The program below produces
output - 1 4 4
#include<stdio.h>
void main()
{
char ch;
ch='A';
printf("%d %d %d\n",sizeof(ch),sizeof('A'),sizeof(3.2f));
}
Why is the size of character constant is 4 bytes ?
Asked
Active
Viewed 3,706 times
1
Kanwaljit Singh
- 4,339
- 2
- 18
- 21
Anjaneyulu
- 434
- 5
- 21
-
3Because `int` on your platform has 4 bytes. – too honest for this site Jan 09 '17 at 13:36
-
1C character literals are `sizeof(int)` large. C++ character literals are `sizeof(char)` large. It's not constant `4`, it's `sizeof(int)`. – cadaniluk Jan 09 '17 at 13:37
2 Answers
4
Because according to the C standard the type of a character constant is int, and not char.
So in effect, this is the sizeof(int) on your platform.
StoryTeller - Unslander Monica
- 165,132
- 21
- 377
- 458
-
1Detail: According to the C standard, there are 2 kinds of literals: _string literal_ and _compound literal_. `'A'` is not identified as a literal but a _character constant_. – chux - Reinstate Monica Jan 09 '17 at 13:51
-
@chux - Thanks. This sort of inconsistency in the standards use of English is killing me sometimes. – StoryTeller - Unslander Monica Jan 09 '17 at 13:54
-
1Maybe it is not so much an English inconsistency as C is differentiating a _literal_, which can have an address `printf("%p\n", (void*) &"abc");`, from a _constant_, which cannot: `printf("%p\n", (void*) & 'a');` – chux - Reinstate Monica Jan 09 '17 at 14:45
-
@chux - Now that you spell it out, I suppose there is room to differentiate. – StoryTeller - Unslander Monica Jan 09 '17 at 14:46
2
ch is char type so 1 byte.
'A' is int type so 4 bytes. Because in C the character constant is an int type.
Last is float value so 4 bytes.
These values according to the machine you are using.
Edit -
The range of int and float depends on the machine you are using, 16 bit int is as common as 32 bit int.
Kanwaljit Singh
- 4,339
- 2
- 18
- 21
-
-
5Nothing is "converted". A character literal *is of type* `int`. – StoryTeller - Unslander Monica Jan 09 '17 at 13:42
-
3I'm not sure about the use of the word "converted" here. Wouldn't it be more accurate to say that it *is* an `int` (and when you do `ch = 'A';` the `int` is converted to `char`)? – sepp2k Jan 09 '17 at 13:42
-
-
1You don't make clear that the standard types lack a fixed size except for `char`, which is always one byte. "so" isn't any better because you cannot draw the conclusion "`int`, so it's four bytes." – cadaniluk Jan 09 '17 at 13:43
-
There is no guarantee `int` has 4 bytes, nor for `float`. – too honest for this site Jan 09 '17 at 14:38
-