Thread: Please help understand following c++ syntax

It's an initialization list. From there, you can set values to your class's variables/constants or even call a specific constructor of a standard or otherwise class.

This :

Code:

class A{   
int a;
public:
       A(int b){a=b;}
}

is the same as:

Code:

class A{   
int a;
public:
       A(int b): a(b) {}
}

But if you're using something other than the built in types:

This :

Code:

class A{   
string a;
public:
       A(string b){a=b;}
}

is slower than:

Code:

class A{   
string a;
public:
       A(string b): a(b) {}
}

Because the first firstly calls the default constructor and then sets a=b, while the second directly calls the appropriate constructor.

It is also easier to achieve exception safety with an initialiser list, as the compiler does the work of cleaning things up properly if an exception is thrown.

To be fair,

Code:

class A{   
int a;
public:
       A(int b) { a = 0; }
}

Tells the compiler to
1) Call the default constructor for a (which does not exist),
2) Assign 0 to a.

Code:

class A{   
int a;
public:
       A(int b): a(b) {}
}

This tells the compiler to call the constructor for a that takes an int.

So they're different, while the later being more efficient for custom types, and sometimes also a requirement (such as when there is no default constructor available).