当应用于字符算术时，模运算符是什么意思？

Question

symbol = (rand() % ('~' - ' ' + 1)) + ' ';

I have understood that every time that i run the program "symbol" is different. But , why can't I use

symbol = rand();

Answer 1

There's nothing called character arithmatic ^Note (AFAIK) and the modulo operator means it's usual business. What you seem to be confused is about the whole expression.

As per the ASCII table

• ~ has a decimal value of 126
• has decimal value of 32.

The output of rand() will be modulo-valued by the range, over the space , so as to produce a printable representation .

To break it down,

• ('~' - ' ' + 1) == 126-32 +1 == 95
• (rand() % ('~' - ' ' + 1)) produces a random number between 0-94 .
• (rand() % ('~' - ' ' + 1)) + ' ' yields a random number between 32-126 , which is the printable range.

why I can't use symbol = rand();

Well, there's nothing wrong in that statement, but looks like the purpose is to produce a printable symbol value.

---

Note:

Quoting C11 , chapter §6.4.4.4

An integer character constant is a sequence of one or more multibyte characters enclosed in single-quotes, as in 'x'. [...]

and,

An integer character constant has type int . [...]

so, it's integer arithmetic, all the way.

Answer 2

Ascii of

• ~ is 126

• space is 32

So symbol = (rand() % ('~' - ' ' + 1)) + ' '; is equivalent to:

symbol = (rand() % (126 - 32 + 1)) + 32;

which gives you a random number between 32 and 126

---

C99 7.20.2.1

The rand function computes a sequence of pseudo-random integers in the range 0 to RAND_MAX .

The value of the RAND_MAX macro shall be at least 32767 .

You can use symbol=rand() if you want a range between 0 and min(max(32767,RAND_MAX),TYPE_SYMBOL_MAX) and as long as the type of symbol permit that .

Answer 3

Character constants in C are integers. So the following:

('~' - ' ' + 1)

Calculates the offset of the encoding for '~' from the encoding for ' ' and add one to it. It's how many different character encodings lie between them, let's call that number m .

(rand() % ('~' - ' ' + 1))

Takes a pseudo-random number and maps it to the range [0, m) , let's call that number n .

(rand() % ('~' - ' ' + 1)) + ' '

This simply takes n , treating it as an offset, and adds it to the encoding of ' ' , thus getting an encoding of a symbol. The symbol will always be one of a certain set.

---

symbol = rand();

The above will convert an integer, that possibly lies way outside the range of char , into some a character value. That will be done in some implementation defined manner. That's not the most reliable way to get correct results, to say the least.