Type of operations

[Saginsky]

I was wondering if something like
int main()
{
double a = 13.7;
while (a < 5.9){
a = a - 1.1;
}
}

would be tested. I am asking because this means that the assembly generated for the comparison can only be determined by looking at the type of the variable within it i.e. a which is a double. This means that we need to do a double precision comparison. From god-bolt there doesn't seem to be any casting involved here just a double precision version of a while loop. Thanks!

[ymherklotz]

I believe that could be tested, yes.

[JaafarRammal]

@ymherklotz I assume the following is a wrong syntax:

...
float a = 3.4;
while(a < 1.2)
...

and should be:

float a = 3.4;
while(a < 1.2f)

I assume therefore that since in our tests there won't be any casting, floating point constant will always be followed by an f

In other words, we can safely remove the last character in a floating point constant assignment. The following would therefore be invalid:

float a = 3.4;

and should be:

float a = 3.4f;

[tobhil98]

Correct, decimal values without the f should be treated as doubles.

[JaafarRammal]

Another question @tobhil98

Can we assume no hex assignment will be used for floating points? Otherwise if we have something like float a = 0xFFF \\or double do we have to remove the last F to convert it or not? Compilers online don't seem to remove and prioritize the hexadecimal representation in the lexer over the floating point indicator, hence cast it from double into float.

[tobhil98]

Yeah, I would say hex values should be treated as integers, but I'm not entirely sure.

[ymherklotz]

Yeah I would agree, I doubt hex representation will be used to initialise floats and doubles.