Should I write long signed int or just long if I want a signed int of 32 bits?

Question

So I just wondered if it would be better to be explicit with types e.g. using long signed int for a signed 32 bit int.
Or if it would be better to just not waste your time and just write long from a best practice perspective in C.
Can some compilers on some platforms misinterpret long as long float or long unsigned int?

Answer

You may also consider using one of the fixed width integer types, to be explicit.

Answer

If you want a 32-bit signed integer, you should write int32_t.

Answer

As you can see in the Arithmetic types reference:

long by itself is always signed by default (same goes to long int).
Explicitly adding signed is superfluous and IMHO will only confuse readers.
long by itself always denotes an integral type.
So no compiler will misinterpret it as a floating point type (unless you explicitly add float/double).

Answer

long signed int for a signed 32 bit int

Why do you believe long signed int is going to be 32 bits?

Answer

I'm guessing they mean "to guarantee to get at least 32 bits", since a signed int can have as little as 16 bits.
One can use int_least32_t for that, or int32_t if one is ok assuming the environment has a 32-bit type.

Should I write long signed int or just long if I want a signed int of 32 bits?

Comments on this question (0)

7 answers

Your answer