ViaVersion/common/src/main/java/com/viaversion/viaversion/rewriter/EntityRewriter.java

616 lines
26 KiB
Java
Raw Normal View History

Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
/*
* This file is part of ViaVersion - https://github.com/ViaVersion/ViaVersion
2024-01-01 12:39:45 +01:00
* Copyright (C) 2016-2024 ViaVersion and contributors
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
package com.viaversion.viaversion.rewriter;
import com.github.steveice10.opennbt.tag.builtin.CompoundTag;
import com.github.steveice10.opennbt.tag.builtin.IntTag;
import com.github.steveice10.opennbt.tag.builtin.ListTag;
2024-01-24 17:34:40 +01:00
import com.github.steveice10.opennbt.tag.builtin.NumberTag;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.github.steveice10.opennbt.tag.builtin.Tag;
import com.google.common.base.Preconditions;
import com.viaversion.viaversion.api.Via;
import com.viaversion.viaversion.api.connection.UserConnection;
import com.viaversion.viaversion.api.data.Int2IntMapMappings;
import com.viaversion.viaversion.api.data.Mappings;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.viaversion.viaversion.api.data.ParticleMappings;
import com.viaversion.viaversion.api.data.entity.DimensionData;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.viaversion.viaversion.api.data.entity.EntityTracker;
import com.viaversion.viaversion.api.data.entity.TrackedEntity;
import com.viaversion.viaversion.api.minecraft.Particle;
2024-01-24 17:34:40 +01:00
import com.viaversion.viaversion.api.minecraft.RegistryEntry;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.viaversion.viaversion.api.minecraft.entities.EntityType;
import com.viaversion.viaversion.api.minecraft.item.Item;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.viaversion.viaversion.api.minecraft.metadata.MetaType;
import com.viaversion.viaversion.api.minecraft.metadata.Metadata;
import com.viaversion.viaversion.api.protocol.Protocol;
import com.viaversion.viaversion.api.protocol.packet.ClientboundPacketType;
2024-02-07 18:29:19 +01:00
import com.viaversion.viaversion.api.protocol.packet.PacketWrapper;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.viaversion.viaversion.api.protocol.remapper.PacketHandler;
import com.viaversion.viaversion.api.protocol.remapper.PacketHandlers;
import com.viaversion.viaversion.api.rewriter.RewriterBase;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.viaversion.viaversion.api.type.Type;
2022-10-16 10:13:05 +02:00
import com.viaversion.viaversion.data.entity.DimensionDataImpl;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import com.viaversion.viaversion.rewriter.meta.MetaFilter;
import com.viaversion.viaversion.rewriter.meta.MetaHandlerEvent;
import com.viaversion.viaversion.rewriter.meta.MetaHandlerEventImpl;
2024-01-24 17:34:40 +01:00
import com.viaversion.viaversion.util.Key;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import java.util.ArrayList;
import java.util.Comparator;
2022-10-16 10:13:05 +02:00
import java.util.HashMap;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import java.util.List;
2022-10-16 10:13:05 +02:00
import java.util.Map;
2023-11-20 19:56:48 +01:00
import java.util.logging.Level;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
import java.util.logging.Logger;
import java.util.stream.Collectors;
2023-08-10 09:14:01 +02:00
import org.checkerframework.checker.nullness.qual.Nullable;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
public abstract class EntityRewriter<C extends ClientboundPacketType, T extends Protocol<C, ?, ?, ?>>
extends RewriterBase<T> implements com.viaversion.viaversion.api.rewriter.EntityRewriter<T> {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
private static final Metadata[] EMPTY_ARRAY = new Metadata[0];
protected final List<MetaFilter> metadataFilters = new ArrayList<>();
protected final boolean trackMappedType;
protected Mappings typeMappings;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
protected EntityRewriter(T protocol) {
this(protocol, true);
}
/**
* Creates a new entity rewriter instance.
*
* @param protocol protocol
* @param trackMappedType whether the mapped or unmapped entity type should be tracked
*/
protected EntityRewriter(T protocol, boolean trackMappedType) {
super(protocol);
this.trackMappedType = trackMappedType;
protocol.put(this);
}
/**
* Returns a meta filter builder.
* <p>
2024-01-29 10:17:00 +01:00
* Calling {@link com.viaversion.viaversion.rewriter.meta.MetaFilter.Builder#register()} will automatically register the filter on this rewriter.
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
*
* @return meta filter builder
*/
public MetaFilter.Builder filter() {
return new MetaFilter.Builder(this);
}
/**
* Registers a metadata filter.
2024-01-29 10:17:00 +01:00
* Note that {@link com.viaversion.viaversion.rewriter.meta.MetaFilter.Builder#register()} already calls this method.
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
*
* @param filter filter to register
* @throws IllegalArgumentException if the filter is already registered
*/
public void registerFilter(MetaFilter filter) {
Preconditions.checkArgument(!metadataFilters.contains(filter));
metadataFilters.add(filter);
}
@Override
public void handleMetadata(final int entityId, final List<Metadata> metadataList, final UserConnection connection) {
final TrackedEntity entity = tracker(connection).entity(entityId);
final EntityType type = entity != null ? entity.entityType() : null;
for (final Metadata metadata : metadataList.toArray(EMPTY_ARRAY)) { // Copy the list to allow mutation
MetaHandlerEvent event = null;
for (final MetaFilter filter : metadataFilters) {
if (!filter.isFiltered(type, metadata)) {
continue;
}
if (event == null) {
// Instantiate lazily and share event instance
event = new MetaHandlerEventImpl(connection, entity, entityId, metadata, metadataList);
}
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
try {
filter.handler().handle(event, metadata);
} catch (final Exception e) {
logException(e, type, metadataList, metadata);
metadataList.remove(metadata);
break;
}
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
if (event.cancelled()) {
// Remove meta, and break current filter loop
metadataList.remove(metadata);
break;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
}
if (event != null && event.hasExtraMeta()) {
2021-08-14 10:37:21 +02:00
// Finally, add newly created meta
metadataList.addAll(event.extraMeta());
}
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
if (entity != null) {
entity.sentMetadata(true);
}
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
@Override
public int newEntityId(int id) {
return typeMappings != null ? typeMappings.getNewIdOrDefault(id, id) : id;
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
/**
* Maps an entity type.
*
* @param type entity type
* @param mappedType mapped entity type
* @throws IllegalArgumentException if the types share the same implementing class
*/
public void mapEntityType(EntityType type, EntityType mappedType) {
Preconditions.checkArgument(type.getClass() != mappedType.getClass(), "EntityTypes should not be of the same class/enum");
mapEntityType(type.getId(), mappedType.getId());
}
protected void mapEntityType(int id, int mappedId) {
if (typeMappings == null) {
typeMappings = Int2IntMapMappings.of();
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
typeMappings.setNewId(id, mappedId);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
/**
* Maps entity ids based on the enum constant's names.
*
* @param oldTypes entity types of the higher version
* @param newTypeClass entity types enum class of the lower version
* @param <E> new enum entity type
*/
public <E extends Enum<E> & EntityType> void mapTypes(EntityType[] oldTypes, Class<E> newTypeClass) {
if (typeMappings == null) {
typeMappings = Int2IntMapMappings.of();
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
for (EntityType oldType : oldTypes) {
try {
E newType = Enum.valueOf(newTypeClass, oldType.name());
typeMappings.setNewId(oldType.getId(), newType.getId());
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
} catch (IllegalArgumentException notFound) {
if (!typeMappings.contains(oldType.getId())) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
Via.getPlatform().getLogger().warning("Could not find new entity type for " + oldType + "! " +
"Old type: " + oldType.getClass().getEnclosingClass().getSimpleName() + ", new type: " + newTypeClass.getEnclosingClass().getSimpleName());
}
}
}
}
/**
* Maps entity ids based on the protocol's mapping data.
*/
public void mapTypes() {
Preconditions.checkArgument(typeMappings == null, "Type mappings have already been set - manual type mappings should be set *after* this");
Preconditions.checkNotNull(protocol.getMappingData().getEntityMappings(), "Protocol does not have entity mappings");
2023-03-04 17:17:47 +01:00
typeMappings = protocol.getMappingData().getEntityMappings();
}
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
/**
* Registers a metadata handler to rewrite, item, block, and particle ids stored in metadata.
*
* @param itemType item meta type if needed
* @param blockStateType block state meta type if needed
* @param optionalBlockStateType optional block state meta type if needed
* @param particleType particle meta type if needed
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
*/
public void registerMetaTypeHandler(@Nullable MetaType itemType, @Nullable MetaType blockStateType, @Nullable MetaType optionalBlockStateType, @Nullable MetaType particleType) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
filter().handler((event, meta) -> {
final MetaType type = meta.metaType();
if (type == itemType) {
protocol.getItemRewriter().handleItemToClient(meta.value());
} else if (type == blockStateType) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
int data = meta.value();
meta.setValue(protocol.getMappingData().getNewBlockStateId(data));
} else if (type == optionalBlockStateType) {
int data = meta.value();
if (data != 0) {
meta.setValue(protocol.getMappingData().getNewBlockStateId(data));
}
} else if (type == particleType) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
rewriteParticle(meta.value());
}
});
}
public void registerTracker(C packetType) {
protocol.registerClientbound(packetType, new PacketHandlers() {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
@Override
public void register() {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
map(Type.VAR_INT); // 0 - Entity ID
map(Type.UUID); // 1 - Entity UUID
map(Type.VAR_INT); // 2 - Entity Type
handler(trackerHandler());
}
});
}
public void registerTrackerWithData(C packetType, EntityType fallingBlockType) {
protocol.registerClientbound(packetType, new PacketHandlers() {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
@Override
public void register() {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
map(Type.VAR_INT); // 0 - Entity id
map(Type.UUID); // 1 - Entity UUID
map(Type.VAR_INT); // 2 - Entity Type
map(Type.DOUBLE); // 3 - X
map(Type.DOUBLE); // 4 - Y
map(Type.DOUBLE); // 5 - Z
map(Type.BYTE); // 6 - Pitch
map(Type.BYTE); // 7 - Yaw
map(Type.INT); // 8 - Data
handler(trackerHandler());
handler(wrapper -> {
int entityId = wrapper.get(Type.VAR_INT, 0);
EntityType entityType = tracker(wrapper.user()).entityType(entityId);
if (entityType == fallingBlockType) {
wrapper.set(Type.INT, 0, protocol.getMappingData().getNewBlockStateId(wrapper.get(Type.INT, 0)));
}
});
}
});
}
public void registerTrackerWithData1_19(C packetType, EntityType fallingBlockType) {
protocol.registerClientbound(packetType, new PacketHandlers() {
2022-04-06 19:07:16 +02:00
@Override
public void register() {
2022-04-06 19:07:16 +02:00
map(Type.VAR_INT); // Entity id
map(Type.UUID); // Entity UUID
map(Type.VAR_INT); // Entity type
map(Type.DOUBLE); // X
map(Type.DOUBLE); // Y
map(Type.DOUBLE); // Z
map(Type.BYTE); // Pitch
map(Type.BYTE); // Yaw
map(Type.BYTE); // Head yaw
map(Type.VAR_INT); // Data
handler(trackerHandler());
handler(wrapper -> {
2023-08-10 09:14:01 +02:00
if (protocol.getMappingData() == null) {
return;
}
2022-04-06 19:07:16 +02:00
int entityId = wrapper.get(Type.VAR_INT, 0);
EntityType entityType = tracker(wrapper.user()).entityType(entityId);
if (entityType == fallingBlockType) {
wrapper.set(Type.VAR_INT, 2, protocol.getMappingData().getNewBlockStateId(wrapper.get(Type.VAR_INT, 2)));
}
});
}
});
}
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
/**
* Registers an entity tracker for the extra spawn packets like player, painting, or xp orb spawns.
*
* @param packetType packet type
* @param entityType entity type
* @param intType int type of the entity id
*/
public void registerTracker(C packetType, EntityType entityType, Type<Integer> intType) {
protocol.registerClientbound(packetType, wrapper -> {
int entityId = wrapper.passthrough(intType);
tracker(wrapper.user()).addEntity(entityId, entityType);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
});
}
/**
* Registers an entity tracker for the extra spawn packets.
*
* @param packetType packet type
* @param entityType entity type
*/
public void registerTracker(C packetType, EntityType entityType) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
registerTracker(packetType, entityType, Type.VAR_INT);
}
/**
* Sub 1.17 method for entity remove packets.
*
* @param packetType remove entities packet type
*/
public void registerRemoveEntities(C packetType) {
protocol.registerClientbound(packetType, wrapper -> {
int[] entityIds = wrapper.passthrough(Type.VAR_INT_ARRAY_PRIMITIVE);
EntityTracker entityTracker = tracker(wrapper.user());
for (int entity : entityIds) {
entityTracker.removeEntity(entity);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
});
}
/**
* 1.17+ method for entity remove packets.
*
* @param packetType remove entities packet type
*/
public void registerRemoveEntity(C packetType) {
protocol.registerClientbound(packetType, wrapper -> {
int entityId = wrapper.passthrough(Type.VAR_INT);
tracker(wrapper.user()).removeEntity(entityId);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
});
}
public void registerMetadataRewriter(C packetType, @Nullable Type<List<Metadata>> oldMetaType, Type<List<Metadata>> newMetaType) {
protocol.registerClientbound(packetType, new PacketHandlers() {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
@Override
public void register() {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
map(Type.VAR_INT); // 0 - Entity ID
if (oldMetaType != null) {
map(oldMetaType, newMetaType);
} else {
map(newMetaType);
}
handler(wrapper -> {
int entityId = wrapper.get(Type.VAR_INT, 0);
List<Metadata> metadata = wrapper.get(newMetaType, 0);
handleMetadata(entityId, metadata, wrapper.user());
});
}
});
}
public void registerMetadataRewriter(C packetType, Type<List<Metadata>> metaType) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
registerMetadataRewriter(packetType, null, metaType);
}
public PacketHandler trackerHandler() {
return trackerAndRewriterHandler(null);
}
2022-04-06 19:07:16 +02:00
public PacketHandler playerTrackerHandler() {
return wrapper -> {
final EntityTracker tracker = tracker(wrapper.user());
final int entityId = wrapper.get(Type.INT, 0);
2022-05-20 09:53:39 +02:00
tracker.setClientEntityId(entityId);
tracker.addEntity(entityId, tracker.playerType());
2022-04-06 19:07:16 +02:00
};
}
/**
* Returns a packet handler storing height, min_y, and name of the current world.
* If the client changes to a new world, the stored entity data will be cleared.
*
* @param nbtIndex index of the current world's nbt
* @return packet handler
*/
public PacketHandler worldDataTrackerHandler(int nbtIndex) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
return wrapper -> {
EntityTracker tracker = tracker(wrapper.user());
CompoundTag registryData = wrapper.get(Type.NAMED_COMPOUND_TAG, nbtIndex);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
Tag height = registryData.get("height");
if (height instanceof IntTag) {
int blockHeight = ((IntTag) height).asInt();
tracker.setCurrentWorldSectionHeight(blockHeight >> 4);
} else {
Via.getPlatform().getLogger().warning("Height missing in dimension data: " + registryData);
}
Tag minY = registryData.get("min_y");
if (minY instanceof IntTag) {
tracker.setCurrentMinY(((IntTag) minY).asInt());
} else {
Via.getPlatform().getLogger().warning("Min Y missing in dimension data: " + registryData);
}
String world = wrapper.get(Type.STRING, 0);
if (tracker.currentWorld() != null && !tracker.currentWorld().equals(world)) {
tracker.clearEntities();
tracker.trackClientEntity();
}
tracker.setCurrentWorld(world);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
};
}
public PacketHandler worldDataTrackerHandlerByKey() {
return wrapper -> {
EntityTracker tracker = tracker(wrapper.user());
String dimensionKey = wrapper.get(Type.STRING, 0);
DimensionData dimensionData = tracker.dimensionData(dimensionKey);
if (dimensionData == null) {
Via.getPlatform().getLogger().severe("Dimension data missing for dimension: " + dimensionKey + ", falling back to overworld");
dimensionData = tracker.dimensionData("minecraft:overworld");
Preconditions.checkNotNull(dimensionData, "Overworld data missing");
}
tracker.setCurrentWorldSectionHeight(dimensionData.height() >> 4);
tracker.setCurrentMinY(dimensionData.minY());
String world = wrapper.get(Type.STRING, 1);
if (tracker.currentWorld() != null && !tracker.currentWorld().equals(world)) {
tracker.clearEntities();
tracker.trackClientEntity();
}
tracker.setCurrentWorld(world);
};
}
2024-01-24 17:34:40 +01:00
public PacketHandler worldDataTrackerHandlerByKey1_20_5(final int dimensionIdIndex) {
return wrapper -> {
EntityTracker tracker = tracker(wrapper.user());
int dimensionId = wrapper.get(Type.VAR_INT, dimensionIdIndex);
DimensionData dimensionData = tracker.dimensionData(dimensionId);
if (dimensionData == null) {
Via.getPlatform().getLogger().severe("Dimension data missing for dimension: " + dimensionId + ", falling back to overworld");
dimensionData = tracker.dimensionData("minecraft:overworld");
Preconditions.checkNotNull(dimensionData, "Overworld data missing");
}
tracker.setCurrentWorldSectionHeight(dimensionData.height() >> 4);
tracker.setCurrentMinY(dimensionData.minY());
String world = wrapper.get(Type.STRING, 0);
if (tracker.currentWorld() != null && !tracker.currentWorld().equals(world)) {
tracker.clearEntities();
tracker.trackClientEntity();
}
tracker.setCurrentWorld(world);
};
}
public PacketHandler biomeSizeTracker() {
return wrapper -> trackBiomeSize(wrapper.user(), wrapper.get(Type.NAMED_COMPOUND_TAG, 0));
2023-08-03 10:36:30 +02:00
}
2023-10-19 02:53:14 +02:00
public PacketHandler configurationBiomeSizeTracker() {
return wrapper -> trackBiomeSize(wrapper.user(), wrapper.get(Type.COMPOUND_TAG, 0));
}
2023-08-03 10:36:30 +02:00
public void trackBiomeSize(final UserConnection connection, final CompoundTag registry) {
final CompoundTag biomeRegistry = registry.get("minecraft:worldgen/biome");
final ListTag biomes = biomeRegistry.get("value");
tracker(connection).setBiomesSent(biomes.size());
}
public PacketHandler dimensionDataHandler() {
return wrapper -> cacheDimensionData(wrapper.user(), wrapper.get(Type.NAMED_COMPOUND_TAG, 0));
}
2023-10-19 02:53:14 +02:00
public PacketHandler configurationDimensionDataHandler() {
return wrapper -> cacheDimensionData(wrapper.user(), wrapper.get(Type.COMPOUND_TAG, 0));
}
2022-10-16 10:13:05 +02:00
/**
2023-08-03 10:36:30 +02:00
* Caches dimension data, later used to get height values and other important info.
2022-10-16 10:13:05 +02:00
*/
2023-08-03 10:36:30 +02:00
public void cacheDimensionData(final UserConnection connection, final CompoundTag registry) {
final ListTag dimensions = ((CompoundTag) registry.get("minecraft:dimension_type")).get("value");
final Map<String, DimensionData> dimensionDataMap = new HashMap<>(dimensions.size());
for (final Tag dimension : dimensions) {
final CompoundTag dimensionCompound = (CompoundTag) dimension;
2024-01-24 17:34:40 +01:00
final NumberTag idTag = dimensionCompound.get("id");
2023-08-03 10:36:30 +02:00
final CompoundTag element = dimensionCompound.get("element");
final String name = (String) dimensionCompound.get("name").getValue();
2024-01-24 17:34:40 +01:00
dimensionDataMap.put(Key.stripMinecraftNamespace(name), new DimensionDataImpl(idTag.asInt(), element));
2023-08-03 10:36:30 +02:00
}
tracker(connection).setDimensions(dimensionDataMap);
2022-10-16 10:13:05 +02:00
}
2024-01-24 17:34:40 +01:00
public PacketHandler registryDataHandler1_20_5() {
return wrapper -> {
final String registryKey = Key.stripMinecraftNamespace(wrapper.get(Type.STRING, 0));
if (registryKey.equals("worldgen/biome")) {
final RegistryEntry[] entries = wrapper.get(Type.REGISTRY_ENTRY_ARRAY, 0);
tracker(wrapper.user()).setBiomesSent(entries.length);
} else if (registryKey.equals("dimension_type")) {
final RegistryEntry[] entries = wrapper.get(Type.REGISTRY_ENTRY_ARRAY, 0);
final Map<String, DimensionData> dimensionDataMap = new HashMap<>(entries.length);
for (int i = 0; i < entries.length; i++) {
final RegistryEntry entry = entries[i];
dimensionDataMap.put(entry.key(), new DimensionDataImpl(i, (CompoundTag) entry.tag()));
}
tracker(wrapper.user()).setDimensions(dimensionDataMap);
}
};
}
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
// ---------------------------------------------------------------------------
// Sub 1.14.1 methods
/**
* Returns a packethandler to track and rewrite an entity.
*
* @param metaType type of the metadata list
* @return handler for tracking and rewriting entities
*/
public PacketHandler trackerAndRewriterHandler(@Nullable Type<List<Metadata>> metaType) {
return wrapper -> {
int entityId = wrapper.get(Type.VAR_INT, 0);
int type = wrapper.get(Type.VAR_INT, 1);
int newType = newEntityId(type);
if (newType != type) {
wrapper.set(Type.VAR_INT, 1, newType);
}
EntityType entType = typeFromId(trackMappedType ? newType : type);
// Register Type ID
tracker(wrapper.user()).addEntity(entityId, entType);
if (metaType != null) {
handleMetadata(entityId, wrapper.get(metaType, 0), wrapper.user());
}
};
}
public PacketHandler trackerAndRewriterHandler(@Nullable Type<List<Metadata>> metaType, EntityType entityType) {
return wrapper -> {
int entityId = wrapper.get(Type.VAR_INT, 0);
// Register Type ID
tracker(wrapper.user()).addEntity(entityId, entityType);
if (metaType != null) {
handleMetadata(entityId, wrapper.get(metaType, 0), wrapper.user());
}
};
}
/**
* Returns a packethandler to track an object entity.
*
* @return handler for tracking and rewriting entities
*/
public PacketHandler objectTrackerHandler() {
return wrapper -> {
int entityId = wrapper.get(Type.VAR_INT, 0);
byte type = wrapper.get(Type.BYTE, 0);
EntityType entType = objectTypeFromId(type);
// Register Type ID
tracker(wrapper.user()).addEntity(entityId, entType);
};
}
// ---------------------------------------------------------------------------
2024-01-17 21:04:44 +01:00
public void rewriteParticle(Particle particle) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
ParticleMappings mappings = protocol.getMappingData().getParticleMappings();
int id = particle.getId();
if (mappings.isBlockParticle(id)) {
2023-10-21 05:00:39 +02:00
Particle.ParticleData<Integer> data = particle.getArgument(0);
data.setValue(protocol.getMappingData().getNewBlockStateId(data.getValue()));
} else if (mappings.isItemParticle(id) && protocol.getItemRewriter() != null) {
2023-10-21 05:00:39 +02:00
Particle.ParticleData<Item> data = particle.getArgument(0);
Item item = data.getValue();
protocol.getItemRewriter().handleItemToClient(item);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
particle.setId(protocol.getMappingData().getNewParticleId(id));
}
2024-02-07 18:29:19 +01:00
public void rewriteParticle(PacketWrapper wrapper, Type<Particle> from, Type<Particle> to) throws Exception {
final Particle particle = wrapper.read(from);
rewriteParticle(particle);
wrapper.write(to, particle);
}
private void logException(Exception e, @Nullable EntityType type, List<Metadata> metadataList, Metadata metadata) {
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
if (!Via.getConfig().isSuppressMetadataErrors() || Via.getManager().isDebug()) {
Logger logger = Via.getPlatform().getLogger();
2021-06-09 18:59:48 +02:00
logger.severe("An error occurred in metadata handler " + this.getClass().getSimpleName()
2021-06-10 08:42:52 +02:00
+ " for " + (type != null ? type.name() : "untracked") + " entity type: " + metadata);
2021-06-09 18:59:48 +02:00
logger.severe(metadataList.stream().sorted(Comparator.comparingInt(Metadata::id))
.map(Metadata::toString).collect(Collectors.joining("\n", "Full metadata: ", "")));
2023-11-20 19:56:48 +01:00
logger.log(Level.SEVERE, "Error: ", e);
Refactor entity tracking and meta handling This essentially merges the two approaches to the metadata handling from ViaVersion and ViaBackwards and improves on both designs. ViaVersion did not track every single entity, but only those needed (at least in theory) and can work with untracked entities' metadata. It had a very simple method overridden by metadata rewriter implementations, directly operating on the full metadata list and manually handling meta index changes as well as item/block/particle id changes. ViaBackwards on the other hand had to track *every single* entity and threw warnings otherwise - while less prone to errors due to giving obvious warnings in the console, it unnecessarily tracks a lot of entities, and those warnings also annoys users when encountering virtual entity plugins (operating asynchronously and sending update packets while already untracked or not yet tracked). Dedicated MetaHandlers made id changes and filtering a lot easier to read and write. However, the actual metadata list handling and its distribution to handlers was not very well implemented and required a lot of list copying and creation as well as exception throws to cancel individual metadata entries. This version has MetaFilters built with a Builder containing multiple helper functions, and the entity tracking is properly given its own map, hashed by a Protocol's class, to be easily and generically accessible from anywhere with only a Protocol class from the UserConnection, along with more optimized metadata list iteration. The entity tracking is largely unchanged, keeping ViaVersion's approach to not having to track *all* entities (and being able to handle null types in meta handlers). All of this is by no means absolutely perfect, but is much less prone to errors than both previous systems and takes a lot less effort to actually write. A last possible change would be to use a primitive int to object map that is built to be concurrency save for the EntityTracker, tho that would have to be chosen carefully.
2021-05-24 23:24:50 +02:00
}
}
}