Making sense of the manual

[Anon1010]

I've come to the formal explanation of literals, and the manual reads as follows.

All literal values have a data type, but this raises a question. As you know, there are several different types of integers, such as int, short int, unsigned long int. There are also different types of floating point. The question is: How does the compiler determine the type of a literal. For example, is 123.23 a float or a double.


Then it gives an answer. The problem I have is that I thought the compiler was told the type, whether it is a float or a double, when the variable was declared.

Can someone tell me why this is not the case?
(can provide more of passage from book if needed)

[Nisheeth]

Literals aren't declared as variables.

For instance

int a;
a = 5;

a = a/2;             //Here 2 is a literal. We haven't declared a variable for it. So the compiler must recognize its Data type itself. 

I hope that clears it!

[Anon1010]

Yep; thanks

[Anon1010]

Nisheeth; I've just come across

float f=1.2F;

Why are both float and F used?

[anonymous23323124]

Because 1.2 is a double literal. 1.2F is a float literal.

Without the F, 1.2 is a double and it will be cast by the compiler to a float. This should generate a warning, and could lead to loss of data (it probably won't in this case, but it the numbers are much large or the precision greater then it will happen).