What do 0LL or 0x0UL mean?

Question

I am reading the Google Go tutorial and saw this in the constants section:

There are no constants like 0LL or 0x0UL

I tried to do a Google search but all that comes up are instances where people are using these constants but no explanation as to what they mean. 0x is supposed to start a hexadecimal literal but these are not characters that are possible in a hexadecimal number.

Answer 1

These are constants in C and C++. The suffix LL means the constant is of type long long, and UL means unsigned long.

In general, each L or l represents a long and each U or u represents an unsigned. So, e.g.

1uLL

means the constant 1 with type unsigned long long.

This also applies to floating point numbers:

1.0f    // of type 'float'
1.0     // of type 'double'
1.0L    // of type 'long double'

and strings and characters, but they are prefixes:

 'A'   // of type 'char'
L'A'   // of type 'wchar_t'
u'A'   // of type 'char16_t' (C++0x only)
U'A'   // of type 'char32_t' (C++0x only)

---

In C and C++ the integer constants are evaluated using their original type, which can cause bugs due to integer overflow:

long long nanosec_wrong = 1000000000 * 600;
// ^ you'll get '-1295421440' since the constants are of type 'int'
//   which is usually only 32-bit long, not big enough to hold the result.

long long nanosec_correct = 1000000000LL * 600;
// ^ you'll correctly get '600000000000' with this

int secs = 600;
long long nanosec_2 = 1000000000LL * secs;
// ^ use the '1000000000LL' to ensure the multiplication is done as 'long long's.

In Google Go, all integers are evaluated as big integers (no truncation happens),

    var nanosec_correct int64 = 1000000000 * 600

and there is no "usual arithmetic promotion"

    var b int32 = 600
    var a int64 = 1000000000 * b
    // ^ cannot use 1000000000 * b (type int32) as type int64 in assignment

so the suffixes are not necessary.

Answer 2

There are several different basic numeric types, and the letters differentiate them:

0   // normal number is interpreted as int
0L  // ending with 'L' makes it a long
0LL // ending with 'LL' makes it long long
0UL // unsigned long

0.0  // decimal point makes it a double
0.0f // 'f' makes it a float