http://www.devx.com/cplus/Article/28105/0/page/3
Look at this page. On one of the lines, the code is
What does this mean? It seems to assign a new StringImp to the pimpl variable, but what is the difference between the above code, and this:Code:String::String(): pimpl (new String::StringImpl) {}
What exactly does the colon after the constructor do?Code:String::String() { pimpl=new String::StringImpl }
The first example illustrates initialisation of a member variable using an initialisation list. The second example illustrates assignment to a member variable in the constructor body. For a pointer type there should be effectively no difference in efficiency, but for some class types the first example would be more efficient since the second might involve an expensive default initialisation followed by an expensive assignment.
Look up a C++ Reference and learn How To Ask Questions The Smart WayOriginally Posted by Bjarne Stroustrup (2000-10-14)
The initializer list will initialize objects in that it calls the constructor directly.
If you try to assign in the body, it will first call the default constructor and then call operator = to assign a new String::StringImpl.
In this case, as laserlight points out, it doesn't really matter which way you do it. But when it comes to more complex objects, it might be better to initialize it since it would save the time it takes to construct it and then assign. And also, if an object does not have a default constructor (a constructor that takes no arguments), then you must use initializer lists to pass the correct arguments to the correct constructor of the object.
