Next topic

[Rusky]

You shouldn't have to dynamically allocate your structs, they should work fine allocated the usual way.

The reason you keep getting "enigma::varaccess_amount(int(test))" is because ENIGMA thinks "test" is a normal EDL variable storing the index of an instance, rather than your struct. Robert's trying to get around that by using a pointer, but even then you shouldn't have to dynamically allocate it, just take its address.

It would also be helpful to know which version of GCC you have.

[Goombert]

After that I went and tested this code again in the create event... and it worked. So it's actually a JDI bug with script locals causing the dot operator to not work. You should not call new, and instead use the following code in events only.

Code (EDL) Select Expand


intStats test;
test.amount = 539;
show_message(string(test.amount));


[rotenKleber]

Thou art sure?

I added the constructor in definitions
And put this code into a create event

Code Select Expand


intStats test;
test.amount = 539;


But I am still getting the all familiar error:

Code Select Expand


.h:43:38: error: invalid cast from type 'itemStats' to type 'int'
     enigma::varaccess_amount(int(test))= 539;


It's as if the period still means nothing to Enigma. And the pointer operator worked not.

So is there a way to get around Enigma treating my struct as a normal variable?

[TheExDeus]

I'm surprised that structs actually work, as I think EDL parser was very far from finished when it was stopped being worked on. I would use structs everywhere if I could.
But in short - EDL != C++. And the things that are written in the wiki about EDL are not entirely true. Those were the planned features. Most of them weren't finished.

[rotenKleber]

@TheExDeus
Well, they really don't for me, still! But yeah, that would be pretty nice if they did. I'll have to think of a hacky workaround now.