Difference in object construction using “X x(42)” and “X x = 42”?

Question

Let's say we have class X with defined constructor X(int value).

Is this semantically equivalent or not?

X x = 42;
X x(42);

I believe the difference will appear only if we add explicit keyword to constructor of X. Otherwise compiler will reduce expression X x = 42; to X x(42);

Please, correct me if I'm wrong.

Answer 1

The form

X x = 42;

requires that the constructor be non-explicit and that there be an accessible copy-constructor. The implementation is allowed to construct a temporary and copy it over, but no implementation I know of does that.

Answer 2

Semantically the two operations are different per the language in the specification, but the effective results after compilation are the same in the following circumstances:

1. There is a single argument constructor for X that can take an integral-type and the argument type is not a non-const l-value reference
2. There is a single argument constructor for X where an integral-type can be converted to the argument type, and where the argument is again not a non-const l-value reference
3. In the case of #2, the constructor is not declared as explicit