[JASS] "code" type variable leak?

[8uY_YoU]

Would be it leaks if I pass a 'code' type variable (which is a function) as a parameter to another function?

Ex:

function DPS takes code c returns nothing
    local group g = CreateGroup()
    call GroupEnumUnitsInRect(g,GetPlayableMapRect(),null)
    call ForGroup(g,c)
    call GroupClear(g)
    call DestroyGroup(g)
endfunction

function f takes nothing returns nothing
    call SetUnitLifeBJ(GetEnumUnit(),GetUnitState(GetEnumUnit(),UNIT_STATE_LIFE)-1)
endfunction

function Trig_Untitled_Trigger_001_Actions takes nothing returns nothing
    call DPS(function f)
endfunction

//===========================================================================
function InitTrig_Untitled_Trigger_001 takes nothing returns nothing
    set gg_trg_Untitled_Trigger_001 = CreateTrigger(  )
    call TriggerRegisterTimerEventPeriodic( gg_trg_Untitled_Trigger_001, 0.01 )
    call TriggerAddAction( gg_trg_Untitled_Trigger_001, function Trig_Untitled_Trigger_001_Actions )
endfunction

 

[WaterKnight]

No, functions are static and function only denotes that the following identifier is a function and its pointer is wanted. No need to instantiate some object.

 

[deathismyfriend]

You should also optimize that code. Get rid of the BJs (red text functions) and never use TriggerAddAction use TriggerAddCondition and run everything from there.

 

[8uY_YoU]

No, functions are static and function only denotes that the following identifier is a function and its pointer is wanted. No need to instantiate some object.

Ah... thanks for the answer, this makes my life easier
I thought it will form an object like TriggerAddAction? how it does (still curious)?
+rep

You should also optimize that code. Get rid of the BJs (red text functions) and never use TriggerAddAction use TriggerAddCondition and run everything from there.

IMO, some of BJs are lovely, ex: BJDebugMsg
I only use TriggerAddAction for trigger that stays permanently (does it still leaks? does it produce an object every time the action is executed?)
Anyways, it just an quick example, but thanks

 

[deathismyfriend]

Look at my tutorial Converting GUI to efficient JASS. It explains why to not use TriggerAddAction.

Also yes some BJs are ok. The ones you have above though are not.

 

[8uY_YoU]

Well from your tutorial.. I took a conclusion that TriggerAddAction generates thread. As far as I know thread is subset of a process, how can it considered as object which is leak?

Looks like we go off-topic lol

 

[deathismyfriend]

It does not leak the thread it creates a new one that is not necessary and slows the trigger down.

 

[8uY_YoU]

Ah.. so it recreates the function that already exist instead of using the current one?
Like function Trig_Untitled_Trigger_001_Actions?

Looks like I need more reference material on this
Could you lend me some?
I already tried to search it myself, but I didn't know the correct key to search, lol

 

[deathismyfriend]

basically all triggers are run on threads. That is why it is possible to hit the op-limit for a trigger.
Using TriggerAddAction creates another thread. So it runs slower than just using TriggerAddCondition.

 

[WaterKnight]

TriggerAddCondition creates a thread as well. It's just faster, can return a boolean and has some other advantages like the destroy functions for triggeractions do not really deallocate the object, only deregister it from the trigger iirc. triggerconditions cannot have waits but those are rarely required. And no, contrary to the codes I read here all the time, returning from a triggercondition entry function is optional.

 

[PurgeandFire]

And no, contrary to the codes I read here all the time, returning from a triggercondition entry function is optional.

Returning a boolean isn't required by the standard editor or Wc3, but PJASS requires it (after PitzerMike updated it post 1.24 to address the return bug), which is why all the codes here return boolean in the callbacks.

 

[WaterKnight]

Due to the memory exhaust issue, originally, I am using JassParserCLI.

 

[PurgeandFire]

Due to the memory exhaust issue, originally, I am using JassParserCLI.

Oh, right. In case you ever want to try switching back, you can use LeP's updated version of PJASS:
http://www.hiveworkshop.com/forums/2458006-post345.html

 

[8uY_YoU]

Wow thanks for the responses..
In other words, it will run faster, isn't it?