Tool items update

[mineLdiver]

Current StationAPI tool items system is convoluted and broken.

There aren't solid plans on how to reimplement the system.

Discussion is welcome.

[mineLdiver]

My proposed layout:

The idea is that each level node is only aware of its predecessors and can have multiple of them.

In code, if tool material only specifies a numeric mining level, it'll stick to the default branch (vanilla + extended).
If a level node is specified, it overrides the numeric mining level, binding the tool material to the appropriate place in the graph.

Any level node can be modified at any time to get more predecessors, including vanilla nodes.

Feedback is greatly appreciated.

[mineLdiver]

Draft implementation without extended levels

import java.util.Arrays;
import java.util.HashSet;
import java.util.Optional;
import java.util.Set;

public class Main {
    static MiningNode[] VANILLA_NODES = new MiningNode[4];
    static {
        var base = new MiningNode(Optional.empty(), Set.of());
        var stone = new MiningNode(Optional.of(new BlockTag("needs_stone_tool")), set(base));
        var iron = new MiningNode(Optional.of(new BlockTag("needs_iron_tool")), set(stone));
        var diamond = new MiningNode(Optional.of(new BlockTag("needs_diamond_tool")), set(iron));
        VANILLA_NODES[0] = base;
        VANILLA_NODES[1] = stone;
        VANILLA_NODES[2] = iron;
        VANILLA_NODES[3] = diamond;
    }

    @SafeVarargs
    static <T> Set<T> set(T... elements) {
        return new HashSet<>(Arrays.asList(elements));
    }

    public static void main(String[] args) {
        var myLevel1 = new MiningNode(Optional.of(new BlockTag("needs_my_tool_level_1")), set(VANILLA_NODES[1]));
        var myLevel2 = new MiningNode(Optional.of(new BlockTag("needs_my_tool_level_2")), set(myLevel1));
        var myLevel3 = new MiningNode(Optional.of(new BlockTag("needs_my_tool_level_3")), set(myLevel2));
        VANILLA_NODES[2].predecessors().add(myLevel3);
        System.out.println(VANILLA_NODES[1].contains(new BlockTag("needs_my_tool_level_2")));
        System.out.println(VANILLA_NODES[2].contains(new BlockTag("needs_my_tool_level_2")));
    }

    record ToolMaterial(
            int miningLevel,
            int itemDurability,
            float miningSpeed,
            int attackDamage,
            Optional<MiningNode> miningNode
    ) {
        MiningNode getNode() {
            return miningNode.orElseGet(() -> VANILLA_NODES[miningLevel]);
        }
    }

    record MiningNode(
            Optional<BlockTag> tag,
            Set<MiningNode> predecessors
    ) {
        boolean contains(BlockTag otherTag) {
            return (tag.isPresent() && tag.get().equals(otherTag)) || predecessors.stream().anyMatch(predecessor -> predecessor.contains(otherTag));
        }
    }

    record BlockTag(String id) {}
}

[mineLdiver]

Actual implementation turned out to be way easier thanks to tags that basically replace the MiningNode and all related logic.

Implementation is being worked on at https://github.com/ModificationStation/StationAPI/tree/2.0-alpha.3