Tool Galaxy parser

[Darthfett]

Well, I'm partly to blame since I forgot to include the example implementation

Anyway, perhaps more noob friendly: http://phyrex1an.net/galaxy.php

int PlayerGetTeam(int whichPlayer) {
    return (whichPlayer + 1) % 2;
}

Compiles, does this mean the % operator is valid in Galaxy?

 

[Jesus4Lyf]

Compiles, does this mean the % operator is valid in Galaxy?

Yes, so is +=, -=, etc. Lots of things are valid, now...
I want to go back to vJass already, nearly... these guys are unfortunately right. At the same time, they haven't seen the struct pointer stuff yet, I believe..

Edit:
Don't expect methods, [LJASS]this[/LJASS] is not a reserved word. However, [LJASS]new[/LJASS] is. But I haven't worked out how to use it, yet...

Edit:
PS. This works fine:

struct node {
	node* next;
	node* prev;
	int val;
};

 

[Vestras]

J4L, YOU ARE GOD! This is so awesome.
Make more experiments and I might make a syntax highlighter for Galaxy for you, lol

 

[phyrex1an]

Lots of new syntax to parse Good job j4l. I'm starting to think that I "just" should parse the entire C language...

 

[Vestras]

Btw, what about ++, --, switch, etc.?

 

[UndeadDragon]

I'm glad that Blizzard included all these new features in Galaxy

 

[Romek]

I don't think it'll be particularly difficult to learn. :thup:

 

[quraji]

Yes, so is +=, -=, etc. Lots of things are valid, now...
I want to go back to vJass already, nearly... these guys are unfortunately right. At the same time, they haven't seen the struct pointer stuff yet, I believe..

Edit:
Don't expect methods, [LJASS]this[/LJASS] is not a reserved word. However, [LJASS]new[/LJASS] is. But I haven't worked out how to use it, yet...

The overwhelming pessimism + dark layout of the site put me in a depressed mood.

As for methods, couldn't the "this" be implicit? I hope that structs aren't just for data, I like the encapsulation. Maybe there will be classes if not?

Jealous that you're getting to mess around with the stuff already...I guess I'll have to attain myself the beta as well.

 

[phyrex1an]

Please test

int[3] foo = {1,2,3};

If that works try to make it const too.

 

[Darthfett]

As for methods, couldn't the "this" be implicit? I hope that structs aren't just for data, I like the encapsulation. Maybe there will be classes if not?

No.

struct Unit

    boolean r

    method attack takes Unit which returns nothing
        if r then
            set r = not r
            call which.attack(this) //this Undefined in Galaxy
        endif
    endmethod
endstruct

From the example code we've seen so far, structs being used only for data is what it looks like it's going to be. I'm not terribly disappointed, because at the very least you get structs, but I agree that encapsulation is nice.

 

[Lyerae]

Yes, so is +=, -=, etc. Lots of things are valid, now...
I want to go back to vJass already, nearly... these guys are unfortunately right. At the same time, they haven't seen the struct pointer stuff yet, I believe..

Edit:
Don't expect methods, [LJASS]this[/LJASS] is not a reserved word. However, [LJASS]new[/LJASS] is. But I haven't worked out how to use it, yet...

Edit:
PS. This works fine:

struct node {
	node* next;
	node* prev;
	int val;
};

Wait, isn't that a linked list?

 

[quraji]

Wait, isn't that a linked list?

Yes.

 

[Jesus4Lyf]

Check this out:

void doTest(){
    WriteLine("Hiya");
}
bool epicTest (bool testConds, bool runActions) {
    Wait(0.5,0)
    (*(&doTest))();
    return true;
}

Compiles, and works, but also gives the error data pointer tried to access code space, when (*(&doTest))(); is called.

 

[Renendaru]

That's weird. Kind of like a warning. How are you figuring all this out, trial and error? One person is doing a lot more than wc3c.net. Haha.

 

[Jesus4Lyf]

How are you figuring all this out, trial and error?

Yep.

Hmm.

int x2(int i){
    return i*2;
}
bool epicTest (bool testConds, bool runActions) {
    Wait(0.5,0);
    WriteLine(IntToString(((*(&x2))(1))));
    return true;
}

This doesn't output anything to the screen. I conclude that function pointers are currently unable to return values properly, then. Must be to do with that error I get. I suppose these are not completely implemented.

At least we know the syntax for using them, brings lots of hope for the future, in my opinion.
(ie. we can write a framework to use function pointers instead of strings for triggers, and stuff like that, if Blizz doesn't change it..)

 

[Renendaru]

I'm surprised you're figuring out so much mroe than wc3c, all they're do is doing tiny experiments, but you already wrote a system.

 

[Lyerae]

Yeah.
Also, in my research for the language file for Notepad++, I found this:

//==================================================================================================
//  AI Native Function Prototypes
//==================================================================================================

Galaxy has prototypes?

 

[Jesus4Lyf]

Prototypes are the means for calling a function before its implementation is declared.

void b(); // prototype
void a(){
    b();
}
void b(){
    // do something
}

 

[Anker]

Unfortunately, J4f. you were not the first...

Multidimensional array
http://bbs.islga.org/read-htm-tid-38337-fpage-4.html

Struct
http://bbs.islga.org/read-htm-tid-38341-fpage-4.html

Pointer & Function Pointer
http://bbs.islga.org/read.php?tid=38425

 

[Vestras]

# function StructTest() {
#     myStruct myStructTest;
#     myStructTest.a = 256;
#     myStructTest.b = 874;
#

#     UIDisplayMessage(PlayerGroupAll(), 4, StringToText(IntToString(myStructTest.b)));
# }

That isn't valid Galaxy...