Twilight is an API for developers creating plugins for Spigot or Paper based Minecraft servers. It contains a wide range of utilities and QOL improvements, from inventories, to schedulers and databases.
Twilight is built using Kotlin, and is recommended for usage with. Many features of Twilight should work with plain Java, though compatibility is not guaranteed.
For support, questions or to chat with the team, come join the Discord:
Twilight should be bundled within your plugin. Add the following repository and dependency to your build tool:
Maven
<repository>
<id>flyte-repository-releases</id>
<name>Flyte Repository</name>
<url>https://repo.flyte.gg/releases</url>
</repository>
<dependency>
<groupId>gg.flyte</groupId>
<artifactId>twilight</artifactId>
<version>1.1.15</version>
</dependency>
Gradle (Groovy DSL)
maven {
url "https://repo.flyte.gg/releases"
}
implementation "gg.flyte:twilight:1.1.15"
Gradle (Kotlin DSL)
maven("https://repo.flyte.gg/releases")
implementation("gg.flyte:twilight:1.1.15")
Certain features of Twilight require configuration, which can be done via the Twilight class. To setup a Twilight class instance, you can use the twilight
function as shown below:
val twilight = twilight(this)
If you want to make use of environment variables (.env files), you can configure your usage of these here, like so:
val twilight = twilight(this) {
env {
useDifferentEnvironments = true
devEnvFileName = ".env.dev"
prodEnvFileName = ".env.prod"
}
}
The above are the default env configuration settings.
If you useDifferentEnvironments
, you'll need a .env
file which contains the following:
ENVIRONMENT=DEV # or PROD
This file determines whether to use your dev .env or your prod .env.
If you do not use this different environments feature, then it will just use the .env (or whatever you specify the name as with prodEnvFileName
)
Throughout your project you can use Environment.get("VARIABLE")
to retrieve a value from your environment variables.
Other features that can be configured in the Twilight class builder will have their own sections later in the README.
Twilight takes advantage of Kotlin's extension functions to add additional functions to various classes used within the API. Many of these are convenience functions, and some add complete new functionality. To see all of the functions added, view them inside the code.
We have a neat way to handle code for events, which register by themselves so you don't have to!
You can make use of it like so:
event<PlayerJoinEvent> {
player.sendMessage("Welcome to the server!")
}
If you need to change the eventPriority
or the ignoreCancelled
, you can pass it to the function call like:
event<PlayerJoinEvent>(EventPriority.HIGHEST, true) {
player.sendMessage("Welcome to the server!")
}
If you ever need an instance of the Listener class that the event gets put in to, it's returned by the function. Specifically, it returns TwilightListener
. This class has a convenience function in for unregistering the listener, it can be used like so:
val listener = event<PlayerJoinEvent> {
player.sendMessage("Welcome to the server!")
}
listener.unregister()
Instead of having to extend the org.bukkit.event.Event yourself, you can extend TwilightEvent!
Here's an example:
class MyCustomEvent : TwilightEvent() {
}
This is much easier than the standard Bukkit Event, as you don't have to worry about defining handles, etc.
The TwilightEvent also includes a timestamp which provides an Instant for when the event was ran:
// with the above event example
event<MyCustomEvent> {
println("time event was executed: $timestamp")
}
You can also declare that it should be asyncronous by passing a true
value to the TwilightEvent constructor:
class MyCustomEvent : TwilightEvent(true) {
}
class MyCustomEvent(async: Boolean) : TwilightEvent(async) {
}
Twilight provides additional events which are not found in Spigot or Paper. These are:
- PlayerMainHandInteractEvent
- PlayerOffHandInteractEvent
- PlayerOpChangeEvent
- PlayerOpEvent
- PlayerDeopEvent
- ChatClickEvent (see below)
You can opt out of Twilight calling these events. For example:
disableCustomEventListeners(OpEventListener, InteractEventListener)
Due to limitations imposed by the Minecraft server software, when interacting with a clickable message in chat or in a book the only response options are RUN_COMMAND
, SUGGEST_COMMAND
, CHANGE_PAGE
, COPY_TO_CLIPBOARD
, OPEN_FILE
and OPEN_URL
. None of these match the most common use case: running custom code. Twilight utilizes the RUN_COMMAND
response to call a custom ChatClickEvent
which can be listened to like a regular event.
To use this feature, where you would normally build your clickable message, use the Twilight extension functions to add a custom click event. Twilight will then redirect any data which you put in the parameters to be accessable as a variable from within the ChatClickEvent
when the player clicks the message.
For Paper/Adventure (recommended):
import net.kyori.adventure.text.Component
Component.text("Click here")
.customClickEvent("openGUI", "warps")
Or for Spigot/BungeeCord:
import net.md_5.bungee.api.chat.TextComponent
TextComponent("Click here")
.customClickEvent("openGUI", "warps")
From there, simply listen to the event as normal, and access the data attached to the message the player has clicked. In this basic example, information to open a "warps" GUI has been passed through as the custom data, and so the correct action can be taken:
event<ChatClickEvent> {
if (data.size != 2) return@event
if (data[0] != "openGUI") return@event
when (data[1]) {
"warps" -> GUIManager.openWarps(player)
..
}
}
Bukkit's built in scheduler is tedious at best, so Twilight takes advantage of beautiful Kotlin syntax to make it easier to write, as well as adding a custom TimeUnit to save you calculating ticks.
How to schedule a single task to run on Bukkit's main thread either sync or async: `
sync {
println("I am a sync BukkitRunnable")
}
async {
println("I am an async BukkitRunnable")
}
How to schedule a delayed task, with optional custom time unit and async parameters:
// SYNC
delay {
println("I am a sync BukkitRunnable delayed by 1 tick")
}
delay(10) {
println("I am a sync BukkitRunnable delayed by 10 ticks")
}
delay(1, TimeUnit.SECONDS) {
println("I am a sync BukkitRunnable delayed by 1 second")
}
// ASYNC
delay(10, true) {
println("I am an async BukkitRunnable delayed by 10 ticks")
}
delay(1, TimeUnit.SECONDS, true) {
println("I am an async BukkitRunnable delayed by 1 second")
}
How to schedule a repeating task, with optional custom time unit and async parameters:
// SYNC
repeat(10) {
println("I am a sync BukkitRunnable running every 10 ticks")
}
repeat(10, TimeUnit.SECONDS) {
println("I am a sync BukkitRunnable running every 10 seconds")
}
repeat(5, 10) {
println("I am a sync BukkitRunnable running every 10 ticks, waiting 5 before starting")
}
repeat(5, 10, TimeUnit.SECONDS) {
println("I am a sync BukkitRunnable running every 10 seconds, waiting 5 before starting")
}
// ASYNC
repeat(10, true) {
println("I am an async BukkitRunnable running every 10 ticks")
}
repeat(10, TimeUnit.SECONDS, true) {
println("I am an async BukkitRunnable running every 10 seconds")
}
repeat(5, 10, true) {
println("I am an async BukkitRunnable running every 10 ticks, waiting 5 before starting")
}
repeat(5, 10, TimeUnit.SECONDS, true) {
println("I am an async BukkitRunnable running every 10 seconds, waiting 5 before starting")
}
Twilight
repeat
conflicting with Kotlin'srepeat
? As an alternative, you can userepeatingTask
.
You can also chain runnables together using onComplete
to nicely nest sync/async executions. Here's an example:
async {
println("I am an async BukkitRunnable called Atom")
}.onComplete() {
println("I am an async BukkitRunnable called Brandon running immediately after Atom finishes executing")
}.onCompleteSync(10) {
println("I am a sync BukkitRunnable called Charlie running 10 ticks after Brandon finishes executing")
}.onComplete(5, TimeUnit.SECONDS) {
println("I am a sync BukkitRunnable called Dawson running 5 seconds after Charlie finishes executing")
}.onCompleteAsync {
println("I am an async BukkitRunnable called Enid running immediately after Dawson finishes executing")
}
As you can see, you can specify whether sync/async (if unspecified, it will not change) and you can pass in an optional delay.
This also works with a delay from the get-go:
delay(20, TimeUnit.SECONDS) {
println("I am a sync BukkitRunnable delayed by 20 seconds")
}.onCompleteAsync(10, TimeUnit.SECONDS) {
println("I am an async BukkitRunnable delayed by a further 10 seconds")
}
Currently, onComplete is incompatible with repeating tasks.
Creating GUI's can be an incredibly long and tedious process, however, in Twilight we offer a clean and efficient way to build GUIs.
Here's an example of a simple, bog-standard GUI:
val basicGui = gui(Component.text("Click the apple!"), 9) {
set(4, ItemStack(Material.APPLE)) {
isCancelled = true
viewer.sendMessage(Component.text("This is an apple!"))
}
}
player.openInventory(basicGui)
As you can see, setting the click event logic has never been easier. You can reference the player at any time using viewer
.
Here's an example of a gui implementing the pattern feature, making it much easier to visualise:
val complexGui = gui {
pattern(
"#########",
"# S #",
"#########"
)
set('S', ItemStack(Material.PLAYER_HEAD).apply {
val meta = itemMeta as SkullMeta
meta.displayName(Component.text("${viewer.name}'s Head!"))
meta.owningPlayer = viewer
itemMeta = meta
}) {
isCancelled = true
viewer.sendMessage(Component.text("Ouch!", NamedTextColor.RED))
}
set('#', ItemStack(Material.LIGHT_GRAY_STAINED_GLASS_PANE).apply {
val meta = itemMeta
meta.displayName(Component.empty())
itemMeta = meta
}) { isCancelled = true }
}
player.openInventory(complexGui)
You can, of course, also implement GUIs for other inventory types:
val dropperGui = gui(Component.text("Title"), 9, InventoryType.DROPPER) {
// You can also set multiple indexes at once
set(listOf(1, 3, 4, 5, 7), ItemStack(Material.GRAY_STAINED_GLASS_PANE))
}
player.openInventory(dropperGui)
To open the gui, simply do Player#openInventory(GUI)
like shown in the examples.
Currently we have support for MongoDB. To configure it, you can take one of two routes:
You can use the following Environment variables for your MongoDB:
MONGO_URI="your URI string"
MONGO_DATABASE="your database name"
When building your Twilight instance, you can specify your URI and database like so:
val twilight = twilight(plugin) {
mongo {
uri = "your URI string"
database = "your database name"
}
}
From here you can use the following function to get a collection from your database:
MongoDB.collection("my-collection")
And use the standard features of the Mongo Sync Driver with your MongoCollection
.
OR you can use some of our custom features, making communicating with a Mongo database infinitely easier. Here's how you do it:
class Profile(
@field:Id val id: UUID,
val name: String
) : MongoSerializable
What's happening here? We're declaring what should be used as the key identifier for our class in the database, we can do so by annotating a field with @field:Id
.
We also implement an interface MongoSerializable
. This gives us access to a bunch of methods which make our lives really easy when it comes to moving between our class instance and our database.
For example, I could do the following:
val profile = Profile(UUID.randomUUID(), "Name")
profile.save()
// this returns a CompletableFuture of the UpdateResult
// or we could do profile.saveSync() if we need it to be sync, does not return wrapped by a CompletableFuture
profile.delete()
// this returns a CompletableFuture of the DeleteResult
// same as save, can do profile.deleteSync() for sync, does not return wrapped by a CompletableFuture
If we ever want to find and load and instance of our class from the database, we can use some functions from the TwilightMongoCollection:
val collection = MongoDB.collection<Profile>() // by default this assumes the name of the collection is the plural camel case of the type, f.x. Profile -> profiles, SomeExampleThing -> someExampleThings
// you can specify the name of the collection if you wish it to be different like so
val collection = MongoDB.collection<Profile>("myCollection")
collection.find() // returns a CompletableFuture<MongoIterable<Profile>>
collection.find(BsonFilter) // returns a CompletableFuture<MongoIterable<Profile>>
collection.findById(id) // id must be the same type as the field marked as the id on the class, returns a CompletableFuture<MongoIterable<Profile>>
collection.delete(BsonFilter) // returns a CompletableFuture<DeleteResult>
collection.deleteById(id) // id must be the same type as the field marked as the id on the class, returns a CompletableFuture<DeleteResult>
// all of these have sync versions which follow the same pattern, f.x. collection.findSync(), where the return value is the same as the async version, just not wrapped by a CompletableFuture
If we need something that isn't already wrapped by the TwilightMongoCollection, it exposes us the MongoCollection of Documents, which we can get with collection.documents
.
To get started you need to create an instance of the SQL Wrapper like so
val db = SQLWrapper(url = "your db url", user="user", password="password")
db.connect()
The QueryBuilder class will help you in creating everything from simple queries like SELECT's to even complex JOIN's. All you need to start is an instance of the query builder, here's an example usage
val queryBuilder = QueryBuilder()
// Example SELECT query
val selectQuery = queryBuilder.select("id", "name").from("person").where("age" gt 18).build()
// Example INSERT query
val insertQuery = queryBuilder.insertInto("person", "id", "name").values(1, "John").build()
// Example UPDATE query
val updateQuery = queryBuilder.update().table("person").set("name", "John Doe").where("id" eq 1).build()
// Example DELETE query
val deleteQuery = queryBuilder.delete().table("person").where("id = 1").build()
If you would like to retrieve and store data as objects within your database there a some methods provided for this
1 - Your object must implement SQLSerializable
2 - You must have a table that fits the structure of your object, you can create by calling convertToSQLTable()
on your object, then execute the statement like so
// NOTE: convertToSQLTable() takes a optional dialect parameter, at this time the only additional dialect is postgres
val createTable = yourObjectInstace.convertToSQLTable()
if(db.execute(createTable)) {
// successfully executed
}
3 - To insert your object call toInsertQuery()
like so
val insertToTable = yourObjectInstace.toInsertQuery()
if(db.execute(insertToTable)) {
// successfully executed
}
4 - To retrieve objects from your database you call a select statement like normal but call toListOfObjects<Type>()
on the returned Results
class
Once you have your query using either the QueryBuilder or your own you can run it like so
val result = result.executeQuery(selectQuery)
once you have run the query it will return a Results
class, it can be used like so
result?.let { res ->
println("MyColumn Value: " + res["my_column"])
}
The results class contains a list of all the rows and columns returned by the database. You can access any of the columns values the same way you would with a map.
There is a basic ternary operator implementation added which can be used like so:
val test = false
println(test then "yes" or "no")
This doesn't yet work for evaluating functions either side of the ternary though, we plan to figure this out in the near future.
Twilight can do the heavy lifting and query the Mojang API to find the UUID from name or name from UUID of a player, particularly useful for networks. Twilight will cache responses in an attempt to not break the rate limit imposed by Mojang.
If you have a UUID and you want to get a name, you can call nameFromUUID
:
NameCacheService.nameFromUUID(UUID.fromString("a008c892-e7e1-48e1-8235-8aa389318b7a"))
This will look up your cache to see if we already know the name, otherwise we will check the MongoDB "cache" of key, value pairs, and finally, we'll query Mojang if we still don't know it.
After each step the key, value pair will be stored so the next call is just on the cache.
Similarly, if you have a name and want to get the UUID, you can call uuidFromName
:
NameCacheService.uuidFromName("stxphen")
Currently the only way to configure your MongoDB "cache" for UUIDs and names, is to have an Environment variable called NAME_CACHE_COLLECTION
with the value being what you want to call the collection. Don't want to use the Mongo cache? Disable useMongoCache
in the settings.
Twilight has a Redis system that lets you set/get/delete string key value pairs, additionally, you are able to publish messages and listen to incoming messages on any channel you'd like.
You can use the following Environment variables for your Redis Server:
REDIS_HOST="your redis server host"
REDIS_PORT="your redis server port"
REDIS_TIMEOUT="your redis connection timeout"
REDIS_AUTHENTICATION="NONE"
Alternatively, if your Redis server requires a Username + Password in order to access, you can use the following:
REDIS_HOST="your redis server host"
REDIS_PORT="your redis server port"
REDIS_TIMEOUT="your redis connection timeout"
REDIS_AUTHENTICATION="USERNAME_PASSWORD"
REDIS_USERNAME:"coolUsername"
REDIS_PASSWORD:"coolPassword"
Alternatively, if your Redis server requires a URL in order to access, you can use the following:
REDIS_HOST="your redis server host"
REDIS_PORT="your redis server port"
REDIS_TIMEOUT="your redis connection timeout"
REDIS_AUTHENTICATION="URL"
REDIS_URL="coolURL"
When building your Twilight instance, you can specify your host and port like so:
val twilight = twilight(plugin) {
redis {
authentication = Authentication.NONE
host = "your redis server host"
port = 6379 // Default Redis Port
timeout = 500 // 500 Milliseconds Timeout
}
}
Alternatively, if your Redis server requires a Username + Password in order to access, you can use the following:
val twilight = twilight(plugin) {
redis {
authentication = Authentication.USERNAME_PASSWORD
host = "your redis server host"
port = 6379 // Default Redis Port
timeout = 500 // 500 Milliseconds Timeout
username = "coolUsername"
password = "coolPassword"
}
}
Alternatively, if your Redis server requires a URL in order to access, you can use the following:
val twilight = twilight(plugin) {
redis {
authentication = Authentication.URL
host = "your redis server host"
port = 6379 // Default Redis Port
timeout = 500 // 500 Milliseconds Timeout
url = "coolURL"
}
}
You can Set/Get/Delete String Key-Value pairs on your Redis server like so: (All of those functions are Async and return a CompleteableFuture)
Redis.set("cool-key", "super-secret-value")
val future = Redis.get("cool-key") // Returns a Completable Future
future.thenApplyAsync {
value -> println("The value is: $value") // Prints: "The value is: super-secret-value"
}.exceptionally {
e -> println("An exception occurred: ${e.message}") // Handle the Exception
}
Thread.sleep(1000)
Redis.delete("cool-key")
You can publish messages like so:
Redis.publish("channel", "message") // Async Publishing
You are able to listen to incoming message on specific channels, using the 'TwilightRedisListener' Class:
// Extend the 'TwilightRedisListener' class and override the 'onMessage' function.
class PlayerConnectionRedisListener(): TwilightRedisListener("player-connection") {
override fun onMessage(message: String) {
// do stuff
}
}
You can add add/register the listener like this: (which also returns the listener which lets you unregister it if you'd like)
val listener = Redis.addListener(PlayerConnectionRedisListener())
listener.unregister() // unregistering the listener.
Alternativley, instead of extending the listener class, you can add a listener using a block of code, which returns the 'RedisMessage' data class, which contains the channel, the message, and the listener:
val listener = Redis.addListener("cool-channel"){
println("The following message was received: '$message' on channel '$channel'")
this.listener.unregister() // unregistering the listener after we recieved the message.
}
You can easily get the hash of a file using this method. The parameter algorithm
is used to define which should be used for the hash, the default is SHA-256.
If you need to hold a reference to a file at a remote location you can use the RemoteFile
class provided here. This allows you to easily get the hash of a remote file with RemoteFile#hash
.
This is particularly useful when used in parallel with our other open-source library, resource-pack-deploy, so you can get the hash of the latest resource pack file, and use it to reset a client's cached version of your resource pack if there are any differences.
Twilight offers a collection of widely used symbols within the Symbols
object.
These include, but are not limited to:
- • (BULLET)
- ❤ (HEART)
- ★ (STAR)
- » (SMALL_RIGHT)
- █ (SQUARE)
- ⬤ (CIRCLE)
- ⛏ (PICKAXE)
- ⛀ (COIN)
- ⛁ (COINS)
Twilight is bundled with some useful libraries, some include:
- Bson
- Dotenv
- Mongo Sync Driver
- GSON
We're aiming to provide some standard GSON Type Adapters for ease of use. Currently, we have adapters for the following:
- Location
We have a useful exclusion strategy, allowing you to exclude marked fields of a class from being in serialized JSON. All you need to do is annotate the class field with @Exclude
!
Make sure to use our GSON
import rather than making your own/own builder, as the Gson instance has to declare the ExclusionStrategy.