On Wed, 17 Oct 2007 09:24:50 -0000, "ianbrn@gmail.com"
<ianbrn@gmail.com> wrote:

>Hello,
>I have a struct defined thus:
>
>typedef struct myStruct
>{
> int j;
>} myStruct;
>
>I saw somewhere calls with the following syntax:
>f1(&*a)

It would if we new what a was.

What do you think the effect of the combined & and * operators is? For
extra points, would it make a difference if the operators were
reversed? What is the only possible type the expression &*a (or *&a)
can have (on those occasions when it is a legal expression)?

>
>What should be the protype of f1 so that it will be correct and
>so that compilation will succeed ?

The prototype is not the problem.

>
>I am talking about two cases:
>In the first, we have the following definition:
>myStruct a;

What happens when you apply the * operator to a struct?

>
>in the second, we have
>myStruct* a;
>
>Is it possible with both case to have a definition of f1() so that
>f1(&*a) will be correct and pass compilation ?

While passing an incompatible type to a function possibly could be
"corrected" by changing the prototype, what makes you think a
prototype can magically correct a syntax error?

Again, what happens when you apply the * operator to a struct?

>
>Any ideas?

Yes. Make a note of the author who wrote the call to f1. Then make
sure you never again take anything written by him seriously.


Remove del for email