C++ type casting [duplicate]

Question

Possible Duplicate:
When should static_cast, dynamic_cast and reinterpret_cast be used?

Until a few days ago, I've always used C style type casting in C++ because it seemed to work good. I recently found out that using C in C++ is very bad..

I've never really used C++ casting before, so I'm wondering if someone could tell me (in their own words preferably) what the difference between static_cast, reinterpret_cast and const_cast are?

const_cast I know removes a "const" from something, but I'm not sure what the difference between them all is, and what one I need to use in different situations.

Answer 1

To say "C casting is bad" is an extremity that by itself is about as bad as using C-style casts all the time.

The areas where "new" C++ style casts should be used are: hierarchical casts (upcasts, downcasts, crosscasts), const-correctness casts and reinterpretation casts. For arithmetical casts C-style casts work perfectly fine and pose no danger, which is why they can safely be used in C++ code. In fact, I would actually recommend using specifically C-style casts as arithmetical casts - just to make arithmetical casts to look different from other cast types.

Answer 2

1. static_cast is the standard c++ way to do a cast at compile time when programmer knows the type of an object and/or wants to let the compiler know.
2. dynamic_cast is like '(T)obj' where the cast is checked at runtime.
3. reinterpret_cast is used to cast between different objects without a runtime check.
4. const_cast explicitly converts to a type that is identical by removing the const and volatile qualifiers.