Is it possible to get macros to compile in a fix order or to do multiple passes? I'm trying to have something use a macro as if it were a function::
e.g.
define GetHashCode(T) = GetHashCode_##T
define Real2Hash(v) = StringHash(R2S(v));
define InitHasher(T,HashMethod) =
{
int GetHashCode(T)(T value)
{
return HashMethod(value);
}
}
InitHasher(real,Real2Hash) causes it to freak out.

just wanted to report the bug i found!

define {
integer = integ ## er
integer(x) = GetHandleId(x)
}

will compile and work correctly if i put that in map directly! but if i use external file and #guard it will display some strange compile errors and characters...

ofc take ur time to test this tomorrow :)

Greets!
~DD

Re: Mailing you.

Yeah.. I'd need the makers permission to do that and such. But the code that is erroring.. if I isolate the vJass library where it is erroring at it compiles perfectly. All I can assume about the map itself is because it uses WEU and such that it is messed up incredibly for any sort of mod like this to ever work perfectly on.

Thank you regardless for your help.. looks like I'm just going to not work on trying to find a bug for the guy.

Good night and thanks for answering the questions!

~DD

", id, i)\n exitwhen true\n endif\n set i = i + 1\n endloop\n endif\n return i\n endfunction\n\n ///Computes a weak hash value, hopefully secure enough for our purposes\n private function poor_hash takes string s, integer seed returns integer\n local integer n = StringLength(s)\n local integer m = n + seed\n local integer i = 0\n loop\n exitwhen i >= n\n set m = m * 41 + C2I(SubString(s, i, i+1))\n set i = i + 1\n endloop\n return m\n endfunction\n\n ///Stores previously sent messages for tamper detection purposes\n private struct QueueNode\n readonly real timeout\n readonly string msg\n readonly integer checksum\n readonly string key\n public QueueNode next = 0\n public static method create takes integer id, string msg returns QueueNode\n local QueueNode this = QueueNode.allocate()\n set .timeout = time() + 7.0 + GetRandomReal(0, 2+0.1*GetPlayerId(GetLocalPlayer()))\n set .msg = msg\n set .checksum = poor_hash(.msg, id)\n set .key = I2S(id)\n return this\n endmethod\n private method onDestroy takes nothing returns nothing\n call FlushStoredInteger(gc, M_KEY_VAL+.key, .msg)\n call FlushStoredInteger(gc, M_KEY_CHK+.key, .key)\n set .msg = null\n set .key = null\n set .next = 0\n endmethod\n public method send takes nothing returns nothing\n call StoreInteger(gc, M_KEY_VAL+.key, .msg, .checksum)\n call StoreInteger(gc, M_KEY_CHK+.key, .key, .checksum)\n call SyncStoredInteger(gc, M_KEY_VAL+.key, .msg)\n call SyncStoredInteger(gc, M_KEY_CHK+.key, .key)\n endmethod\n endstruct\n \n ///Returns true for a fixed size uniform random subset of players in the game\n private function isEmitter takes nothing returns boolean\n local integer i = 0\n local integer n = 0\n local integer r\n local int

I am appearing to get "string too long errors" and this occured only after my new update. It also whined at me about a COMMENT.

Sigh errors that never have happened before.

n endglobals\n\n ///////////////////////////////////////////////////////////////\n /// Private variables and constants\n ///////////////////////////////////////////////////////////////\n globals\n private constant boolean SHOW_DEBUG_MESSAGES = true\n \n private constant string chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789_-+=\\!@#%^&*()/?>.<,
:0x5C227B7D5B

The last "B" there is highlighted, and it says incorrect literal. I can personally assure you that it is NOT incorrect because it has compiled fine before.

The actual line it appears to be erroring at is:

private constant string chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789_-+= \\!@#$%^&*()/?>.<,;:'\"{}[]|`~"

Is there something wrong with endline manipulation and ;?

Good night everyone, support session is over for today xD

ohh sry for spam but i posted just at time where u answerd Mooglefrooglian's question...

so about #guard! i am just not sure what name should it be... i mean SOMENAME should be file name? so like this:

#guard cj_antibj_base.j

or smth else?

as well FUNCNAME how exactly should i use it?

i mean is it string or is it just macro name?

i mean if u could write an example xD it would be great!

Greets!
~DD

Thanks again for everything.. interestingly, I did find a way to avoid hooking globally initalized CreateTiemrs(). I merely did:

define CreateTimer = Create ## Timer

Randomly in my little debug trigger, and then at the top of the map script, I put:

setdef CreateTimer = CreateTimer_Hook

Which makes it so I hook only non globalised CreateTimer() calls.. I would like to have the globally initted ones as well, buuut.

Also, odd bug:

function Trip_AbilCode takes nothing returns integer
return 'A05B'
endfunction

Errors for me and according to my compiler:

function Trip_AbilCode takes nothing returns integer
return0x41303542
endfunction

And errors because it has no return and such.

Is this cJass or vJass erroring? I never had this problem before I installed cJass, so..

thanks VD i thought it worked that way but was not sure, so once again thanks :)

Greets!
~DD

# + Added #error directive.
# + Added #guard directive.
# + Added FUNCNAME system define.

Can you go into further detail about what all of these will do? FUNCNAME assumably will tell the name of the function? But what about the rest?

ahh sry xD

i actually never understanded how this setdef and all that works...

defines are replaced before and after the line: define smth = smth_else

now smth will become smth_else

however setdef means that define will now be changed but from where to where :S

undef as well what does that mean that it will be removed or smth? but i cant fully understand from which point and what will happen after its removed...

so here example:

nothing smth() {}

define smth = blah
setdef smth = smth_else

integer smth = 0
undef smth
real smth = 0.0

nothing smth() {integer i = smth_else; blah()}

is that correct or did i understand it wrong?

thanks :)
~DD

very nice update, gj guys!

hope to see 1.4 as well ;)

about while loop, well if you say so then i guess ill define it that other way.

@Mooglefrooglian

about what u asked i think u can do this:

define CreateTimer = Create ## Timer
setdef CreateTimer = CreateTimer_hook

this code should be at the top but after main GUI globals

Regards!
~DD

Thanks a lot. A custom common.j never occured to me. I'd love to see what ADOLF comes up with.

Have a good one.

Hi, I posted the problem earlier, but you seem to have misunderstood me.

Because some of the variables are GUI, and initialised in the Varible Editor, they are moved to the top of the map script. It is IMPOSSIBLE to put a function above those, and as such I can't hook a function like CreateTimer(). I would like to know if there is a way I can get it to avoid hooking globals that are initted, or else find a way to get a function above those globals.

THank you for the response though. My original question was:"Small problem. Defines affect the globals in the map header. Therefore I am unable to hook CreateTimer() and such because it is defined with GUI variables.

Any fix? Or did I not read syntax correctly?"

yeah i once defined it like:

define while (expr) = whilenot not (expr)

but Van Damm told me it was better the other way, it was coz i had to type arguments like this:

while (<my_expr>)

i guess now thats fixed... :)

about while)

this is the way to do it correctly:

define UnitData(u) = unitdata(GetUnitUserData(u))

struct unitdata extends array
integer ID = xxx
integer Type = xxx
...
...
endstruct

nothing Test () {
unitdata ud = UnitData(GetTriggerUnit())
}

if define has arguments it must be called with arguments so you cant have with same name struct and define.

btw my stile is to always have types with lower-case.

Greets!
~DD

Sorry, my examples arr too bad.

a new example:

define <UnitData>(u) = UnitData(GetUnitUserData(u))

struct UnitData extends array
integer ID = xxx
integer Type = xxx
...
...
endstruct

nothing Test () {
UnitData = UnitData(GetTriggerUnit())
}

Sometimes i want to

<jass>
define <Test> = Test
define <Test>(x) = Test(x)

struct Test extends array
// xxxxxxxxx
// xxxxxxxxx
endstruct

nothing CantWork () {
Test a = Test(1)
Test b = Test.Create()
}
</jass>

What should i do ?

yeah when we are talking about blizzard.j!

is it possible for adolf to recode blizzard.j by fixing blizzards own bad code and then insted of using blizzard.j from .mpq he injects his own code directly in the map.

that would really be great but not neccessary.

about my wish list, ill add few more which i forgot about last time.

numbers mean priority to me :)

this 7 are important to me and i hope that ADOLF will like them as well. so maybe in 1.5 ill be able to see this requests but i know i cant wish to much since this all might be hard to add, i mean i dont know asm so... :)

as well nice that this bug with while loop is fixed in 1.3.4.30!

Regards!
~DD

Small problem. Defines affect the globals in the map header. Therefore I am unable to hook CreateTimer() and such because it is defined with GUI variables.

Any fix? Or did I not read syntax correctly?

k anyway thanks and i hope it gets fixed :)

Greets!
~DD

found a crash!

while (this.qn-- > 0) { this.qi[this.qn] = null; this.in[this.qn] = null }

qn = integer
qi = questitem array [32]
in = string array [32]

as well while is defined like you said

define while = whilenot not

Greets!
~DD

k thanks!

i never actually learned (C) i started with (C++) so i dont know what the difference is xD

glade to hear enums can be treated as types, for now i guess ill have to do this:

enum {smth, smth_else...}; define name = integer

for now that will work :)

about overloading defines, yes i knew that if i call it with () it will work but i had already made it in whole script without () so i only wanted to know will ADOLF remove this limitation in feature :)

once again thanks for all your answers and help!

Regards!
~Dark Dragon