[JASS] Need help with hashtable.

[Geshishouhu]

Hi guys, I wonder how hashtable access its data, please see the following example:

function example takes nothing returns nothing
	...
	call SaveReal(hash,98765432,0,1)
	....
	call LoadReal(hash,98765432,0)
endfunction

//Compare to the following parent key

function example takes nothing returns nothing
	...
	call SaveReal(hash,1,0,1)
	....
	call LoadReal(hash,1,0)
endfunction

//The acessing speed is different and my questions are:
//1. how big of a difference does it make?
//2. in the first example, does the parent key value got accessed immediately? or is it added up from 0 to 98765432? I hope this makes sense. :D
//   I mean how does hashtable access its keys?
//3. how about using hex number as parent key? Is it faster in that way? (I don't think so but not sure)

Edit:
I know that with more data stored in one hashtable, the accessing speed decreases. So do the two examples have the same reduction in terms of accessing speed?

How about creating a unit inside of a hashtable? Is it inefficient? If so, why?

function example takes nothing returns nothing
	call SaveUnitHandle(hash,0,0,CreateUnit(.....))
endfunction

 

[Ceday]

1)You will never notice the difference.
2)Immediately.
3)Hard question, someone with better knowlodge about w3 hashtables need to answer this one.

 

[Geshishouhu]

1)You will never notice the difference.
2)Immediately.
3)Hard question, someone with better knowlodge about w3 hashtables need to answer this one.

Thanks.
So why bothering using arrays than?

 

[chobibo]

Because accessing/storing on arrays don't require function calls.

 

[Geshishouhu]

Because accessing/storing on arrays don't require function calls.

Damn, I should have figured that out.....
Thanks.

 

[muzzel]

//3. how about using hex number as parent key? Is it faster in that way? (I don't think so but not sure)

Short answer: no. The internal integer representation is always the same.

Because accessing/storing on arrays don't require function calls.

Correct, but im pretty sure that this is not the reason why hashtables are slower than arrays. Hashtables do a bit more than just storing data at a predefined place, which is the reason why you have so huge keys and still only use small amount of memory (sparse).

 

[Ceday]

Your last edited question answer is also no, there is no difference, in terms of efficiency it is even better to do this way.

 

[Geshishouhu]

Thanks for all of you, +Rep