update to 26.1.2

[feeeedox]

Fixes #401

[feeeedox]

maven("https://jitpack.io")

implementation("com.github.feeeedox.AnvilGUI:anvilgui:8d6ec7efad")

[Ghosty920]

the jitpack file is useless for a PR btw, since @WesJD doesn't use it anyway

[feeeedox]

the jitpack file is useless for a PR btw, since @WesJD doesn't use it anyway

Yeah, he can remove it. I just added it, so im able to use it in my projects without referring to a local jar file.

[WesJD]

Thanks for submitting this! Just a few things to fix and this is good to merge.

[0dinD]

@WesJD @feeeedox I could be wrong because I don't have time to test/investigate this for a few days, but my initial reaction is that this seems like the wrong way to fix #401.

This is the diff between the 26_R1 wrapper and the 26_R2 wrapper:

diff --git a/Wrapper26_R1.java b/Wrapper26_R2.java
index 020216b..a1ad266 100644
--- a/Wrapper26_R1.java
+++ b/Wrapper26_R2.java
@@ -9,11 +9,7 @@ import net.minecraft.network.protocol.game.ClientboundOpenScreenPacket;
 import net.minecraft.network.protocol.game.ClientboundSetExperiencePacket;
 import net.minecraft.server.level.ServerPlayer;
 import net.minecraft.world.Container;
-import net.minecraft.world.inventory.AbstractContainerMenu;
-import net.minecraft.world.inventory.AnvilMenu;
-import net.minecraft.world.inventory.ContainerLevelAccess;
-import net.minecraft.world.inventory.MenuType;
-import net.minecraft.world.inventory.Slot;
+import net.minecraft.world.inventory.*;
 import org.bukkit.craftbukkit.CraftWorld;
 import org.bukkit.craftbukkit.entity.CraftPlayer;
 import org.bukkit.craftbukkit.event.CraftEventFactory;
@@ -21,7 +17,7 @@ import org.bukkit.craftbukkit.util.CraftChatMessage;
 import org.bukkit.entity.Player;
 import org.bukkit.inventory.Inventory;
 
-public final class Wrapper26_R1 implements VersionWrapper {
+public final class Wrapper26_R2 implements VersionWrapper {
     private int getRealNextContainerId(Player player) {
         return toNMS(player).nextContainerCounter();
     }
@@ -51,14 +47,18 @@ public final class Wrapper26_R1 implements VersionWrapper {
         } catch (NoSuchMethodError ignored) {
             // Workaround for Paper servers
             try {
-                Class<?> inventoryCloseEventReasonClass =
-                        Class.forName("org.bukkit.event.inventory.InventoryCloseEvent$Reason");
+                ServerPlayer nmsPlayer = toNMS(player);
+
+                Object reasonPlugin = Enum.valueOf(
+                        (Class<? extends Enum>) Class.forName("org.bukkit.event.inventory.InventoryCloseEvent$Reason"),
+                        "PLUGIN");
+
                 Method handleInventoryCloseEventMethod = CraftEventFactory.class.getMethod(
-                        "handleInventoryCloseEvent", ServerPlayer.class, inventoryCloseEventReasonClass);
-                handleInventoryCloseEventMethod.invoke(
-                        null,
-                        toNMS(player),
-                        inventoryCloseEventReasonClass.getField("UNKNOWN").get(null));
+                        "handleInventoryCloseEvent",
+                        net.minecraft.world.entity.player.Player.class,
+                        Class.forName("org.bukkit.event.inventory.InventoryCloseEvent$Reason"));
+
+                handleInventoryCloseEventMethod.invoke(null, nmsPlayer, reasonPlugin);
             } catch (ReflectiveOperationException e) {
                 throw new RuntimeException(e);
             }

The only meaningful changes I see are:

• The close reason was changed from UNKNOWN to PLUGIN. Why? That seems wrong to me at first glance.
• The signature of handleInventoryCloseEvent was changed from net.minecraft.server.level.ServerPlayer.class to net.minecraft.world.entity.player.Player.class. This looks right, but I don't think it should ever have been net.minecraft.server.level.ServerPlayer.class in the 26_R1 wrapper to begin with.

@WesJD did you ever test #398 on a Paper 26.1.1 server? The PR message seems to indicate that you only tested on Spigot because Paper was not released at that time? Again, I haven't had time to test, but my suspicion is that #401 affects 26_R1 (26.1.1) as well, not just 26.1.2. In that case, we should fix the 26_R1 wrapper, not add a new 26_R2 wrapper.

For context, in the 1_21_R7 wrapper, net.minecraft.world.entity.player.EntityHuman.class is used, which would translate to net.minecraft.world.entity.player.Player.class, not net.minecraft.server.level.ServerPlayer.class: https://mappings.dev/1.21.11/net/minecraft/world/entity/player/Player.html