https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102609
--- Comment #7 from waffl3x <waffl3x at protonmail dot com> ---
struct S {
int f(this S*) {
return 5;
}
};
int main()
{
S s{};
return s.f();
}
Here is my current progress, this code works. I have a good feeling that the
rest is going to be easy. Except for deduction maybe, but I have a minor hunch
that it might -just work- without any extra tinkering after everything else has
been implemented.
Also yes, I know that `int f(this S*)` is not a valid declaration, I just
didn't have to change any member function call code for it to work this way so
I made it my first goal to implement it this way.
Actually, I guess it's not explicitly invalid is it? I don't recall seeing it
in the pathological cases section of the paper, indeed, I just checked and
there doesn't seem to be mention of it. I believe it should probably be valid
though, not useful if you ask me. Even less so than the other pathological
cases (which I think might actually prove to be useful in some cases.) However,
regardless of my opinion, following the same direction as the original paper, I
imagine it should be allowed. The class would just need an implicit conversion
to pointer, maybe a warning would be in order though?
