5

Possible Duplicate:
Negative ASCII value 

int main() {
    char b = 8-'3';
    printf("%c\n",b);

    return 0;
}

I run this program and I get a sign which looks like a question mark (?).

My question to you is why is it prints that and not printing nothing, beacause as far as I know the value of b by the ASCII table is minus 43 which is not exist.

by the way, when I compile this code: 

int main() {
    char b = -16;
    printf("%c\n",b);

    return 0; 
}

I get nothing.

Community
  • 1
  • 1
wantToLearn
  • 481
  • 4
  • 7
  • 19
  • 2
    You may want to look at http://stackoverflow.com/questions/4690415/negative-ascii-value – CCoder Oct 31 '12 at 15:22
  • I do get a non-printable character for the second case also – CCoder Oct 31 '12 at 15:23
  • how can I know which sign will appear when I give him a value ? – wantToLearn Oct 31 '12 at 15:29
  • I would say that really depends on the compiler. Please read up on http://stackoverflow.com/questions/4690415/negative-ascii-value. and you can refer extended ASCII table here - http://www.ascii-code.com/ – CCoder Oct 31 '12 at 15:33
  • char may be signed or unsigned, depending on the implementation. If a character is represented as a negative number, that doesn't mean it does not exist. There are two other types, signed char, and unsigned char, and those are of course signed and unsigned, respectively. – Mikkel K. Oct 31 '12 at 15:38

6 Answers6

5

Here's the behavior as specified in the C 2011 standard

7.21.6.1 The fprintf function

...
8 The conversion specifiers and their meanings are:
...
c If no l length modifier is present, the int argument is converted to an unsigned char, and the resulting character is written.

If an l length modifier is present, the wint_t argument is converted as if by an ls conversion specification with no precision and an argument that points to the initial element of a two-element array of wchar_t, the first element containing the wint_t argument to the lc conversion specification and the second a null wide character.

That -43 is being converted to an unsigned value (213), so it's printing an extended ASCII character.

John Bode
  • 119,563
  • 19
  • 122
  • 198
  • And since the signedness of `char` is implementation-defined, we don't know if it is signed or unsigned for this particular compiler. It is "something" and then converted to unsigned char when passed to printf. – Lundin Oct 31 '12 at 16:07
3

That's because -16 is the same as hex 0xf0 which is the same as 240. In my character set (ISO 8859-1) that character is LATIN SMALL LETTER ETH or 'ð'.

Some programmer dude
  • 400,186
  • 35
  • 402
  • 621
2

There is no such thing as a "negative ASCII" value. ASCII defines character and control codes for values from 0 to 127.

Most systems don't use plain ASCII. They use some other character set. Almost all other characters sets match ASCII for the values from 0 to 127, but they define additional characters and control codes for higher values.

In some C and C++ implementations char corresponds to a signed char that can take values from -127 to +127 (and in most cases from -128 to +127). In other implementations char is an unsigned value from 0 to 255.

If you have a signed char with a negative value and try to print it, the library will treat that as an unsigned char (e.g., -1 becomes 255), and it will print whichever character in the platform's character set corresponds to that value. If there is no character assigned to a particular value, the library may choose to display some default placeholder character, like a question mark or a box.

Adrian McCarthy
  • 45,555
  • 16
  • 123
  • 175
0

characters are internally interpreted as "unsigned" while printing. Hence, it will be 255 - 16 = 239th character. The ascii value with the order 239 is printed on the screen console. Similarly -43 is order 212.

Aniket Inge
  • 25,375
  • 5
  • 50
  • 78
0

The %c formatting will always print a character, so presumably it just takes the bits of your char and treats it as an unsigned value, which ends up being something with the highest bit set. Your terminal may display these characters as blank or as placeholders with question marks, but mine shows Õ and ð…

If you want to print nothing when the value is negative, you need to check for it (and use either a larger data type, such as int so that you can fit the non-char values along with all possible chars, or signed char if you wish to be sure that you are getting a signed type; a plain char may be either signed or unsigned).

Arkku
  • 41,011
  • 10
  • 62
  • 84
0

You can refer the ASCII table. And all numbers should map to (256+(value))%256 (This is the way overflows are handled). You should be able to find which symbol is printed for which number.

CCoder
  • 2,305
  • 19
  • 41
  • ASCII only defines values from 0 to 127. There are (many) other character sets that define additional characters for values above that range and match the ASCII definitions for the lower range. Those other character sets are sometimes called "extended ASCII", but there is no single definition of "extended ASCII". The mappings for the values from 128-255 depend on which character set (or code page) is in use. – Adrian McCarthy Oct 31 '12 at 16:40
  • The link I shared does contain 'a version' of the extended ASCII table. And the logic I proposed for deciding which is the character works fine for me. – CCoder Oct 31 '12 at 19:28