Strange results when using SELECT in a function #5

New Issue

2021-04-05T13:45:30Z

bloggy commented

2021-04-05 13:45:30 +00:00

(Migrated from github.com)

I get <none> on the first run and the correct value on the second run of my script, which is very strange. Maybe you can tell me what's happening here.
A broadcast within the function will always display the correct value on the first run.
With first run and second run I mean entering the command "/test".

Versions:

This server is running Tuinity version git-Tuinity-"94f025a" (MC: 1.16.5) (Implementing API version 1.16.5-R0.1-SNAPSHOT)
Loading Skript v2.5.3
[skript-db] Loading skript-db v1.2.1

The script:

function testing(p: player) :: number:
	set {_uuid} to {_p}'s uuid
	execute "SELECT uuid, player, points FROM `mytable` WHERE `uuid` = %{_uuid}%;" in {sql} and store the result in {_r::*}
	set {_points} to {_r::points::1}
	# This will return the correct value!
	#broadcast "Points: %{_points}%"
	return {_points}


command /test:
	trigger:
		set {_p} to the player
		set {_result} to testing({_p})
		# This will return the wrong value (it will return <none> on the first run, and the correct value on the second run)!
		message "Final: %{_result}%"

Table Structure (Screenshot):
https://i.imgur.com/p9Vh15W.png

How can this happen? How to fix?

I get `<none>` on the first run and the correct value on the second run of my script, which is very strange. Maybe you can tell me what's happening here. A broadcast within the function will always display the correct value on the first run. With first run and second run I mean entering the command "/test". **Versions:** - This server is running Tuinity version git-Tuinity-"94f025a" (MC: 1.16.5) (Implementing API version 1.16.5-R0.1-SNAPSHOT) - Loading Skript v2.5.3 - [skript-db] Loading skript-db v1.2.1 - **The script:** ``` function testing(p: player) :: number: set {_uuid} to {_p}'s uuid execute "SELECT uuid, player, points FROM `mytable` WHERE `uuid` = %{_uuid}%;" in {sql} and store the result in {_r::*} set {_points} to {_r::points::1} # This will return the correct value! #broadcast "Points: %{_points}%" return {_points} command /test: trigger: set {_p} to the player set {_result} to testing({_p}) # This will return the wrong value (it will return <none> on the first run, and the correct value on the second run)! message "Final: %{_result}%" ``` **Table Structure (Screenshot):** https://i.imgur.com/p9Vh15W.png How can this happen? How to fix?

GovindasOM commented

2021-04-05 13:46:30 +00:00

(Migrated from github.com)

it is because you cannot use delayed effects in return functions, execute is delayed

ham1255 commented

2021-04-05 13:48:46 +00:00

(Migrated from github.com)

Why execution is delayed here?

bloggy commented

2021-04-05 13:49:25 +00:00

(Migrated from github.com)

Thank you for your quick reply.
Same question:

Why would execute be delayed? Reason for this?
What would be the best to fix it?

Thank you for your quick reply. Same question: - Why would execute be delayed? Reason for this? - What would be the best to fix it?

GovindasOM commented

2021-04-05 13:49:55 +00:00

(Migrated from github.com)

Thank you for your quick reply.
Same question:
* Why would execute be delayed? Reason for this?

* What would be the best to fix it?

it is delayed to not lag the server, the queries are ran on separate thread

> Thank you for your quick reply. > Same question: > > * Why would execute be delayed? Reason for this? > > * What would be the best to fix it? it is delayed to not lag the server, the queries are ran on separate thread

GovindasOM commented

2021-04-05 13:50:37 +00:00

(Migrated from github.com)

in future version I will make "synchronous execute" to also be asynchronous, but converting the current code to async, thus avoiding delay, so maybe that'll be able to solve the issue partially

bloggy commented

2021-04-05 13:53:46 +00:00

(Migrated from github.com)

So that means for now I am unable to use return values that come from SELECT statements in a skript function?

GovindasOM commented

2021-04-05 13:57:36 +00:00

(Migrated from github.com)

yes, that's the case with all similar add-ons, a workaround is skript-reflect's delayed effect

bloggy commented

2021-04-05 14:02:03 +00:00

(Migrated from github.com)

And what about this situation:

Can it happen that when I send two EXECUTE commands one right after the other, the result of the second event is not displayed correctly to me?

Example:

execute "INSERT INTO {@table} (`uuid`, `player`, `points`) VALUES (%{_uuid}%, %{_playername}%, %{_points}%) ON DUPLICATE KEY UPDATE `player` = %{_playername}%, `points` = %{_points}%" in {sql}
execute "SELECT uuid, player, points FROM `{@table}` WHERE `uuid` = %{_uuid}%;" in {sql} and store the result in {_r::*}

Could the SELECT result then be empty the first time because the INSERT INTO wasn't fast enough for the SELECT to receive the new value?

Regarding skript-reflect's delayed effect: it may be inconsistent as well, or not? Because you never know how long to wait?

And what about this situation: Can it happen that when I send two EXECUTE commands one right after the other, the result of the second event is not displayed correctly to me? Example: ``` execute "INSERT INTO {@table} (`uuid`, `player`, `points`) VALUES (%{_uuid}%, %{_playername}%, %{_points}%) ON DUPLICATE KEY UPDATE `player` = %{_playername}%, `points` = %{_points}%" in {sql} execute "SELECT uuid, player, points FROM `{@table}` WHERE `uuid` = %{_uuid}%;" in {sql} and store the result in {_r::*} ``` Could the SELECT result then be empty the first time because the INSERT INTO wasn't fast enough for the SELECT to receive the new value? Regarding skript-reflect's delayed effect: it may be inconsistent as well, or not? Because you never know how long to wait?

GovindasOM commented

2021-04-05 14:03:24 +00:00

(Migrated from github.com)

it should be consistent and shouldn't have any issues, mysql should handle it well
you know how long to wait, wait until a value is returned (which is when the code continues after EXECUTE)

1. it should be consistent and shouldn't have any issues, mysql should handle it well 2. you know how long to wait, wait until a value is returned (which is when the code continues after EXECUTE)

bloggy commented

2021-04-05 14:04:56 +00:00

(Migrated from github.com)

Could you by any chance post an example of such a skript-reflect "delay" syntax?

GovindasOM commented

2021-04-05 14:05:28 +00:00

(Migrated from github.com)

sadly I do not have an example, but if I remember correctly, you just type:
delay the effect or something like that

sadly I do not have an example, but if I remember correctly, you just type: ```delay the effect``` or something like that

GovindasOM commented

2021-04-12 13:30:02 +00:00

(Migrated from github.com)