Inventory effects & expressions are really slow in 1.12

[TrademarkTM]

Hello again.
I was trying to make a kits script. However, I noticed that giving items is working insanely slowly in the new versions (5 ms to give 5 items). You can even feel that there's a big delay between each item appearing in your inventory

Here's how I am doing it:

command /kit [<text>]:
	permission: itempackage.kit
	permission message: {@P}
	trigger:
		set {_s} to size of {packages::package::%arg 1%::*}
		loop 36 times:
			if {_n} = {_s}:
				stop loop
			slot loop-number - 1 of player = air
			add 1 to {_n}
			set {_slots::%{_n}%} to loop-number - 1
		loop {packages::package::%arg 1%::*}:
			set slot {_slots::%loop-index%} of player to loop-value

Takes 1~1.5 ms to run.

However, simply giving the items is laggy:

command /kit [<text>]:
	permission: itempackage.kit
	permission message: {@P}
	trigger:
		loop {packages::package::%arg 1%::*}:
			give loop-value to player

Takes 5 ms.

Hope you can fix this.
Thanks!

[bensku]

Hmm, sounds like a possible bug. Could you check what happens if you just give player same item in loop, without using loop-value. Does it cause performance issues?

[TrademarkTM]

Made a full test here:

Insanely slow:

command /test0:
	trigger:
		remove all items from player's inventory

[15:02:38 INFO]: [Skript] # /test0
[15:02:38 INFO]: [Skript] # test0 took 12.803814 milliseconds

Reasonable, but slower than manual clearing

command /test1:
	trigger:	
		clear player's inventory

[15:03:24 INFO]: [Skript] # /test1
[15:03:24 INFO]: [Skript] # test1 took 1.709404 milliseconds

I think it's slow for just 4 items

command /test3:
	trigger:
		set {_a::*} to diamond sword, iron sword, iron axe and stone shovel
		loop {_a::*}:
			give loop-value to player

[15:03:54 INFO]: [Skript] # /test3
[15:03:54 INFO]: [Skript] # test3 took 1.852104 milliseconds

Most performance-friendly inventory clearing:

command /test4:
	trigger:
		loop all items in player's inventory:
			clear loop-item

[15:04:27 INFO]: [Skript] # /test4
[15:04:27 INFO]: [Skript] # test4 took 0.475441 milliseconds

0.6 ms for one item.

command /test5:
	trigger:
		set {_i} to diamond sword of sharpness 5, knockback 2 named "TEST" with lore "HEY"
		give {_i} to player

[15:10:58 INFO]: [Skript] # /test5
[15:10:58 INFO]: [Skript] # test5 took 0.645621 milliseconds

Not bad, but could be better I think.

command /test6:
	trigger:
		if player can hold 64 stone:
			broadcast "CAN HOLD STONE"

[15:11:38 INFO]: [Skript] # /test6
[15:11:38 INFO]: CAN HOLD STONE
[15:11:38 INFO]: [Skript] # test6 took 0.533713 milliseconds

Faster than looping, so fine

[15:12:22 INFO]: [Skript] # /test7
[15:12:22 INFO]: HAS DIAMOND SWORD
[15:12:22 INFO]: [Skript] # test7 took 0.455576 milliseconds

command /test8:
	trigger:
		broadcast "%number of diamond sword in player's inventory%"

[15:14:56 INFO]: [Skript] # /test8
[15:14:56 INFO]: 5
[15:14:56 INFO]: [Skript] # test8 took 0.404589 milliseconds

command /test9:
	trigger:
		remove 64 diamond from player's inventory

[15:17:53 INFO]: [Skript] # /test9
[15:17:53 INFO]: [Skript] # test9 took 0.372805 milliseconds

I am comparing only player's tool, I don't think it should take 0.5 ms.

command /test10:
	trigger:
		tool = diamond sword
		broadcast "IS DIAMOND SWORD"

[15:20:04 INFO]: [Skript] # /test10
[15:20:04 INFO]: IS DIAMOND SWORD
[15:20:04 INFO]: [Skript] # test10 took 0.439022 milliseconds

Another thing:
If I give this sword to player give diamond sword of sharpness 5, knockback 2 named "TEST" with lore "HEY" to player, then use player has diamond sword of sharpness 5, knockback 2 named "TEST" it won't work, because the lore doesn't match.

The same works for NBT, if I add ANYTHING different in NBT, it won't match the item, even though it matches the conditions.

Using only name and enchantments, it'll match normally.

inventory of %player% is empty doesn't work too.

That's it basically.

[bensku]

I will take look at performance issues. About your "another thing", are you sure it is new with latest release? Sounds like old "feature" for me.

[TrademarkTM]

http://prntscr.com/g4sg78

command /test11:
	trigger:
		give a diamond sword of sharpness 5, knockback 2 named "TEST" with lore "HEY" to player
		if player has diamond sword of sharpness 5, knockback 2 named "TEST":
			send "Indeed"
		else:
			send "Negative"

[Tuke-Nuke]

That's because the diamond sword of sharpness 5, knockback 2 named "TEST" isn't a part of condition, so it creates an item and pass to the condition, and the condition check if it has that item, not similiar item.

[TheLimeGlass]

@TrademarkTM Is this still an issue? Bensku has changed the Slot system recently

[bensku]

The changes are not relevant for this issue.

[sovdeeth]

Tested on 2.7-beta3, 1.20.1:

[22:28:54 INFO]: [Skript] # /test0
[22:28:54 INFO]: [Skript] # test0 took 1.3959 milliseconds

[22:28:37 INFO]: [Skript] # /test1
[22:28:37 INFO]: [Skript] # test1 took 0.1617 milliseconds

[22:28:40 INFO]: [Skript] # /test3
[22:28:40 INFO]: [Skript] # test3 took 0.5178 milliseconds

[22:28:42 INFO]: [Skript] # /test4
[22:28:42 INFO]: [Skript] # test4 took 0.1335 milliseconds

[22:29:52 INFO]: [Skript] # /test5
[22:29:52 INFO]: [Skript] # test5 took 0.6364 milliseconds

[22:30:35 INFO]: [Skript] # /test6
[22:30:35 INFO]: CAN HOLD STONE
[22:30:35 INFO]: [Skript] # test6 took 0.2426 milliseconds

[22:32:31 INFO]: [Skript] # /test8
[22:32:31 INFO]: 5
[22:32:31 INFO]: [Skript] # test8 took 0.6906 milliseconds
[22:33:18 INFO]: [Skript] # /test8
[22:33:18 INFO]: 36
[22:33:18 INFO]: [Skript] # test8 took 1.5131 milliseconds

[22:34:08 INFO]: [Skript] # /test9
[22:34:08 INFO]: [Skript] # test9 took 0.3518 milliseconds

[22:34:27 INFO]: [Skript] # /test10
[22:34:27 INFO]: [Skript] # test10 took 0.0839 milliseconds
[22:34:29 INFO]: [Skript] # /test10
[22:34:29 INFO]: IS DIAMOND SWORD
[22:34:29 INFO]: [Skript] # test10 took 0.3762 milliseconds

It seems to me the main issue is item comparison performance, not inventory manipulation.

[AyhamAl-Ali]

Since we dropped support for legacy versions (1.12-) and there has been many improvements to item comparison I guess this has been improved/fixed on the way. If anyone thinks this is still a problem please open a new issue.