What's the purpose of the char data type in C++? Why not just use strings?

Question

I don't see the point of having the use of single quotes reserved for single characters. So, when is this used? i.e char a = 'a'; Coming from Javascript and PHP, I'm used to being able to use single quotes for entire strings, but after learning that single quotes are reserved for characters in C++, I'm curious to know why? Why would you need a single character?

Answer 1

PHP and JavaScript are languages that operate at quite a high level. This means that the basic types are essentially just a few different types, whose implementation is hidden inside a set of functions in the actual script engine.

C and C++, as well as most other low level languages expose more of "how the machine works". A string, in C, is a sequence of characters. If you want to deal with strings, you need to be able to deal with their components, which is char. A single character becomes useful when you want to build strings, compare the contents of strings, etc. Naturally, for normal string operations in C++, you'd use std::string, and then, like in script languages, most aspects of how the string is actually represented is hidden inside the std::string class implementation, so you don't really need to care about it. But if you were to "look inside" a std::string, it would somewhere sooner or later, become a char *, which is a pointer to a piece of memory that contains a sequence of characters, individual char elements.

One could look at it like going from having "ready made big lumps of Lego" to having only small pieces to work with. You can still build the same things, but it requires more pieces, and requires a bit more effort to construct. What you win is flexibility and speed. A char is really easy to deal with for the processor, where a single character in PHP is still represented as a string - it just happens to be one element long. As such, there is extra overhead in keeping track of this one character string, where it's stored, how long it is, etc, because the functionality in the language doesn't make any distinction between a single character and a string of a megabyte.

The purpose of C, and to a large degree also C++, is to closely represent the hardware. So your basic types are much closer to what the actual hardware representation is, and this is something you will need to learn more about if you are going to understand C and C++ well. Unfortunately, to cover ALL of that would be far beyond a single answer in SO. You will need to get yourself a good C and/or C++ book.

Answer 2

To run things faster and efficient you must learn how to use less space needed. When you have an 'a', this is actually a number (see ASCII table), but when you got "a" it is an array of 2 characters {'a','\0'}. The zero is to know when your string ends because the computer is not sure when the string ends. Do you want to add a length property like in javascript, to know directly the string's length? You use more space that may not be needed. Somehow you have to distinguish these two values to run efficient code. Learning C/C++ first you actually learn how things work on the low level of your computer and understand more than by learning php/javascript/ruby/python first. C++ is more customizable than higher level programming languages.

Answer 3

Javascript and PHP are scripting languages while C++ (and especially its predecessor C) is quite low-level native programming language where you have to consider how your variables are stored in memory.

char a = 'a'; creates a 8-bit-long numeric variable that can hold character value (ASCII code) and put the value of the character a into it. So char a = 97; does the same work.

const char* s = "a"; creates a null terminated string which is an array with two elements: the value of character a and the terminating 0 character (just number 0 or the character '\0'). The * means we create a pointer to the array. Its type is const char because it contains string literal which is constant. We could create an identical array using const char s[2] = { 97, 0 }; or const char s[2] = { 'a', '\0' };.

---

By the way, single quotes are not reserved exclusively for single characters. You can put a few characters into single quotes. See What do single quotes do in C++ when used on multiple characters?.

Answer 4

The language C++ inherited scalar types from the language C. So you should get a good book about C++ or C to get the details.

The type char is a numeric type. It's an integer that can old a character. It's usually signed but can be overwritten with the signed or unsigned prefix. Additionally the signedness can be configured at the compiler. The 'A' literal defines a number that is identical to the code of the character A.

The "a" string is an array of char's and has a zero termination. In the example you have two bytes 'a','\0'. When you use the literal the compiler passes the address as a pointer to the array. This pointer can be assigned to pointer variable

char *s = "A";

or passed to a function

foo("A");

Additionally there are all the pointer arithmentics possible that you can grasp when you got the meaning of a char array.

Edit Both the number literal 'a' and the char array "A" are const objects. It's obvious that you can't assign anything to a number like

'a' = 23; // wrong!

When you assigned the literal to variable you can change the variable later tough.

But when you stored the pointer in a pointer variable it's illegal and you get undefined behavior when you try to modify the char array:

char *s = "A";
*s = 'B';  // try to change the first byte of the char array, causes undefined behavior.

To express this it's good style to use a pointer to const char variable:

const char *s = "A";
*s = 'B'; // compiler diagnostic says: not allowed