer to "Why are character constants of type int rather
than char?" is "Because the standard says so". *Why* the standard
says so is another question.

--
Keith Thompson (The_Other_Keit h) ks***@mib.org <http://www.ghoti.net/~kst>
San Diego Supercomputer Center <*> <http://users.sdsc.edu/~kst>
We must do something. This is something. Therefore, we must do this.

[Joona I Palaste]

Old Wolf <ol*****@inspir e.net.nz> scribbled the following:

Jack Klein wrote:

Let's just assume ASCII for a moment.

Consider that the only differences between 'A', 65, 0x41,
and 081 exist syntactically in the way they are parsed

081 ?

Shouldn't that be 0101?

--
/-- Joona Palaste (pa*****@cc.hel sinki.fi) ------------- Finland --------\
\-------------------------------------------------------- rules! --------/
"Bad things only happen to scoundrels."
- Moominmamma