adventure-help

[sterile rampart]

we're not your personal servants, read for yourself

[jolly oyster]

woo, who said that? i just asked a simple question no need to be rude saying such thing lol

[upper galleon]

I’m currently using:
<dependency>
<groupId>net.kyori</groupId>
<artifactId>adventure-api</artifactId>
<version>4.21.0</version>
</dependency>
<dependency>
<groupId>net.kyori</groupId>
<artifactId>adventure-text-minimessage</artifactId>
<version>4.21.0</version>
</dependency>
But for some reason on Minecraft 1.16.5 I’m getting this error:

https://mclo.gs/cFtPG85

[sterile rampart]

are you relocating? you probably have a conflict with whatever version the server has

[sterile rampart]

though at the end of the day, this applies

[upper galleon]

No, I'm not using relocate.

[sterile rampart]

that's your problem

[upper galleon]

Okay, I'll give it a try. Thanks.

[oak hamlet]

Hi guys, I am getting weird error: https://paste.md-5.net/etaroquqix.md it is caused by MiniMessageHolder
Here is my code:

public interface MiniMessageHolder {

    MiniMessage MINI_MESSAGE = MiniMessage.builder()
            .preProcessor(new LegacyColorPreProcessor())
            .postProcessor(new LegacyColorProcessor())
            .build();

}

public class LegacyColorProcessor implements UnaryOperator<Component> {

    private static final TextReplacementConfig LEGACY_REPLACEMENT_CONFIG = TextReplacementConfig.builder()
            .match(Pattern.compile(".*"))
            .replacement((matchResult, build) -> Legacy.component(matchResult.group()))
            .build();

    @Override
    public Component apply(Component component) {
        return component.replaceText(LEGACY_REPLACEMENT_CONFIG);
    }

}

public class LegacyColorPreProcessor implements UnaryOperator<String> {

    @Override
    public String apply(String component) {
        return component.replace("§", "&");
    }

}

public class Legacy {

    public static final LegacyComponentSerializer AMPERSAND_SERIALIZER = LegacyComponentSerializer.builder()
            .character('&')
            .hexColors()
            .useUnusualXRepeatedCharacterHexFormat()
            .build();

    public final static Component RESET_ITALIC = Component.text()
            .decoration(TextDecoration.ITALIC, false)
            .build();

    public static Component component(String text) {
        return RESET_ITALIC.append(AMPERSAND_SERIALIZER.deserialize(text));
    }

    public static List<Component> component(List<String> texts) {
        return texts.stream()
                .map(Legacy::component)
                .collect(Collectors.toList());
    }

}

plugins {
    `java-library`

    id("xyz.jpenilla.run-paper") version "2.3.1"
    id("com.gradleup.shadow") version "9.0.0-beta12"
}

group = "pl.karoldronia"
version = "1.0-SNAPSHOT"

val mainPackage = "pl.karoldronia.bounties"
val projectPrefix = "stormcore-zlecenia"

repositories {
    mavenCentral()

    maven("https://storehouse.okaeri.eu/repository/maven-public/")
    maven("https://repo.papermc.io/repository/maven-public/")
    maven("https://repo.eternalcode.pl/releases")
    maven("https://jitpack.io")
    maven("https://s01.oss.sonatype.org/content/repositories/snapshots/")
}

dependencies {
    // -- paper --
    compileOnly("io.papermc.paper:paper-api:1.17.1-R0.1-SNAPSHOT")

    // -- configs --
    implementation("eu.okaeri:okaeri-configs-yaml-bukkit:5.0.6")
    implementation("eu.okaeri:okaeri-configs-serdes-bukkit:5.0.6")
    implementation("eu.okaeri:okaeri-configs-serdes-commons:5.0.6")
    implementation("eu.okaeri:okaeri-configs-json-simple:5.0.6")

    // -- commands --
    implementation("dev.rollczi:litecommands-bukkit:3.9.7")

    // adventure
    implementation("net.kyori:adventure-platform-bukkit:4.3.1")
    implementation("net.kyori:adventure-text-minimessage:4.16.0")
    implementation("net.kyori:adventure-api:4.18.0")

    // -- notifications --
    implementation("com.eternalcode:multification-bukkit:1.1.4")
    implementation("com.eternalcode:multification-okaeri:1.1.4")

    // -- gui --
    implementation("dev.triumphteam:triumph-gui:3.1.12")

    // -- database --
    implementation("eu.okaeri:okaeri-persistence-flat:3.0.1-beta.2")

    // -- economy --
    compileOnly("com.github.MilkBowl:VaultAPI:1.7.1")

}

tasks.shadowJar {
    archiveFileName.set("${project.name} v${project.version}.jar")

    exclude(
        "org/intellij/lang/annotations/**",
        "org/jetbrains/annotations/**",
        "META-INF/**"
    )
}

can somebody help me solve this error please?

[sweet hornet]

Legacy support is not encouraged here

[oak hamlet]

yea I know, but I think legacy support is not the problem

I am making plugins for some time already and this is the first time when I am making something for 1.17.1 and got this error 🥴

[sweet hornet]

It is an extremely outdated version

[prisma mason]

Have you tried setting minimessage version to the same one as adventurw api?

[robust wharf]

Adventure supports all the versions

[oak hamlet]

yes - the same error

[sweet hornet]

Legacy formatting

[obsidian nimbus]

you can not use minimessage on such outdated version. You need to use adventure-platform and relocate adventure, lacking paper adventure support if you want minimessage

oh wait

yeah, thats it

you need to relocate adventure

[oak hamlet]

just it?

[obsidian nimbus]

yes, but you are lacking paper methods which support components then

[oak hamlet]

relocate entire net.kyori package?

[obsidian nimbus]

net.kyori.adventure i believe

[oak hamlet]

that's working!!! Thank you so much!

[tight seal]

Is it possible to change Audience Language or resolve a key in a specific language?

[robust wharf]

Only english is available server side

[tight seal]

I mean my own Plugin translation which I Feed into global Translator

[dense veldt]

Currently it's not possible natively in adventure. There is adventure/#1221 open, which adds support for this though.

Some platforms allow changing the locale of an audience themselves, like Player#setEffectiveLocale(Locale) in velocity for example, but if you're looking for something in paper it doesn't exist iirc

[tight seal]

Okay, and is it possible to get the translated component from the global translator in a spezific language?, i guess i can use that: https://jd.advntr.dev/api/latest/net/kyori/adventure/translation/GlobalTranslator.html#render(net.kyori.adventure.text.Component,java.util.Locale)
Hope this pr get implemented 💯

[last horizon]

i have a Formatter.choice and want to display the time in singular and plural accordingly
this is my current MM format

time.day=<time:'1#one day|days'>
time.hour=<time:'1#one hour|hours'>
time.minute=<time:'1#one minute|minutes'>
time.month=<time:'1#one month|months'>
time.week=<time:'1#one week|weeks'>

how do i show the time (number) in case it is not 1

i want 1 to be one day
and 5 to be 5 days

thats my formatter

at the moment no matter what the input is it always renders as "one x"

[cloud vapor]

isn't it 1#day|1<days

I don't think you can just raw | at the end

[last horizon]

i analyzed the ChoiceFormat class and it doesnt support inserting the number itself

i use a boolean choice instead now

[cloud vapor]

generally you use choiceformat to format the text

and insert the number separately

[last horizon]

yeah for singular/plural a boolean choice is much nicer
i just added a boolean choice named "singular"
and a number formatter for time

[winter flint]

Trying to send this minimessage string to the player with sendRichMessage <hover:show_text:'test'><green>Your bio has been updated! Hover to see! but for some reason the hover text just isn't showing. It's really puzzling me as in the web viewer it works just fine so I don't think it's the formatting

Any insight would be appreciated!

[sweet hornet]

Which platform / versions?

[winter flint]

Paper 1.21.5

Through Velocity but I doubt that would make any difference

[sweet hornet]

Paper 1.21.5 is still experimental, no? I know there are some changes in the Minecraft 1.21.5 formats for text events (hover / click), so maybe it comes from here

[winter flint]

Might be why, I'll do some further experimentation but it's a weird bug 😭

I fixed it, was doing some weird stuff elsewhere that was affecting it

[sweet hornet]

👍

[void slate]

adventure hover text isn't being displayed when the message is sent through Velocity

on Paper, it works fine

[sand drift]

Make sure that you're on the latest version of velocity

[void slate]

i'm on the latest version

[sand drift]

client version?

[void slate]

1.21.5

[sand drift]

are you using protocol hacks?

[void slate]

not on velocity

[sand drift]

Can you test without other plugins on the proxy?

My understanding is that support for the new hover text format was added a few weeks ago, but, I've not toyed around with it in a while

[void slate]

i'll test it right now

i tested with no other plugins on the proxy and the hover text still isn't showing

click events also do not work

[void slate]

it makes sense

[sand drift]

what client are you using?

[void slate]

1.21.5, vanilla

[sand drift]

I mean, are you using the vanilla client or somethging else

[void slate]

and it works on 1.21.4

[sterile rampart]

and which velocity build?

[void slate]

"Velocity 3.4.0-SNAPSHOT (git-a549880d-b491)"

[sand drift]

That is, in fact, not the latest.

[void slate]

is there a command to check the build?

[sterile rampart]

it's easier to get help if you don't lie to us

[void slate]

how can i check the build

[sterile rampart]

it's right there, 491

the latest is 509

[void slate]

i didn't know that meant the build 😅

[void slate]

my bad! didn't realize that b491 meant the build number. thanks for explaining, tho

[elder torrent]

Does anyone know if it's currently possible to nest MiniMessage translated components? For example: Component.translatable("example.key", Component.translatable("example.argument"))

• Where example.key represents the string <blue>Info » </blue><green><arg:0></green><dark_aqua> is a complex argument!</dark_aqua>
• Where example.argument represents the string This <bold>complex</bold> string

Currently, example.key is translated but example.argument is interpreted literally, resulting in Info » example.argument is a complex argument!

The use case for this would be a wall of text displaying info for the current arena. The titles/descriptions are strings which currently can only be hardcoded in one language via a Component.text() argument.

[dense veldt]

no, its not. but its being worked on

[sleek delta]

private Component message;
String json = GsonComponentSerializer.gson().serialize(this.message);
this.message = GsonComponentSerializer.gson().deserialize(json);```
4.21.0
The json string contains the hover and clickevent, but when I deserialize it into the message Component, hover and clickevent disappear. Should I use another serializer?

[native cargo]

what if you serialize that this.message again and print it lol

are hover and clickevents in the json payload

if not that's probably a bug that should be fixed

serialize/deserialize should be isomorphic

[sleek delta]

yes they are

[native cargo]

then it doesn't sound like an adventure issue it sounds like somethings up with your client

[sleek delta]

I sent them via a packet - the json string is with hover&click but after the third line with deserialize() they disappear in the Component where it gets null in the specific hover and click thing
I dont know how I can change something in this path 😄

[upper galleon]

Hello, could you please help me fix this error?
Error: https://mclo.gs/lMxu6bU
Code: https://mclo.gs/nbKvh00

[sand drift]

You can't relocate adventure and then use the native adventure methods in paper

[upper galleon]

I'm using relocate, but nothing seems to change.

[sand drift]

Yes, and I'm telling you not to

surprise, paper doesn't have a method that uses your relocated version of adventure

[upper galleon]

Unfortunately, that won't work — if I remove it, my plugin stops working on version 1.16.5.

Is it possible to fix it without removing it from relocate?

[sand drift]

No

Read the method name, the error is fairly clear

[next adder]

How can I get the in-game locale code by passing a java.util.Locale ?

I am currently figuring out https://github.com/KyoriPowered/adventure/blob/main/4/api/src/main/java/net/kyori/adventure/translation/Translator.java

[cloud vapor]

What platform are you using?

[next adder]

Paper

[primal dome]

Can I somehow add multilanguage support like eng:"bla",otherlang:"bla"?

[cedar wind]

https://paste.denizenscript.com/View/133865
getting this error when shading adventure 4.21.0 on 1.19.4 paper

Is the solution to stay on an older version indefinitely?

[graceful kestrel]

relocate

[cloud vapor]

It is not possible then, the server doesn't know about all the locales the client has

You could download them manually or smth

[cloud vapor]

Not like that but yes, see the docs for localisation

[cedar wind]

I'm using the libraries feature, shouldn't it be relocated for me?

[sand drift]

No, libraries stuff doesn't relocate, it has no means to accept patterns, etc

[cedar wind]

hmm, annoying. guess I'll go back to libby/runtime relocations

[sweet hornet]

The parameter of Key#key method is annotated with @KeyPattern, thus every time I want to create a key from an unknown string (from a user, or a config value, for example), I have an Unsubstituted expression warning from IntelliJ. While I understand why, I'd like to know what is, according to you, the best way to create a Key from an unknown string (note that I am perfectly fine with the InvalidKeyException, which is caught later on)

I am aware that I can just @SuppressWarnings("PatternValidation"), but there must be a better solution to such a common use-case, no?

[rare sage]

something like

@KeyPattern String str = readKey(...);
Key key = Key.key(str);

should work, or annotating your key-reading method, although these only make sense if you make sure to validate the read string

[obsidian nimbus]

i jus ignore the warning

[sweet hornet]

But why should I bother validating the string when the Key#key already does?

[obsidian nimbus]

well thats more documentation thingy, can be helpful if your key is literal

[obsidian nimbus]

otherwise jus do this

im sorry for ping

[rare sage]

or configure intellij to ignore the keypattern annotation altogether

[cloud vapor]

i'd say that exceptions aren't "normal" flow so unless we had some sort of Key#parseKey method (that was nullable or had a default or smth) then it does make sense for the annotation to be there

happy for a Key#parseKey(String, @Nullable Key) method to be added

that doesn't have that annotation

but yeah, just putting the annotation elsewhere or ignoring it is fine

[sweet hornet]

And I agree. In fact, in my scenario, when the key is invalid, the exception is caught later only to be wrapped in a more precise one, for logging purpose, and then disable the feature that is not properly configured

I do not personally consider this the "normal" flow of execution, am I wrong?

[cloud vapor]

Yeah you're right

which is why it makes sense for that warning to be there, bc something that doesn't match the pattern will throw an exception

[sweet hornet]

Hmm, understandable. I guess the main issue is that IntelliJ is not aware that the exception is caught and this is fine, which explains why I need to tell him to suppress the warning then

[cloud vapor]

yeah i'd just @SuppressWarning for that line/statement

[amber prawn]

Is there any tag to enable default text shadow generated based on text color without having to specify the color?

[cloud vapor]

You could make one p easily

[amber prawn]

Inside the minimessage there is no tag for this?

[cloud vapor]

Correct

[copper marsh]

Seems like adventure-nbt's SNBT reader is flawed

Made an issue here -> https://github.com/KyoriPowered/adventure/issues/1240

Is the fix to just add an if check or is there some other approach I could take?

[cloud vapor]

Your fix looks good to me, although can you add a test just so we don't miss it?

[copper marsh]

Hopping to bed right now, will do it when I wake up if nobody else gets to it first

It's been a long 14 hours of coding

[copper marsh]

Done

[small cloud]

why does the white not reset the color to white?
<#38a5e3>SizeChanger <#d4cf60>»<white>

it only removes the 2nd hex color?

no wait...

I guess doing the insert of the prefix via

final TagResolver subSectionResolver = TagResolver.resolver("ref", (argumentQueue, context) -> {
  final String reference = argumentQueue.popOr("reference expected").value();
  return Tag.inserting(parseComponent(reference));
});

is the wrong way?

I mean with Tag.inserting, I thought it would continue the lats color from the component but aparently its the first color of the component?

[midnight spire]

probably has something to do with how components are structured, if you remove the color of a child component it will get the color of the parent

and what you describe sounds exactly like that

[small cloud]

I mean if I do <#38a5e3>SizeChanger <#d4cf60>»<red>t it comes goes to that? is there some way to continue the color of the last component?

[cloud vapor]

show your full code please

and remember that components are trees, if you have two sibling components, the style of each component does not impact the other

[small cloud]

https://github.com/Hugo5000/PaperMC-ConfigLib/blob/master/src/at/hugob/plugin/library/config/ConfigUtils.java#L518

[prisma mason]

are you appending a different component to the minimassge one?

[small cloud]

I am inserting it via Tag.inserting

I thought it would just continue the last color... but what would be the best way to continue the last color?

[midnight spire]

add it as a child of the last child component, not the parent

no idea how that works with Tags though

[small cloud]

I guess I could get the last child and copy its color and make an empty parent component that color?

[midnight spire]

maybe just Tag#selfClosingInserting? kinda sounds like what you want

[small cloud]

well self closing would remove all overflowing colors, no?

I'd like the last color to continue

[midnight spire]

yes, it would use the styling of the part from before your tag

[small cloud]

ah, I think your misunderstanding something I'd like the last color of the tag the continue, not before the tag

[midnight spire]

what is a "last color of the tag"? a tag is something like <red> afaik? Or am I misunderstanding that Tag concept?

[small cloud]

if I have <red><ref:some_reference>text
and some_reference points to <gold>this is golden text should be gold now

[midnight spire]

so Tag#inserting requires a </red> later to stop any parts after it from becoming a child, Tag#selfClosingInserting automatically ads a </red> after the passed Component

I don't think that's how Tags are supposed to work. All Components will be children of your <ref:some_reference> Tag, not of the <gold> Tag

[small cloud]

okay.. I guess I need to be a bit hacky then

[midnight spire]

is there a reason you are not using replacements?

[small cloud]

bc each ref could have its own ref

[midnight spire]

well you could just replace in replacement values too I guess xD

[small cloud]

that... works?

[midnight spire]

I mean, you can make anything work with computers ;)

[small cloud]

true true

[midnight spire]

no idea if Adventure has inbuilt functionality like that, I use my own logic lol

[cloud vapor]

an inserting tag would work literally fine in this example, inserting a component of e.g. Component.text("Hi", RED) will cause subsequent components to be red

unless you explicitely close the tag itself

ofc you can't close it with e.g. </red> because there is no matching opening red tag

[small cloud]

well the problem is when the ref has multiple colors

[cloud vapor]

yes because subsequent components become children of the root component

[cloud vapor]

in your exact scenario here - you should be a) using one minimessage string or b) using a preProcessParsed tag

[small cloud]

ye ye I know that now too qwq
I just meant "its not really working as I want it to"

on a tangental note, can I somehow get the color where a tag is in the TagResolver?

[cloud vapor]

what do you mean by that?

can you explain a bit more about what you want to achieve?

[small cloud]

well rn I want to

get the color the ref should start with (i.e. the color before the tag)
set the tags component color to that
box the tag with an empty component with the color the text should continue in

[cloud vapor]

yeah that's terrible

again, just use a preProcessParsed tag

[rare ridge]

My point (which was given as far as I see, but) did you just try to close your additional colors in the Component?

[small cloud]

you mean like <#38a5e3>SizeChanger </#38a5e3><#d4cf60>»</#d4cf60><red>t ? that only makes the continuation text white

[rare ridge]

Yeah I mean that :/

[small cloud]

yaknow I somehow always hoped that insertion would just append the following text after the last child

yaknow I do wonder what the hell <reset> does tho, since if I use that the following text is white xD

well okay I'll just use the preprocess one, is there some way to not process a matched tag if some requirements are not met? or what exactly does a ParsingException lead to?

[sweet hornet]

What do you mean "not process"? A ParsingException stops the parsing of that tag, and cause a failure in strict mode, and is otherwise ignored (inserting the raw text instead)

[small cloud]

raw text as in the tag in plain text form?

[sweet hornet]

Raw text as the exact input text

If the processing of my-tag throws, <red><my-tag>Text will show <my-tag>Text in red

[small cloud]

kay thanks!

whats the proper way to throw one?

[sweet hornet]

From a tag resolver?

throw ctx.newException("message", args) (or maybe swap the arguments I don't remember)

[small cloud]

ah thanks missed that method

[cloud vapor]

closes all currently open tags

obvs only works in one single parse context

[small cloud]

do preprocessors proprocess other preprocessor tags?

[sweet hornet]

What?

[cloud vapor]

the preprocessor is run multiple times until the string doesn't change

or up to some arbitrary maximum

[small cloud]

kay

good to know

thanks

[small cloud]

huh does the preprocessor of a minimessage instance not get applied to preProcessParsed tags?

[sweet hornet]

It wouldn’t surprise me

[cloud vapor]

no, the pre processor operates on the string before the parse

[small cloud]

sad

[cloud vapor]

why?

what are you trying to do? there's likely a better solution

[small cloud]

nah I'm just replacing the legacy color codes with minimessage tags in my preprocessor

[sweet hornet]

Which is bad

[small cloud]

...why?

[sweet hornet]

You shouldn’t try to mess with multiple formats

[small cloud]

thats why i'm converting it to one format?

[cloud vapor]

idek wht that has to do with pre process tags

[sweet hornet]

And in this case, legacy is what it means, a legacy, now deprecated, format

[small cloud]

well I'd like the minimessage instance to also treat preprocess tags the same, as a deserialize string

[sweet hornet]

You should only use MiniMessage, and nothing else

If you really want to shoot yourself in the foot, you can still call the preprocessor yourself before returning preProcessParsed

But it’s just an open invitation for more issues later on

[small cloud]

nah not worth it I was just checking I didnt do something else wrong

[cloud vapor]

you should just do that in your pre process tags if you're gonna be cursed

[small cloud]

ye

[sterile star]

🌳

[regal fable]

Hey. Does anyone know, what is going wrong?

I tested it for debugging with an simple command that should actually display the displayname of an item into the chat.

Code:

p.sendMessage(Objects.requireNonNull(MiniMessage.miniMessage().serialize(Objects.requireNonNull(p.getInventory().getItemInMainHand().getItemMeta().displayName()))));

Output: See attached picture. Thanks

In config file: (not the same item, but well...)

? <!italic><!underlined><!strikethrough><bold><!obfuscated><gold>M16A1</gold></!obfuscated></bold></!strikethrough></!underlined></!italic><!italic><!bold><white>
    ▪ «11»
  : 0.1

[fiery oracle]

what exactly is wrong?

[regal fable]

I want the item to be serialized as a String to later deserialize it. The config value is peculiar.

guns:
  pistol1: 0.1
names:
  ? <!italic><!underlined><!strikethrough><bold><!obfuscated><gold>M16A1</gold></!obfuscated></bold></!strikethrough></!underlined></!italic><!italic><!bold><white>
    ▪ «11»
  : 0.1
  <!italic><!underlined><!strikethrough><bold><!obfuscated>Kugelsichere Weste T2: 0.1
materials:
  NETHERITE_HELMET: 1.0

[cloud vapor]

Peculiar how?

If you're talking about all the false style tags, people often create components with a "false" style in order to have a "blank slate" that isn't impacted by e.g. item names being italic by default

On Paper, legacy formatting items also undergo the same conversion when turned to components

[regal fable]

Aight, I see. For some reason there is a "?" when I try to save a displayname for a crackshot-item item into the config. Its fine. Thanks for now.

[sterile rampart]

that's just yaml lol

one of the formats for a block mapping

[regal fable]

Allright. Never realized it before.

[sleek delta]

can somebody still help?

[cloud vapor]

show the input string? also what platform/version/etc

[sleek delta]

Components: https://mclo.gs/Ww8WabE
GSON (de-)encode logic: https://mclo.gs/mStZ6kf
GSON string from Decoder: https://mclo.gs/wOG3ELd
string encode to component: https://mclo.gs/RpUyJvo
net-kyori-adventure-api = "4.21.0"
net-kyori-adventure-text-legacy = "4.21.0"
net-kyori-adventure-text-gson = "4.21.0"

[fiery oracle]

version of minecraft, they meant

client & server

[sleek delta]

lol guess some of my java applications wasnt up to date hours of debugging... thanks for the help!!

[molten jasper]

okay insane so i have the bundle of translations for some languages how do i register them I was trying to use some class
named ResourceBundleTranslationSource but in my IDE it looks like it doesnt exist xd
so ig the way is to go with translation store but how do i do this im so tired ive been trying to learn this
for last 4 hours but there is always some wall im meeting with

what i want to do is just send player message like "key",playerData.getLocale

ig i was close at some point

[cloud vapor]

https://docs.advntr.dev/localization.html

https://docs.advntr.dev/minimessage/translator.html

Give these both a thorough read over and let me know if you've got any questions !

[molten jasper]

„There are additional methods on the message format translation store to bulk register from resource bundles. You may also want to use Adventure’s UTF8ResourceBundleControl utility class to create your bundle.”

[molten jasper]

I read that so many times yesterday
But this is too general like what do you mean by there are additional methods

[hardy estuary]

Hello, I was checking in the GitHub Repo and I see in the past this error was reported but was related to people using Gson.

I get this error only in 1.21.5 with papermc 1.21.5 API.

Failed to convert data component value of type class net.kyori.adventure.nbt.api.BinaryTagHolderImpl to type class io.papermc.paper.adventure.PaperAdventure$DataComponentValueImpl due to an error in a converter provided by adventure:platform/paper!
https://mclo.gs/0peJPDX

This happens when I try to render in chat an Item that has a lore, if needed I can provide how I do it.

Tests has been done on:

Paper: This server is running Paper version 1.21.5-113-main@ba7fb23

ViaVersion/ViaBackwards has not been used in the server

[cloud vapor]

com.mojang.brigadier.exceptions.CommandSyntaxException: Expected compound tag at position 26: ...,text:""}]<--[HERE]
Looks like just bad NBT?

As always check you're running the latest versions

[hardy estuary]

yes im running latest papermc available in the website

This server is running Paper version 1.21.5-113-main@ba7fb23

this is the item im trying to render like usually with Component.hoverEvent(source) and as source the itemStack

[cloud vapor]

Can you show some code?

[hardy estuary]

sure

[hardy estuary]

https://pastebin.com/eK4kyJ6k

here it is, btw its

[cloud vapor]

can you make an MRE to reproduce it?

[hardy estuary]

passing as source the itemstack gives problem, but since its only in 1.21.5 I believe its due to SNBT changes?

[cloud vapor]

there's a lot of fluff there

[hardy estuary]

yes and Ni, I mean mostly theoretical but lemme do it

    // REQUESTED MRE

    public void doReplaceInChat(@NotNull Component messageFromChat, final @NotNull Player sender){
        final String prefix = "[item]";

        // Lets think that the message has the prefix and we skip the check if its present just for the sake of this example
        final ItemStack itemToShow = sender.getInventory().getItemInMainHand();
        final Component replacement = itemToShow.displayName().hoverEvent(itemToShow);

        messageFromChat = messageFromChat.replaceText(builder -> builder.matchLiteral(prefix).replacement(replacement));
        
    }```

[rare sage]

it means there additional methods, java methods that you can call, .foo(), on the TranslationStore class that you get for a MessageFormat to register a whole ResourceBundle
https://jd.advntr.dev/api/4.21.0/net/kyori/adventure/translation/TranslationStore.html#messageFormat(net.kyori.adventure.key.Key)
https://jd.advntr.dev/api/4.21.0/net/kyori/adventure/translation/TranslationStore.StringBased.html

[hardy estuary]

To recap this happens in 1.21.5 when using as HoverEventSource and ItemStack

[hardy estuary]

@cloud vapor should I open an issue on papermc github?

[rare sage]

can you send the output of the /paper dumpitem whilst holding it

[hardy estuary]

https://pastebin.com/MUnzfZak here it is, thank you

ah I forgot to say, this happens when item has a lore

if it doesn't have a lore, its fine

[rare sage]

idk it kinda just works fine for me, something else is going on on your server, whether it be your code or some other plugin

[hardy estuary]

mmm strange, what version are you running? of papermc

[rare sage]

pulled directly from main branch, latest available build too

which is what i see you are running too

[hardy estuary]

can you try with AsynChatEvent? its the only difference I see

and another difference I spot its you are in offline mode I believe? I will test this too

Thanks for helping btw

[sweet hornet]

Simple question, have you actually tested your MRE to confirm you have the issue with it?

Because looking at your original code, you’ve done a lot of simplification in your MRE, of things that could go wrong and cause your original issue

And I’m asking because it looks a little bit like you did your MRE using IA (sorry if not the case, the comments made me think that)

[hardy estuary]

you know, you are right, I have not tested that, mostly cause I noticed this happened when item has lore.

Btw its not IA made, it was written with my left feet xd

[rare sage]

i am not in a position to test with my account right now but from what i can see on the server code, that shouldn't matter, you should be using the sync event anyway if you're gonna interact with entities tho :p

but i will test the async event just for you

[hardy estuary]

thank for that

[hardy estuary]

I will clone the itemstack object to avoid possible issues, you are right

[rare sage]

cloning isn't going to fix it

for cloning the item still needs to be read

[hardy estuary]

will use a future, jump on main thread and get a clone

[rare sage]

async event works fine too

(not an excuse to use it)

[rare sage]

that'd be fine yeah

[hardy estuary]

thanks for that, at least I can exclude issues related to server code

[heady yew]

Does the LegacyComponentSerializer.legacySection() support the font tag? As in we are using it to deserialize the MiniMessage component into a string, however it after inserting the <font> tag it is removed from the string but the font isn't applied, but color fonts work however.
Or do we need to use the bukkit audiances instead to send the component to the player, instead of the player#sendMessage(String) method?

[midnight spire]

legacy formatting had no concepts of fonts

[sweet hornet]

No, you should really try to avoid as much as possible using legacy stuff

There is a high chance that you're actually doing something wrong if you "need" to do that

Esecially trying to go MM => legacy sounds REALLY wierd

[heady yew]

Right, so how would one set the title of a GUI to a format?

[midnight spire]

Inventory#title iirc?

[quaint monolith]

Changing the title of an open inventory isn't really supported, thus why there is no Component version of that method

The server is doing a hacky thing behind that method that causes problems sometimes, like flickering and desync

[midnight spire]

ah wait, it's on InventoryView and can only be set on opening :S

you can just reopen the inventory tbh

unless you plan to change the title all the time without user interaction it should be fine

[sweet hornet]

If you really need to, yes reopen, but otherwise you could maybe find a better alternative. What are you trying to achieve?

[heady yew]

Ok, thank you

[quaint monolith]

People like to change the title of an open inventory and use a custom font because they're making the title overdraw the entire inventory UI with a custom interface

And they change the title to animate it

[heady yew]

I was trying to add a option which allows the placeholders to be updated every few ticks for example

[quaint monolith]

Oh, nothing so fancy then

[stark badger]

It would be cool if minecraft added a way to update the title

[quaint monolith]

iirc it used to actually be a thing back when that method was added

[stark badger]

thats annoying

[midnight spire]

inventory gui usage will drop with Dialogs anyways

[stark badger]

why would they remove it

[heady yew]

Doesn't reopening the inventory cause a flicker?

[midnight spire]

it only flickers when you close the old one before iirc

[quaint monolith]

That method is reopening the inventory

iirc you can't do what it does without direct packet access though

[midnight spire]

@stark badger well if it's not used in the game itself then they sometimes clean things up

[quaint monolith]

What it does still causes flickering and desync issues sometimes though

[stark badger]

it would be nice if they added what the developer community wanted

[sweet hornet]

They did, and it's called dialogs

[quaint monolith]

They do add things just for developers sometimes but other times they don't want to allow what you want

Or they have some other idea on how to accomplish it that takes longer to implement so it seems like they ignored you then years later an alternative to what you wanted shows up

[heady yew]

Dialogs won't be able to display the items from a custom inv, or is there already a way?

[midnight spire]

you can display anything you want with a resource pack

but yes, if you want to show an inventory an inventory is probably the better choice

[heady yew]

Lets hope they won't replace inventory's with dialogs then, it would also be ashame as its been a thing from the beginning of MC

[midnight spire]

wat

how would you access your items without an inventory? xD

[heady yew]

I was not talking about player and block inventory's, I was just replying to your message

inventory gui usage will drop with Dialogs anyways

[quaint monolith]

I think they meant using an inventory to make a GUI, not using it as an inventory

[midnight spire]

the custom inventory API is literally using block inventories, they can't remove that from the game without removing chests and stuff

but with Dialogs being a thing the usefullness of custom inventories as GUIs that display stuff that isn't item-based (or was shoehorned into being item-based) will decline

[heady yew]

Ah, fair point then

[sweet hornet]

Dialogs can show items, iirc

[hardy estuary]

I was digging more, adding some souting to understand in which moment the error is triggered and the error is triggered in the moment paper send the chat message, so not in the moment I apply the replacement code logic (which I reduced to the examples above to test)

The photo with red line is when item has a lore, it seems server isnt able to find an Implementation for the BinaryTagHolder? while without lore it seems good!

I don't know what to think honestly. If there was a problem with the component, the error should have been triggered in the moment I create/replace such component right?

[cloud vapor]

You solution is to binary search through your plugins to find what's causing it

If it's your plugin, well, figure out what weird stuff you're doing outside of simple usages

[hardy estuary]

you think it could be a plugin incorrectly setupping the lore into the item?

[cloud vapor]

Well Emily showed that it isn't an issue with normal setup/usage

The usual check for packet hacks, client version, vanilla client, etc

[hardy estuary]

the strange thing is that this only happens in 1.21.5, previous versions same code works fine.

I will check like you said

[hardy estuary]

I solved, has something changed btw with TextReplacementConfig? it was that

[cloud vapor]

Nothing has changed with that in probably over a year

[hardy estuary]

dont know what to say, thanks for helping btw

[cobalt finch]

For some reason when I serialize a Component containing a showItem HoverEvent, I can deserialize it on the same server just fine. But if I send the serialized data to the Velocity proxy and deserialize in the same manor, I get this error. Is there a better way?

Paper side

                    Component itemComponent = item.displayName().hoverEvent(item.asHoverEvent());
                    itemInHand = JSONComponentSerializer.json().serialize(itemComponent);```

Velocity side
```        if (itemInHand != null) {
            Component itemComponent = JSONComponentSerializer.json().deserialize(itemInHand);
            chatEvent.getSender().sendMessage(itemComponent);
        }```

[fiery oracle]

i mean, try/catch and try and print the exception for debugging?
i imagine it's because of problems with versions, since the format for hover events/click actions has changed multiple times
you likely need to adjust the options of the deserializer on the velocity front to match whatever the serializer on the backend has

[cobalt finch]

Ah. You're right, but it's a client/server version mismatch. It's not an issue on the Paper server because I have ViaVer, but on the proxy there's nothing in between.

1.21.5 servers, 1.21.4 client. On a 1.21.5 client it's fine.

[fiery oracle]

viaversion would indeed handle translating the 1.21.5 format to the 1.21.4 format

[cobalt finch]

I wonder if it's at all related to this that was fixed 2 versions ago https://github.com/PaperMC/Velocity/pull/1578

[molten jasper]

if I use bundle would you recommend me to use only component style of message coding
or to prepare code for hybrid like usage with components and minimessages formats?

[sand drift]

"component style"?

[wet kestrel]

and what is "bundle" in this context?

[cloud vapor]

and what do you mean by "hybrid", as in multiple formats in one file?

[molten jasper]

hahaha so actually i just went with componnet style but for clarification
i meant like for mini messages i can just get bundle with this UTF8resourcebundlecontrol smth like this
while for basic style of resource i can get bundles manually and then save them in TranslationStore

and the hybrid i wanted to do was to get bundles with mini messages and then translate them to components on call

however as its recommended to use translation store on docs then i went with it

[cloud vapor]

yeah you can just use MiniMessageTranslationStore that's fine

all the translation stores give you components out

[molten jasper]

oh well

that was the response i was looking for then

[sage river]

is there a way to handle component lists with minimessage translator? e.g. for item lores

[sand drift]

paper doesn't deal with server sided translations for item stacks

if that's what you're expecting to solve here

[sage river]

oh, right

yeah then nvm

[molten jasper]

   
public static void loadTranslations(){
        MiniMessageTranslationStore store = MiniMessageTranslationStore.create(STORE_KEY, MiniMessage.miniMessage());

        for (Locale locale : SUPPORTED_LANGUAGES) {
            ResourceBundle bundle = ResourceBundle.getBundle(
                    BASE_NAME,
                    locale,
                    UTF8ResourceBundleControl.get()
            );

            for (String key : bundle.keySet()) {
                String message = bundle.getString(key);
                store.register(key, locale, message);
            }
        }

        GlobalTranslator.translator().addSource(store);

    }

    public static Component translated(Locale locale, String key, Component... args) {
        return GlobalTranslator.translator().translate(Component.translatable(key, args), locale);
    }

is this correct way of doing this?

[cloud vapor]

you can load a whole bundle into the store

but yes that looks fine to me

[fiery oracle]

https://jd.advntr.dev/api/4.21.0/net/kyori/adventure/translation/TranslationStore.StringBased.html#registerAll(java.util.Locale,java.nio.file.Path,boolean)
https://jd.advntr.dev/api/4.21.0/net/kyori/adventure/translation/TranslationStore.StringBased.html#registerAll(java.util.Locale,java.util.ResourceBundle,boolean)

iirc you probably want to call #render instead of #translate on the global translator

[molten jasper]

how?

[rare sage]

with the methods linked right above

[untold quarry]

Are there any intentions on making Sign/Entity Names/Block etc.. being actually translatable as messages? 🤔

[ebon talon]

are shadows supported by the chat? I can't make it work with MiniMessage, I tried adding explicitly StandardTags.shadow. I'm on latest paper 1.21.4 and also tried custom paper jar with adventure 4.21.0. They work in tab or in display entities.

[untold quarry]

I have different languages registered and works fine with messages, but when it comes to signs for example, it just displays the key of the translation, which makes me avoid using translatable components directly and just translate them to a fallback lang e.g English and then set sign lines.

So I was thinking what if there was a pre-packet-send translation approach where messages get translated only before getting sent to client, by this, each client gets their own view of the sign lines, same applies for everything else, items, titles, etc..

And best part is, when player A drops item X and player B pick it up, it will be viewed in the lang of player B, that would be awesome and top tier i18n

[cloud vapor]

You'll need to ask the platform you're working on

[cloud vapor]

Yes, they do work on chat. Make sure you're using a vanilla client and not one that's breaking it

[untold quarry]

Paper 😂

[cloud vapor]

Then go ask paper, this is adventure help :p

[untold quarry]

Isn't adventure the one that wraps platforms? Not the other way around?

[rare sage]

adventure is an abstraction that platforms implement

[cloud vapor]

Adventure doesn't know about signs or entities

[untold quarry]

Oh ok that makes sense

[spice grail]

Are text components with just whitespace characters optimized to be one space character?

[sweet hornet]

If it contains only a single space, iirc there is a shared instance yes

Component.space()

But not any whitespace, for example a tab will remain a tab

[fiery oracle]

i think only the empty component is guaranteed to share the same instance

https://github.com/KyoriPowered/adventure/blob/main/4/api/src/main/java/net/kyori/adventure/text/TextComponentImpl.java#L50

[sweet hornet]

Yeah I had a doubt about this

However it should not particularly matter for you, unless you really want to use a lot of space components lol

[fiery oracle]

what prompted this original question, anyway?

[molten jasper]

hey so i have this question i created this function:

    public static Component userError(Locale locale, String messageKey, Component... args) {
        Component errorSignature = Component.text("✗ ")
                .color(NamedTextColor.RED)
                .decorate(TextDecoration.BOLD);

        Component messageComponent = translated(locale, messageKey, args)
                .decoration(TextDecoration.BOLD, false);

        return signature().append(errorSignature).append(messageComponent);

    }

and now does this have sense to then do smth like this
error.info.party.leave.leader=You can't leave the party as a leader; <yellow><hover:show_text:'<yellow>click to disband'><click:run_command:'/party disband'>[disband]</click></hover></yellow> it instead.

[naive snow]

        this.miniMessage = MiniMessage.builder()
                .tags(TagResolver.builder()
                        .resolvers(TagResolver.standard())
                        .tag("pc", Tag.inserting(this.getMessage(MessageKeys.PRIMARY_COLOR)))
                        .tag("sc", Tag.inserting(this.getMessage(MessageKeys.SECONDARY_COLOR)))
                        .tag("prefix", Tag.inserting(this.getMessage(MessageKeys.PREFIX)))
                        .build())
                .build();

I have this minimessage builder with some custom tags added
the tag "<prefix>" is not working.
the values for these tags are fetched from a config,
this is the config content:

primary-color: "<red>"
secondary-color: "<white>"
prefix: "<pc>My</pc><sc>Plugin</sc> <dark_gray>»"

So what's the wrong thing im doing ?

[sweet hornet]

You probably want to use Tag.preProcessParsed(...) instead of Tag.inserting

[naive snow]

That fixed it, thanks !

[molten jasper]

.

[cloud vapor]

i mean, both feel like it would be fine

[hardy crest]

i dont get the question

[molten jasper]

actually the quesiton is does this minimessage formatting override message data in function

[cloud vapor]

i cant see you using minimessage anywhere in that code

so no, it's just a string

[molten jasper]

this is minimessage "You can't leave the party as a leader; <yellow><hover:show_text:'<yellow>click to disband'><click:run_command:'/party disband'>[disband]</click></hover></yellow> it instead."

which is stored in minimessagestore smth

for translation

so like my question is if there is red message like signature| X something went wrong <yellow>click

is it gonna be still red or this yellow from minimessage should override it

[sterile rampart]

more specific overrides less specific

components are a tree

[molten jasper]

ig ill just test it

[sour raft]

heres the yml

[wet kestrel]

I feel like you've forgotten something here

[sour raft]

yeah its in #paper-dev now

Or it was

i accidentally uploaded it twice

[molten jasper]

I dont know i get error: class MiniMessageTranslationStore not found during proxy initizlaization

[sterile rampart]

u p d a t e

[molten jasper]

im on velocity 3.4.0

[fiery oracle]

update to the latest build of velocity

[molten jasper]

this is latest build of velocity????

[fiery oracle]

what's your build #?

[molten jasper]

im using snapshot as it is recommended in docs it should use latest build always but i also refreshed dependencies
and it still doesnt find the class

[fiery oracle]

what?
im talking about on your velocity server
you said you got the error during proxy init

also, provide the full error?

[molten jasper]

im working on it

[shadow dove]

Hello

Just had a question regarding adventure.
on 1.19.4 the message I'm sending looks how I expect it to be, but for some reason on newer versions (we are using viaversion) it appears to break (breaks differently on different versions). I was wondering why that is and if there is a way to fix it.

[frank glacier]

Hello, do the new versions support 1.21.5 for hover, click event and everything else? Is support complete for 1.21.5?

[sand drift]

yes

[frank glacier]

hover and click doesn't seem to work in 1.21.5

[sand drift]

1.21.5 server and vanilla client?

[frank glacier]

yes

[sand drift]

generally, no idea; would need to see code, etc; as that generally should work

[frank glacier]

Loading Paper 1.21.5-103-main@2ba1675 (2025-05-29T12:14:46Z) for Minecraft 1.21.5
' <gray>- Type <red>cancel</red> <gray>or <click:run_command:"/{_clan} {_cancel}"><hover:show_text:"<red>Click to cancel."><red>[CLICK HERE]</red></hover></click> to cancel.'

Audience audience = this.plugin.getAdventure().player((@NotNull Player) sender); audience.sendMessage(text);

in 1.21.4 it's ok, 1.21.5 the texts in the chat don't work

a simple text is sent

this.adventure = BukkitAudiences.create(this); onEnable()

[sand drift]

you mean, 1.21.4 server and client, and a 1.21.5 server and client?

[frank glacier]

server and client 1.21.4 is ok
server and client 1.21.5 does not work

[sand drift]

No idea, support for that stuff was merged afaik

make sure that you test without other plugins

mojang did change the format fairly substantially in .5, many protocol library fluffery might break it

[frank glacier]

There are only 4 plugins in this instance started from 1.21.5. I will delete all of them for testing and leave only 1

same thing, only 1 plugin installed

[sterile rampart]

if this is adventure-platform-bukkit, #adventure-platform-mod-help

[cloud vapor]

Break how? What platform? Versions? etc

[summer comet]

Hey guys, how do I support ItemStacks automatically? (For example Nexo uses MiniMessage for the items but on my plugin is not compatible)

[cloud vapor]

What do you mean? Support them how?

[sweet hornet]

Explain what you need to do, with examples, because "support ItemStacks automatically" is too broad

[summer comet]

Someone told me this is not working because MiniMessage is not supported (item from Nexo)

[cloud vapor]

What isn't working? That looks fine to me

[summer comet]

I have no idea what could be the issue because if i hold this item on my hand it works fine but it gets unformatted when in the menu

This is the item:

[cloud vapor]

and MiniMessage spits out components which work fine when put in lore

You're gonna have to share some code or something 😅

[sweet hornet]

This not even looks like the same item

In terms of lore

[summer comet]

I'll get more information on this beforehand, just wanted to know if you guys knew something

[summer comet]

Yes it looks, the difference is that it has 2 extra lines which are related to the crate preview menu

[sweet hornet]

Then I don't understand what your issue is

If you have a valid item (with name / lore created with the help of MM or not), which becomes invalid after going through a modification process like your crate plugin, I don't think the culprit is the creation process (which may have involved MM), but rather it seems to be the modification process (your crate plugin, as you mentioned it)

[summer comet]

Thats correct, my plugin is designed to give compability to legacy format as well as Spigot but it seems paper has a displayName(MiniMessage) or at least a setDisplayName(BaseComponent)

And i'm using the setDisplayName/getDisplayName that works with a String

[sweet hornet]

The item does not store any MM-related stuff inside, it just have components (which, again, could have been created using MM, but this is just like a "factory" stuff, once it is created, it becomes independent from MM)

[summer comet]

Yeah I mean, i was expecting that

But it seems BaseComponent[] doesnt get parsed as expected for the String getDisplayName

[sweet hornet]

Wanting to support multiple formats at once is not recommended

[summer comet]

lol

[sweet hornet]

You shouldn't use deprecated methods in new code

And those two you showed here are deprecated

[summer comet]

I mentioned previously my plugin is designed to give compability to legacy

Thank you I'll take a look on my side now

[sweet hornet]

Which is not recommended

[summer comet]

Just because is not recommended, doesn't mean I shouldnt or can't work with them

I work with them for a reason

Appreciate the time, I'll step out now! 😉

[cloud vapor]

Legacy strings don't have the concept of fonts - you can't use fonts and legacy at the same time

So it's not about compatibility, because fundamentally you cannot get this to work at all using the legacy deprecated methods

[shadow dove]

1.19.4 Paper (native server version)

1.21.1

[cloud vapor]

how are you sending the messages?

[shadow dove]

Just player.sendMessage(comp);

[cloud vapor]

so you're not using adventure-platform or anything

[sweet hornet]

Are you sure the components aren't serialized using legacy serializer somehow?

[shadow dove]

Some parts of it are

[sweet hornet]

Because it really looks like this is the case

Then this is probably the root of your problems

[molten jasper]

   
 public static void loadTranslations(){
        MiniMessageTranslationStore store = MiniMessageTranslationStore.create(STORE_KEY, MiniMessage.miniMessage());

        for (Locale locale : SUPPORTED_LANGUAGES) {
            ResourceBundle bundle = ResourceBundle.getBundle(
                    BASE_NAME,
                    locale,
                    UTF8ResourceBundleControl.get()
            );

            for (String key : bundle.keySet()) {
                String message = bundle.getString(key);
                store.register(key, locale, message);
            }
        }

        GlobalTranslator.translator().addSource(store);

    }

    public static Component translated(Locale locale, String key, Component... args) {
        return GlobalTranslator.translator().translate(Component.translatable(key, args), locale);
    }

When i have something like this

error.usage=usage: <0>

and i provide

error.usage, Component.text("/party help")
it still says usage: <0>

i also tried with {0}

[cloud vapor]

arg:0

[molten jasper]

i know

okay so

arg:0

??/

[cloud vapor]

https://docs.advntr.dev/minimessage/translator.html#using-a-minimessage-translator

[molten jasper]

right

fr

[cloud vapor]

ong

[untold quarry]

Bruh why the tag resolver gets called twice for the same tag

💀 💀

[fiery oracle]

context?

[untold quarry]

I have an iterator that proceeds on each tag it finds, but it seems the tag resolver gets fired twice each time a tag is found

So the iterator literally skips a value for each tag it finds

I will open dis from PC to give more context

so i have this

    private TagResolver resolver(Iterator<Component> iterator){
        return TagResolver.resolver("n", (queue, context) -> {
            System.out.println(context.target());
            return Tag.selfClosingInserting(Component.newline().append(iterator.hasNext() ?  iterator.next() : Component.empty()));
        });
    }

and this message

<gold><b>Hey <receiver>!</b></gold><n><n><n><gray>We're glad you're here. Enjoy the show!</gray>

The sysout prints:

[20:42:27 INFO]: null
[20:42:27 INFO]: null
[20:42:27 INFO]: null
[20:42:27 INFO]: null
[20:42:27 INFO]: null
[20:42:27 INFO]: null

Though it should print 3 nulls instead, since it found only 3 <n> tags.

this results in losin X iterations from my iterator, where X is the number of <n> tags

I can get it working by a hacky way, but that's not really what I intend to

so should I go with String#replace 😭?

[robust wharf]

Tag resolver needs to be pure

[untold quarry]

Pure what?

stateless?

[sweet hornet]

https://en.wikipedia.org/wiki/Pure_function

[cloud vapor]

yep

you can use arguments e.g. <n:1> and so on

it's called multiple times because hasName defaults to checking if the resolve method returns non null for any given name

you can obvs avoid that partially by overriding hasName but note that resolve is still called in the pre process phase of the parse to check if you are providing a preprocess tag

i do want to improve this to cache resolve/name calls, but it's very low prio

[untold quarry]

I see, that makes sense. Didn't know there is a hasName method exists.

[untold quarry]

That would be nice

[cloud vapor]

yeah it's easy to miss when you're using tag resolver as a functional interface

[hardy estuary]

@cloud vapor can you kindly accept my friend request? I would like to send you a private message related to a paid request. Hope this is not a problem, thanks in advance

[cloud vapor]

Unless you're offering full-time employment and are okay with a 3 month notice period then I'm not interested, sorry

[hardy estuary]

nope nothing like that, mostly is paid support for adventure

[cloud vapor]

i can give you free support for adventure if you just ask here lol

[hardy estuary]

It would require me to screenshare cause Im not able to explain it. This is why Im willing to pay for your time

[cloud vapor]

def not interested, sorry

[hardy estuary]

no problem with that! thanks anyway

[odd moth]

so I did this now:

        MiniMessage allowedTagsMiniMessage = MiniMessage.builder().editTags(t -> t.resolvers(tagResolvers)).build();
        MiniMessage notAllowedBuilder = MiniMessage.miniMessage();
        Component parsed = allowedTagsMiniMessage.deserialize(chatColor + notAllowedBuilder.escapeTags(originalMessage));

tagResolvers containing: color, gradient for people with said permissions (op in this case) and a custom TagResolver, the question is: can i make an exception for escapeTags to allow certain things (colors, gradient) depending on permission

escapeTags takes in a TagResolver which according to the docs 'escapes all known tags in input messages' but if I give it StandardTags.Color it doesnt keep gradient in

actually giving the notAllowedBuilder an empty TagResolver works, damn

[sweet hornet]

I think you're deeply overcomplicating this

Can you explain your main goal?

[odd moth]

im figuring out a way to make a chat plugin with some cases like

1. the player should be able to type [i] which gets replaced by <item> that replaced the <item> by a hover component showing their item
2. the player should not be able to use colors in chat, unless they are OP
3. the player should be able to set a chat color/gradient

[sweet hornet]

Ok, so first things first, for your [i] replacement, I suggest you use a pre-processor on your MM instances (doing basic string replacement of [i] into <item>)

Then, to make players able to use only some tags, depending on their permission, as you tried, I think using a dedicated MM instance for each player (or each "role" or "rank", if it makes more sense in your case) with only the appropriate tag resolvers is the good way. You don't need to escape tags since you removed the resolvers, so for example <b> won't match any known resolver in your bare MM instance, so it will just stay <b> and not become bold (same applies for colors, and other tags)

Finally, regarding the global chat style registered for each player, you cannot just include the raw tag text because you removed the resolver, but you can apply the style directly on the resulting component (with .mergeStyle(...)). However, I understand it is bit more difficult with gradients, so if this is an absolute requirement, you may need another approach

[odd moth]

yeah gradients is a must..

i found something good enough for now :p

        MiniMessage allowedMM = MiniMessage.builder().preProcessor(chatHoverHandler::replacePlaceholders).editTags(t -> t.resolvers(tagResolvers)).build();
        MiniMessage disallowedMM = MiniMessage.builder().tags(player.isOp() ? TagResolver.empty() : StandardTags.defaults()).build();
        Component parsed = allowedMM.deserialize(chatColor + disallowedMM.stripTags(originalMessage));

[hoary gulch]

👋

This is my current approach for splitting a sentence into multiple parts: ```java
String text = PlainTextComponentSerializer.plainText().serialize(content);
if (text.isEmpty()) return null;
List<String> sentences = StringUtils.splitSentence(text, MainConfig.SubtitleSettings.MAX_SENTENCE_LENGTH);

`StringUtils.splitSentence(String, int);` splits a sentence into multiple parts, but it only works for strings. My goal is to achieve the same functionality for `Component`s. How would that be done?

I could possibly serialize it into MiniMessage but tags shouldn't get involved in this.

[small cloud]

what does StringUtils#splitSentence do? does it like split it after x characters, the first space before x characters?

[hoary gulch]

For example, "Hello, how are you? Hope you are doing good." to:

Hello, how are
you? Hope you
are doing good.

Basically splits without breaking words

[small cloud]

so the first space before the xth character?

or whatever number MainConfig.SubtitleSettings.MAX_SENTENCE_LENGTH is, right?

[hoary gulch]

Yeah, like that

[sweet hornet]

I recently made a proof of concept about this, but there are quite a few limitations. Feel free to have a look at what I shared there: #adventure-contrib message

[rare ridge]

I also have a limited ComponentLineWrapper in the repo I maintain. I should look at your approach and perhaps use that to improve, indyteo ^^

[storm spoke]

Hello, I have a problem with minimessages. Specifically, I can't deserialize the placeholder in the <click> tag.

This is what the raw text of the message looks like:

<click:suggest_command:'/resident <username>'><hover:show_text:'<username>'><<username>></hover></click>: <message>

I using this code for working with this text:

MiniMessage.miniMessage().deserialize(text, additionalResolver)

There is additionalResolver:

resolvers.add(Placeholder.component("message", message));
resolvers.add(Placeholder.component("username", player.name()));

In the game itself, everything works correctly EXCEPT that <username> is not inserted into the <click> tag! Please help.

This is how it looks in the game:

[wet kestrel]

Pretty sure it has to be a parsed placeholder inside the hover and click.

[storm spoke]

And one more thing! If I remove the single quotes and write <click:suggest_command:/resident <username>> instead of <click:suggest_command:'/resident <username>'>, then <username> will be inserted correctly, but the <click> itself will break

[wet kestrel]

Pretty sure my reply covers your fix.

[storm spoke]

I don't quite understand. Should I do something like this firstly?..

        MiniMessage test = MiniMessage.builder()
                .tags(TagResolver.builder()
                        .resolver(additionalResolver)
                        .build()
                ).build();
        test.deserialize(text);

[wet kestrel]

You're doing Placeholder.component when you should be doing Placeholder.parsed for stuff that gets parsed into a minimessage tag.

[storm spoke]

Oh! It's really works! Thanks a lot!

[pliant escarp]

hi, how can I get a list of compound tags using adventure's NBT?
Found it:

tag.getList("name", BinaryTagTypes.COMPOUND);

[odd moth]

small question, is there a way for a TagResolver to actually return a complete empty string? everything I do seems to return in a space being returned anyway

[sweet hornet]

return Tag.selfClosingInserting(Component.empty())?

[odd moth]

also adds a space :/

[sweet hornet]

What is your input string?

[odd moth]

good spot

thanks lol

i changed everything except the one i was testing

[tepid fable]

is it possible to display dialogs with the new update?

[wet kestrel]

I believe it's up to platforms to implement most of the dialog stuff right now because it's still in flux. Paper, for example, is working on API.

[unkempt cloud]

........

[cloud vapor]

The exception tells you what is wrong p clearly :p

[unkempt cloud]

Yes, but what is the alternative?

[rare sage]

MiniMessage

[cloud vapor]

in your case, the <gradient> tag

[unkempt cloud]

And how can I replace a variable inside a component?

[quaint monolith]

You can define custom tags in minimessage

[unkempt cloud]

If I have %variable% and I want to replace it with the display name of an item stack, how can I do it?

[quaint monolith]

Or just string replace on the MM string before you deserialize it to components

Custom tag would be better though

https://docs.advntr.dev/minimessage/dynamic-replacements.html

[wet kestrel]

Just do "my item is <item>" and use tag resolvers

[unkempt cloud]

hmm

after this I sound stupid :C

let's hope that what I want to do works

And can I only use < >?

Because what would happen if I want to use the <red> tag for my own use?

[cloud vapor]

you can replace the red tag

and you can only use <> yes

[wet kestrel]

Using only <> is a good thing as lots of plugins use minimessage so users get familiar with it

[unkempt cloud]

What is this limitation due to?

[wet kestrel]

MiniMessage is all about <> stuffs. It would be silly to make it handle multiple different formats that nobody uses.

[proper sandal]

planning on making part of a plugin standalone, is there a standard way to get stdout as an audience?

[cloud vapor]

It's not a limitation, it's a conscious decision to make it easier for the end user

[rare sage]

it's just an interface, you could implement it and redirect the various sendMessage methods to a ComponentLogger

[cloud vapor]

with the ANSI component serializer if you want colour

[jolly sage]

How can I customize the formatter for ComponentLogger?
I noticed that adventure loggers use slf4j while craftbukkit uses java.util.logging and i am unsure how to change this
(basically I want to include the thread name, right now it's [21:32:49 INFO]: [plugin-name] log message i want it to be [21:32:49 INFO]: [plugin-name] [Server thread] log message)

[sand drift]

log4j hijacks jul

You would need to modify the log4j2 config of your platform

[jolly sage]

is it like a command-line parameter to starting the server?

ah it's the xml files

imma mess around with these ty for point it out

[sand drift]

https://logging.apache.org/log4j/2.x/manual/pattern-layout.html#converter-thread-name

[jolly sage]

do you maybe know where I can find the default one (for paper or for bukkit/spigot if it's the same)?

[sand drift]

you would grab it from the patched server jar or from the github repo

[jolly sage]

got it, tyvm

[warped marsh]

Hi! Is there already a way to abbreviate a component by putting "..." at the end? Kind of like StringUtils.abbreviate from org.apache.
Alternatively, is there a way to figure out the number of characters in the component, I see there's no .size() or something like that

[cloud vapor]

Nothing in the API for that no

[outer knot]

I am working on a plugin and want to use dialogs. Is there an ETA for 0.4.24 release?

[outer knot]

should have expected that

[cloud vapor]

4.24.0 doesn't even include a dialog API

Adventure won't have one for a while

[outer knot]

It doesn't? I was looking arount the commits and it looked like it did. OK

[cloud vapor]

Read the docs for the methods added

[sweet hornet]

As Kezz said, there is no way to do it properly into the API for now, because there are quite a lot of technical challenges to take into account, but you can have a look here: #adventure-contrib message
It is a proof of concept of component wrapping (with some important limitations, but if they does not apply to you, you might want to take inspiration)

You should be able to adapt the wrap method to create an abbreviate method instead

But the same important limitation will apply. Basically, if you want to support anything other than text component, you’re screwed

[unkempt willow]

was the maven repository changed? this URL is 404 but 4.22 works

[robust wharf]

no clue how that is deployed

https://repo1.maven.org/maven2/net/kyori/adventure-api/4.23.0/ its def all there

[unkempt willow]

oh so it's now on maven central? the adventure-platform-mod repo still uses the original URL

[robust wharf]

its always been on central?

oh its just manually unzipped into here https://github.com/KyoriPowered/adventure-javadocs

ah no theres some python scripts

[unkempt willow]

obviously it's published using an internet explorer activex browser extension, and someone forgot what button to press /s

[robust wharf]

idk what one has to do to trigger the action, not really important, zml will see this and do it when time permits

[sterile rampart]

should be updated

[ocean agate]

it should be A but shows "Strafe Left"

is that a mojang bug or client bug

[sweet hornet]

translate is not keybind

[ocean agate]

real

I used wrong key

[naive gale]

do components auto serialize w/ jackson objectmapper or do i need to make / register a custom serailizer?

or I guess just have a jsonserializer at all?

[prisma mason]

https://docs.advntr.dev/serializer/index.html

[high raptor]

is it intended that all virtual components are equal?

they dont override equals so inherit it from text component, which only compares the content, but for virtual components the content is always empty

[rare sage]

text component compares more than just content

also compares style and children

[high raptor]

ah

but the fact that they dont compare beyond that is annoying

cause im using them in the minimessage way as named translation arguments but they are all comparing equal so the bossbar name (which compares the old/new names for equality) doesn't update

[sweet hornet]

The bossbar optimization has been removed in the latest version iirc

[high raptor]

would be nice to fix the problem inside virtual component itself instead of simply removing this specific bossbar case, but good to know, thanks 👍

[sweet hornet]

It was removed for another reason (translatable components being technically equals before translation): #adventure-help message

[cloud vapor]

feel free to submit a PR

[rare sage]

i'm not entirely sure how to feel about this, i know there is no contract or spec about this but is component == jsonDeser(jsonSer(component)) an invariant we care to maintain?

hm, no, that concept goes sideways the second you construct a component with modern features but have the serializer configured for older versions

[cloud vapor]

i don't really think component equality has ever been something we really consider tbh

[rare sage]

which is why i was asking

but anyway it already isn't something that happens

[ocean agate]

how to reset shadow color

nvm I use a empty component as root now

[sweet hornet]

Is there a Collector to join components from a Stream? I thought so, but I can't find it anymore. Or maybe I just mix up with Component#join

[rugged leaf]

Component.toComponent()

[sweet hornet]

Ahhhhh I was searching with "join". Thx!

[rugged leaf]

Yeah, I had a hard time finding it as well

[deft adder]

Hi, I'm having an issue with the MiniMessage format related to bold text. This is the format I'm using:
<b><gradient:#FFE259:#FFA751>ɪɴғɪɴɪᴛʏ ᴄʀᴀғᴛ</gradient></b> <gray>|</gray> <green>1.21 - 1.21.4</green> <white>Java & Bedrock</white><newline> <yellow>¡Únete a nuestro servidor Survival!</yellow>
The problem is that all the text becomes bold.

If I remove the bold text, everything works fine. This happens with the latest version of Adventure 4.23.0 — with version 4.21.0 it worked perfectly.

<gradient:#FFE259:#FFA751>ɪɴғɪɴɪᴛʏ ᴄʀᴀғᴛ</gradient> <gray>|</gray> <green>1.21 - 1.21.4</green> <white>Java & Bedrock</white><newline> <yellow>¡Únete a nuestro servidor Survival!</yellow>

[deft adder]

This only happens in paper, could it be a ProtocolLib problem?

[last horizon]

try without other plugins than your motd one

[obsidian nimbus]

shadowColor is nullable

[wispy folio]

hi, how can i display to the user to press a key to open his inventory, relative to his client side controls? and, preferably, when is that possible without adventure? ive been managing to stick to the bukkit api so far

[robust wharf]

Keybind component

MiniMessage has a tag for that

[sand drift]

You mean, you want to send them a message to hit their key for opening their inv? if so, you need a component because that stuff is configurable

[robust wharf]

It's not possible with legacy chat, you have to use components or MiniMessage

(the Bukkit project died a decade ago, the Bukkit API only exists in forks called paper-api and the legacy spigot-api)

[wispy folio]

does that keybind component work inside of itemstack lore?

oh wait im speaking bukkit lang 😨

[robust wharf]

I don't see why it wouldn't work there

[wispy folio]

how do I add an adventure component to the lore of a bukkit itemstack?

[robust wharf]

https://jd.papermc.io/paper/1.21.6/org/bukkit/inventory/meta/ItemMeta.html#lore(java.util.List)

But that's more a question for #paper-dev

[wispy folio]

but how can i add an adventure keybind component to it?

[robust wharf]

Have you used adventure before?

[wispy folio]

nope, should i go read the docs :P?

oh yeah TextComponents, sounds familiar to bukkit, is it compatible?

[robust wharf]

https://docs.papermc.io/paper/dev/component-api/introduction/

Bukkit never had text components

[wispy folio]

oops i mean net.md_5.bungee.api.chat.TextComponent

[robust wharf]

What you are referring to are the deprecated bungee components spigot added to spigot-api

[wispy folio]

i suppose i can set the itemmeta lore to either spigot textcomponent or the newer adventure one?

[sand drift]

I mean, we don't provide support for anything that isn't adventure these days

[wispy folio]

understandably

[robust wharf]

If you use legacy spigot methods we will try to convert it to adventure internally

[sand drift]

you would need to use an component API to do it, and anything using the legacy API will wipe out your thing

[silk scarab]

I'm kind of lost rn. When havign a setup like:

Component.text("lorem").color(firstColor).append(
    Component.text("ipsum").color(secondColor)
);

Everything works fine and the colors are applied correctly. But when instead of a Component#text, a Component#translatable is used, the second color isn't applied and the first one is full through. What could be the cause ofr this and is there a solution?

[fiery oracle]

What platform & version of adventure? Are you using any server-side translations?

[silk scarab]

Paper 1.21.4 build 226, Adventure 4.20.0 and yes I am using server-side translations. The translations are components too (but I think the color should still be applied after it's translated, atleast I hope that's the case). The use-case is for a customName of an entity

[robust wharf]

And show code

[silk scarab]

entity.customName(Component.text("[")
        .color(NamedTextColor.RED)
        .append(Component.text("Lvl." + level).color(NamedTextColor.BLUE))
        .append(Component.text("] ").color(NamedTextColor.RED))
        .append(Component.translatable(translationKey).color(NamedTextColor.BLUE))
);```
`entity` is a normal Entity and `level` an integer
For registering translations I use `TranslationStore#register(String key, Locale locale, T translation)` where `T` is a normal Component as my TranslationStore is `TranslationStore<Component>`

[ebon talon]

is there an easy way I can split Component into List of components? For example split on <br>

[fiery oracle]

there were a variety of issues with the component-based translation stuff that have since been fixed
you may be running into one of 'em

[fiery oracle]

no

[silk scarab]

Hm, okay. I was gonna update to 1.21.6 anyway when it's out of experimental, hopefully that will fix it then, thanks :)

[sour raft]

is there a way to support both Minimessage and legacy color code (&c)

[wet kestrel]

Not healthily. Legacy codes are over a decade dead.

[sweet hornet]

No, this is not intended

Explain why and we'll give you alternative

[wet kestrel]

If you have existing configs using legacy, a one-time deserialization to component and then serialization to minimessage is the way to go.

[sour raft]

I was thinking of making a simple chat plugin that supports both minimessage and legacy code

[wet kestrel]

It's 2025 - there's no need for new plugins to support stuff that Mojang moved away from in 2013.

[sour raft]

Mojang itself doesnt use Legacy color codes?

[wet kestrel]

There are a few spots that still support it because they're slow to remove things, but Mojang has been using components for chat since MC 1.7

[sour raft]

OH SHI alright thats uhh a long time ago

[tawny bolt]

You can string-replace the & codes with minimessage tags, if its just for player-facing chat stuff

I've never succeeded in teaching players Minimessage, so I get that

[sour raft]

Alright, Ill probably wont make it tbh.

Ty for the help tho guys!

[obsidian nimbus]

deserialize legacy, serialize to minimessage, replace \< with < and then deserialize from minimessage again

[wet kestrel]

That doesn't really work the way you think it does. Please don't support bad and incomplete approaches.

[sour raft]

uhhh 😅

[obsidian nimbus]

ik thats bad workaround

[sour raft]

I get what you mean but its fine ill not going to start on that project anyway and if i was i'd probably only support minimessage

[wet kestrel]

Legacy and components do not operate the same way, and users will be confused when your string-replace or mixed serializer approach causes formatting to be inconsistent.

[hardy estuary]

After how much time is usually an answer provided for an issue on github?

[amber quail]

How to create an persistent bossbar with Adventure? Like how bukkit does it.

[hardy estuary]

https://docs.advntr.dev/bossbar.html isn't already persistent? Never used it

[cloud vapor]

create a bukkit one if you want to use a bukkit one

[rare sage]

no component support :(

[cloud vapor]

i swear it does

wtf why doesn't it lmao

that's weird

[tawny bolt]

https://github.com/PaperMC/Paper/pull/9097 😦

[cloud vapor]

i think that's a fair reason to close it

but a decision does need to be made about whether or not to deprecate the old boss bars

[tawny bolt]

that would require the adventure bossbars to be persistent, interact with /bossbar and also be used for mobs (e.g. Boss#getBossBar). All three things you can't do with them right now

idk how adventure platform works nowadays, but I remember the concern being there that you can't have those things on e.g. Velocity, so you can't have them in the adventure API

[cloud vapor]

none of those things require adventure api changes

i also dont think persistency in the boss bars really matter

[tawny bolt]

it gets asked about in -dev sometimes, some people seem to want it. And thats how vanilla does it

[quaint monolith]

I suppose the only issue would be that the adventure API doesn't have any way to say if one should be persistent or not

So you would either need all persistent, none persistent, or add some additional API for managing which ones are persistent that is not a part of the actual boss bar API

[cloud vapor]

yeah, Bukkit.makeBossBarPersistent or w/e

we could add a flag for it in adventure but why, it's so easy to just persist them yourself as a dev

[quaint monolith]

In this case persistent doesn't just mean saving them to add again later, it also means vanilla things know they exist

[sleek flare]

// Converts a legacy string (using formatting codes) to a TextComponent
final Component component = LegacyComponentSerializer.legacyAmpersand().deserialize("&6Hello &b&lworld&c!");

I found this on the wiki, does this still work? It says:

Required type:
net. kyori. adventure. text. TextComponent
Provided:
me. clip. placeholderapi. libs. kyori. adventure. text. TextComponent

[sour totem]

you imported the wrong textcomponent class

[sleek flare]

So I should be using
me. clip. placeholderapi. libs. kyori. adventure. text. TextComponent
?
Because all of my code uses this one net. kyori. adventure. text. TextComponent

Is there a way to convert between the two?

[graceful kestrel]

you need to use net.kyori.adventure

not me.clip.placeholderapi

[fiery oracle]

you're importing placeholder api's shaded version of adventure somewhere

the full error should tell you

[sleek flare]

Ahhh, the LegacyComponentSerializer was the wrong version. thank you both

[proper sandal]

which java version is adventure built against?

[prisma mason]

seems to be a minimum of 17

tests are also done on 8 it seems like

[cloud vapor]

8

[proper sandal]

awesome, thanks

[prisma mason]

https://github.com/KyoriPowered/adventure/blob/1c16bdd2ec4a1f1b115e4c0c5acb0d1376a37f32/build-logic/src/main/kotlin/adventure.base-conventions.gradle.kts#L10
Then why is this set to 17

[cloud vapor]

A toolchain is just what java version is required to build and run the project - the actual code that is built is Java 8 language level

indra defaults to 8 for the target, which is why you dont see it explicitly being mentioned in that build file

[prisma mason]

I see, never looked into Indra really

[proper sandal]

how come the component logger isn't an audience?

[rare sage]

it's just an slf4j logger that accepts components

it doesn't make sense to make it an audience, log at what log level? should it use any markers? what format?

[elder torrent]

If I add Adventure 4.22.0 to my plugin's gradle.build like in this page (https://docs.advntr.dev/getting-started.html), will it override Adventure 4.21.0 currently bundled with Paper and let me access newer Adventure features?

It doesn't seem like my plugin can override the bundled Adventure version because the feature I tried isn't there. Is there a way to check what Adventure version my plugin is currently using once the server starts?

[sterile rampart]

no

[elder torrent]

So I just have to wait for Paper to update their Adventure version, right?

[sterile rampart]

yes

[prisma mason]

can also build your own Paper JAR with the updated Adventure, don't know what the side effects of doing this would be

[sweet hornet]

Is there a way to use the built-in date (or any other) formatter with dynamic tag? I think what I did would work, but it looks a bit hacky, so I'd like to have your thought about it...

public static final MiniMessage MM = MiniMessage.builder()
        .editTags(t -> t.tag("event_start", (args, ctx) -> {
            String name = args.popOr("event name expected").value();
            return Formatter.date("event_start", getEventStart(name)).resolve("event_start", args, ctx);
        }))
        .build();

// Omitted for simplicity
private static LocalDateTime getEventStart(String eventName) { /* ... */ }

As you may guess, there will be a similar event_end tag as well.
There can be multiple events simultaneously and their name is not restricted at all, so I'd really like to keep them as an argument of a single tag resolver

The main advantage would be to not have to duplicate any date formatting logic in my own code, since there's already a tag resolver for that.
Typical usages would be:

<gray>The <gold>mining</gold> event is starting at <red><event_start:mining:'HH:mm'></red></gray>

<green>Summer event will be live between <b><event_start:'Summer 2025':'yyyy-MM-dd'></b> and <b><event_end:'Summer 2025':'yyyy-MM-dd'></b></green>

[cloud vapor]

Well the date formatting logic is literally just calling SimpleDateFormat so I wouldn't worry too much about duplicating it tbh

[sweet hornet]

Hmm, in a way, yes, but in the same time, there's the argument handling logic (which is also not that much I agree, but not much + not much = well, a little bit much)
Furthermore, there is the same question for other resolvers, I guess (number being the first one that comes to my mind, but choice or boolean also maybe)

Example that comes to my mind for the number formatter:
Inserting a score of a player (not from the vanilla scoreboard thing, a score from an hypothetical plugin), for a leaderboard.
Raw text would be:

<gold>1st: <name:1> (score: <score:1>)
<gray>2nd: <name:2> (score: <score:2>)
<#CD7F32>3rd: <name:3> (score: <score:3>)

And the score tag would fetch the value at the given rank, and one may want to take advantage of built-in number formatter tag features (such as localization, decimal management...)

And just while I was writing this example, I realized that it could actually be a cleaver way to ease the creation of leaderboard holograms in my leaderboards plugins, so the example is not that hypothetical (at least, in the next few months, when I'll have time to work on that plugin 😅)

[sweet hornet]

Side question, I just want to be sure this is not my (mis)understanding of English that is fooling me:

https://github.com/KyoriPowered/adventure/blob/main/4/text-minimessage/src/main/java/net/kyori/adventure/text/minimessage/tag/resolver/Formatter.java#L100
Here the JavaDocs says (L.93):

This replacement is auto-closing, so its style will not influence the style of following components.
However (L.103):

return Tag.inserting(...); // Shouldn't it be Tag#selfClosingInserting instead then?

(Same applies to all the formatters, hence why I think it is just a misunderstanding on my side rather than an actual error in every formatter)

[fiery oracle]

https://github.com/KyoriPowered/adventure/issues/960

[sweet hornet]

Then let’s make a PR and we’ll see, I guess 😅

[fathom mulch]

where can i see all the translation keys and their english values?

[sand drift]

you would need to grab the english locale file

[tawny bolt]

https://github.com/misode/mcmeta/blob/assets/assets/minecraft/lang/en_us.json

(or open your Minecraft .jar file with a zip tool)

[carmine steppe]

Hey I am working with the localisation feature with MiniMessageTranslator. It works well, but I am encoutering issues with sub-translations.
For example, my.key=hello <value> where value is just a number or player name is ok.
but my.key=You have <time_remaining> where time_remaining links back to time.format.days=<day> days. My MiniMessageTranslator get call for my.key but gets no calls for time.format.days even though its passed through as an argument and is a TranslatableComponent.
Then I end up with "You have <missing: time.format.days>"

Are subtranslations not supported? I couldn't find any info about them on the docs.
thank you, any pointers are appreciated.

[cloud vapor]

This is fixed in Adventure 4.22.0

[carmine steppe]

oh perfect, thank you. 🙏

[tough peak]

is there a way to override style of a tag resolver?
I mean
lets say I have this string: <red>Hello <player_name>
player_name is replaced with a compoent of the player's display name.
and I can't make the player's name red

[cloud vapor]

If the component with the display name has colour, then you'd want to strip the colour beforehand or use a plain string tag or smth

[tough peak]

    private Component getChatMsg(String groupName, Player player, Component displayName, Component message) {

        String msg = configManager().getGroupMsgTemplate(groupName);
        TagResolver tagResolver = TagResolver.builder()
                .tag("player_display_name", Tag.selfClosingInserting(displayName))
                .tag("message", Tag.selfClosingInserting(message))
                .build();
        if (msg == null) msg = "";

        return miniMessage().deserialize(msg, TagResolverHandler.finalPlayerTagResolver(player), tagResolver);
    }

this is my current code

[cloud vapor]

my answer doesn't change after seeing the code

[tough peak]

yk

[tough peak]

but how I replace it without making it a component

[cloud vapor]

huh

either remove the color from the displayName component, or plain text the display name and use a string inserting tag

[sweet hornet]

You could make your own <unstyle> tag to strip style from its children (allowing it to inherit the one from its parent then)

It would be fairly simple using Modifying

[tough peak]

.tag("player_display_name", Tag.selfClosingInserting(displayName.style(Style.empty())))

well this doesnt work

how can I strip and then insert

[sweet hornet]

Create a custom tag implementing Modifying like I said above.
Let me know if you need help for the implementation (I just tested it, it works, with 1 line of code, or 2 if you consider the tag registration in your MM instance)

[sand drift]

components a trees

you would probably want to just use the plain text serialiser

[tough peak]

and then how I would insert it to the text?

[sand drift]

slap it in a text component?

[tough peak]

 .tag("player_display_name", Tag.selfClosingInserting(Component.text(PlainTextComponentSerializer.plainText().serialize(displayName))))

like that?

cause that doesnt work

[sand drift]

idk about the selfClosing part, but, that would work in terms of the component part

"doesnt work"

[cloud vapor]

maybe self closing doesnt work like that

i cant remember

[tough peak]

inserting also doesnt work

its like it sill has a style

[sand drift]

You're not using malformed components, no?

[tough peak]

what is that

[sand drift]

what is the strings contents after you run it through the serialiser

Using legacy cruft inside of text components

[tough peak]

I guess not

I use <>

wait I think I know how to solve

[sweet hornet]

Show us the output of PlainTextComponentSerializer.plainText().serialize(displayName), just out of curiosity plz

[tough peak]

ok

stivendarsi (player name)

[sweet hornet]

Ok, so it has no style at all, thus if you really executed the code you shown, it should have worked. What is the result (what color is the name)?

[tough peak]

white

I am using hex color from a placeholder to set the color - I think thats where the problem comes from

[tough peak]

tho the color it self is translated somehow

Its like it doesnt apply to the text after the placeholder

oh

wait

wait

[sweet hornet]

Can you please confirm that this works:

    private Component getChatMsg(String groupName, Player player, Component displayName, Component message) {
        return miniMessage().deserialize("<red>Hello <player_name>", Placeholder.component("player_name", displayName));
    }

[tough peak]

ahh

thats why 💔

[sweet hornet]

What are you talking about?

[tough peak]

We learn new things every day - and that's how we build up

[tough peak]

I used selfClosingInserting to phrase the placeholders

[tough peak]

so its like closing

[sweet hornet]

Yes, but there's nothing after your placeholder

So it makes no difference in the example you shown

[tough peak]

wait will try

[cloud vapor]

selfClosing just means it closes itself

e.g. style from that tag won't effect subsequent tags

[sweet hornet]

UNLESS, you shown us something that is not what you really try, which is like asking help on a math question with a different formula: pointless and a waste of time for everyone

[cloud vapor]

yep

that's what it smells like

[tough peak]

well I am having trouble with a problem

and I don't know what causing it

[cloud vapor]

please provide an MRE and we can look into it

[tough peak]

ok I fixed it

[tough peak]

🔥

[cloud vapor]

ofc it was papi tags

that means legac haha

so you were inserting legacy into the resulting component or smth

[tough peak]

The phrased papi was this text "<#74cafc>"

[cloud vapor]

that would not even have turned into a colour with your old code

[sweet hornet]

I think he inserted (self closed) empty component containing just a color

Thus, it was not affecting the following

Anyway, it was a completely unrelated issue because he did not actually tested the example he shown us

[tough peak]

Bruh I did

[fossil prawn]

Clicking components does not work and I updated to the latest version. This seems to only be happening in .6

I don't see anything in changelogs, just curious if this is a known issue or not

I'm not seeing anything in github issues either

cc @tawny marsh

[tawny marsh]

specifically uh click for command suggestions

[wet kestrel]

What platform are you using it on?

[full onyx]

can u make custom tags? so for example if I put <hi> it puts "Hi" in the string

    private static final TagResolver RESOLVER =  TagResolver.builder().tag("hi", Tag.inserting(text("HI"))).build();
    private static final MiniMessage MINI_MESSAGE = MiniMessage
            .builder()
            .tags(RESOLVER)
            .preProcessor(str -> "<!i>" + str).build();``` maybe this

would it make it laggy? 👀

[full onyx]

well uh that was something

Maybe I can just add the tags I need

[full onyx]

If its possible can I use reflect and only keep the resolver for default Minecraft colors

I saw all of these are private classes

💀 it works ig

I swear if there is easy way to do it I will cry

[sterile rampart]

look at the methods?

[full onyx]

I did 2 mins ago and my heart is broken

[fossil prawn]

paper 1.21.6 latest release

I’ve had two people complain about it so far

[median gust]

Hi, seem to be running into some weird issue on 1.21.6 with the Translator
Causes a StackOverflow for some reason
I am gonna guess it relates to my system for intercepting packets to translate stuff, but it has not happened on earlier versions

full log; https://mclo.gs/ufyoImR

[median gust]

tested by removing most fancy formatting and it just complains whenever i try and get the translation string for an argument

[restive spruce]

seems to be the same problem as #paper-help message

[median gust]

ok nice, fixed it temporarily by just not registering it for 1.21.6 servers, ill keep an eye on that

[cloud vapor]

ong what the hell are you doing, they are all public - StandardTags.color()

[full onyx]

I saw that very late 😭

[cloud vapor]

are you not just registering a translator that calls itself

[full onyx]

its fine now :)

I just removed all the tags I didn't need so its slightly faster

[cloud vapor]

It would be picoseconds faster if that

A complete non issue

[full onyx]

probably

[cloud vapor]

Not probably, definitely

[full onyx]

should I benchmark it

[cloud vapor]

We already have benchmarks

[median gust]

no its called outside of the class as a static method basically

whenever it enters the translate method and finds a key it just infinitely loops, hence i just dont register it and it works

[cloud vapor]

definitely looks like an issue with the recursive rendering pr

im not going to be able to take a look this weekend, but if you wanna have a little adventure adventure to see if you can figure it out that'd be super!

this pr in particular https://github.com/KyoriPowered/adventure/pull/1226

although i dont know why it wouldn't have been caught in any of our unit tests

[median gust]

in all likelyhood its some combo of some scuff in my packet interception to transform a component + translation being recursive

tried printing out the component and the first here is likely the original one, and after its being kinda funky

[median gust]

okay yeah made a simple plugin that just registers it and it just recursives it with no modifications or anything

[wet kestrel]

Using paper API directly or using adventure-platform-bukkit?

[fossil prawn]

adventure platform bukkit 4.4.0

[wet kestrel]

It doesn't support .6. See blood's PR, and be sure to use #adventure-platform-mod-help in the future 🙂

[fossil prawn]

Will do, thank you!

[fiery oracle]

shouldn't you be returning null instead of the component here, when no translation is done?
afaict it's infinitely recursing because returning a component indicates a translation was done, and translated components may need further translation, so it's passed back to the global translator

should probably remove the recursion there and introduce a limit, though

[rare sage]

your translator is flawed yes

[median gust]

i see so returning the same just means it thinks it was translated into another, returning null indicates it should be "the same"?

[rare sage]

the translation process will keep checking as long as there are translatables

null means there is no translation

[fiery oracle]

returning null means "skip this translator"

the javadocs are a bit obtuse here since that bit is mentioned only for the string key one https://jd.advntr.dev/api/4.23.0/net/kyori/adventure/translation/Translator.html#translate(java.lang.String,java.util.Locale)

but returning null for the TranslatableComponent overload means to call the string one, and the string one returning null means "skip this translator"

[median gust]

that seemed to fix yeah my bad..
guess it worked before cuz it wasnt recursively doing it
thanks

[frozen tree]

Does adventure support setting the Component name/lore of items? Like I know paper stores it as components, but I would imagine this is only supported on the latest paper versions.
So would it be possible to set the name/lore of items via the adventure lib aswell, or would I need to convert them into minecraft components using the API?

[sweet hornet]

Adventure does not know about items

What is your underlying question? Please provide some context

[frozen tree]

Im basically trying to set the item name/lore as components using MiniMessage format, but since only the latest versions of paper support it Im asking if there is a way to set the name/lore of items on older versions of paper aswell, for example using Adventure#setLore(ItemStack, Component)

[wet kestrel]

It's not only the latest versions of paper. It's been there for many, many years. Are you talking about super ancient servers (that aren't typically worth supporting anyway) ?

[graceful kestrel]

No because adventure doesn't known about bukkit itemstacks

[frozen tree]

I mean pre 1.19, yeah

Good enough then, I saw adventure has this handy method: MinecraftComponentSerializer.get().serialize() so I might use that

[wet kestrel]

Not worth your time, honestly. Nearly no servers run it, and certainly no servers that actually download new plugins.

But, if you really want to add component support properly to versions prior to adventure-in-paper (which I think is actually closer to 1.16?) you'd need to poke internals.

[rare sage]

yeah paper has had native adventure integration since before 1.19!

[sweet hornet]

Which was in June 2022 (3 years ago), fyi

[harsh night]

Was markup support removed from MiniMessage?

[robust wharf]

Ages ago, ye

Or was it moved to a module? I don't remember

[harsh night]

I haven't used this library since probably 1.17 so makes sense lol

[harsh night]

I don't see anything in the docs so I was wondering

Since the only options now are tags, does this mean that tags have to be in the format of <{tag_name}>, and I can't just add ** for bold?

oh wait I can do a preProcessor to change ** to <bold>, maybe? hmm...

[wet kestrel]

For survey purposes, may I ask why you wish to use markdown when users are getting used to minimessage tags?

[harsh night]

efficiency and familiarity purposes

** is a lot easier than <bold>

and familiarity via Discord

[robust wharf]

MiniMessage is widely used, I would expect most server owners to know it by now

[harsh night]

I'm using it for chat formatting

so this is for players

[robust wharf]

Better to have a dedicated parser for this then

[sand drift]

There is minedown if you just want to have markdown support

[robust wharf]

The preprocessing approach is what that old markdown module used and it was really cursed and bad

[harsh night]

do my own parsing on the string to replace markdown with tags, then shove it into minimessage?

[robust wharf]

I personally wouldn't mix it

Straight into components

[frozen tree]

Would there be any way to see if the server(paper) supports native MiniMessage support or would I just use reflection to check if its a supported method?

[hallow tangle]

<@&748618676189528155>

[delicate forge]

too slow

[green monolith]

sniped 😦

[green monolith]

At least I wasn't that slow 😛

[grave fox]

Mobile

Honestly don't know why I even try

[cloud vapor]

Just check if the class exists

or check the server version

[frozen tree]

So which version to check for?

And what class?

I mean, CraftMetaItem exists even on bukkit so

[cloud vapor]

uh, the MiniMessage class lol

like Class.forName or just run it and catch a ClassNotFoundException

many ways

[frozen tree]

MiniMessage itself doesn't support setting item meta, atleast that was what was said here

[wet kestrel]

On the legacy spigot loader another plugin loading in minimessage would cause trouble if you're looking to use methods that take components.

[frozen tree]

Yea, so I want to check if those methods are available or not

[wet kestrel]

Check if the method exists, then.

[frozen tree]

I might just use reflection then

[wet kestrel]

If your plugin will only work where supported, could just compile in a use of it in a try/catch during onenable and bail out with a helpful "your server is in ancient history" message (or don't try/catch it and just error out)

[frozen tree]

We have a multi module plugin, so I want to check which module to initialize

[frozen tree]

Is there a way to translate the PAPI placeholders inside a Component without recompiling it? Like not having to parse it to a string, translate it, and parse it again to a Compoment

[sweet hornet]

"recompiling"?

[frozen tree]

Edited my message

[graceful kestrel]

https://docs.advntr.dev/faq.html#how-can-i-use-bukkits-placeholderapi-in-minimessage-messages

[frozen tree]

This example assumes you know the placeholder you are parsing (Also not what Im looking for)

[graceful kestrel]

..?

ah