Does bool return an int?

Question

I know that for the Arduino (and C in general) that for comparison sake, "false" is 0, while any non-zero integer is "true". The question that I have is does the bool data type "return" a concrete integer for "true" or "false"?

When I tried Serial.print(true == 1), it printed 1 (which I assumed to be "true"), but when I tried Serial.print(true == 2), it printed 0 (which I assumed to be "false"). When I tried Serial.print(false == 1), it printed 0 and Serial.print(false == 0) printed a 1.

I didn't know if "true" always evaluated to 1 and "false" always evaluated to 0, or if this is not supposed to happen.

• Jot: Never use if(flag==1), don't even use if(flag==true), but use if(flag), when the variable flag is a bool or boolean. The bool is part of the c++ language and the boolean is made up by arduino. The Serial.print function prints integers and floats, but sadly not "true" or "false" for a bool, so it uses the integer value and prints "1" or "0".
• DuluthIsSuperior: @Jot, I was mainly seeing if it would be the same because I have function foo(int tf) where tf is either a 1 or a 0, and merely wanted to see if tf == true would be the same as tf == 1. But if the printer is translating true to 1, then it sounds like what I'm proposing is a bad idea.

Answer 1

If you look at Arduino.h, you'll see true, false and bool defined as:

#define true 0x1
#define false 0x0
typedef uint8_t boolean;

Using a simple test sketch, it appears that comparing true or false to anything besides a 1 or 0 does not "work".

void setup(){
 Serial.begin(9600);
 if(false == 0){Serial.println("false=0");}
 if(false == 1){Serial.println("false=1");}
 if(false == 2){Serial.println("false=2");}
 if(false == -1){Serial.println("false=-1");}
 if(true == 0){Serial.println("true=0");}
 if(true == 1){Serial.println("true=1");}
 if(true == 2){Serial.println("true=2");}
 if(true == -1){Serial.println("true=-1");}
}
void loop(){}

Answer 2

true is indeed 1, and false is 0.

This follows your tests:

(true == 1) -> prints 1, because 1 == 1
(true == 2) -> prints 0, because 1 != 2
(false == 1) -> prints 0, because 0 != 1

In C true and false are defines and a bool/boolean is a typedef (mostly to an int, unsigned int or unsigned char), where normally false is defined to 0 and true to 1 OR 255. So you cannot rely on a specific value.

In C++ the values of true and false are resp. 1 and 0, when casted to an int.

Answer 3

A 'bool' is defined data type based on (and identical to) an int. It's reason for existence is documentary: it's name describes how you intend to use it. There is no other difference.

'true' and 'false' are the constants 1 and 0, respectively. If you assign 'true' to a bool or any other integer type, it will get the value '1'. If you make a comparison or any other logical test (one that evaluates to true or false), that test results in a numerical value, either 1 or 0.

However you can evaluate (i.e. test) any integral value logically (true or false): if it is equal to zero the result will be false (0); otherwise the result will be true (1) - no matter what non-zero value the expression itself has. That is because the logical operators, '==', '!=', '>', etc., are (similar to) functions which can only return one of two values 'true' and 'false', regardless of the values of their inputs. That is another meaning of the bool datatype: when it is applied to a function, it implies to the compiler and to the reader that this function returns either true or false.