Re: More keyword abomination by C++0x
mytype* operator=(const mytype* Rhs) { ...; return this;}
Would have worked equally as well.
Then how does the expression:
a = ++b;
work?
Same way as overloading unary operator->() works in C++.
Combining two separate concepts (pointer and dereferenced pointer, which
is what you seem to be trying to do) almost always turns out to be a bad
idea.
I'm not combining any concepts. Now where the compiler dereferences a
reference for you it _could_ have done a pointer.
With references you don't get any control at all over dereferencing and you
get no 'perhaps you made a mistake' prompts from the compiler because the
value and value& are generally interchangeable.
What I think would have been more useful would have been references as an
alias with scope control. It'd help with RVO too.
consider something along the lines of... (where references are name aliases
and they can extend the scope of the object until the reference goes out of
scope too)
std::string& foo() {
std::string temp; // a's ctor called
stuff;
return temp; // return as '&' pulls temp out to scope outside foo
}
void bar()
{
if (test) {
std::string& a=foo(); // call foo
...;
dosomethingwith(a);
} // a (temp) dtor called here
}
i.e. temp becomes a local of bar with scope inside the if {} block in this
example.
Nesting would drag the scope back further. I'd like that more than returning
temps by value or manually creating the return temporary & passing
pointer/reference as an argument.
Chris
--
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]