Download
Choose Minecraft Version
1.18.11.21.1v1.4.0
1.18.11.21v1.4.0
1.18.11.20.5v1.3.1
1.18.11.20.4v1.3.1
1.18.11.20.3v1.3.1
1.18.11.20.2v1.3.1
1.18.11.20.1v1.2.0
1.18.11.20v1.2.0
1.18.11.19.4v1.2.0
1.18.11.19v1.2.0
1.18.11.18.2v1.2.0
1.18.11.18.1v1.2.0
1.18.11.18v1.2.0
Close

ChunkScan.Ores

An extension if ChunkScan that enables generating custom ores via postgen. This library supports:

  • Place groups of blocks (ores) in the world at random locations, in the specified ranges.
  • Filtering by biome. Supports 3d biomes- filter is based on individual ore vein location.
  • Filtering by dimension, including support for custom dimensions. Supports any world height between -2048 - 2048, however world bottom at 0 or -64 are most efficient.
  • Compatible with other worldgen, including custom worldgen (like Terralith) and other postgen methods.

This is an embedded library, so you package it inside your datapack as opposed to having a separate download. Requires LanternLoad and ChunkScan to operate.

Scoreboards

These are scoreboard objectives used to represent a value of some kind.

chunk_scan.ores.data
  Used for math and passing variables between functions.

Function Calls

Function calls are called by you to trigger certian events or features.

function chunk_scan.ores:v1/api/register_ore
  Call to add an ore to the registry.
  
  Input:
    #registry.min_y chunk_scan.ores.data [-2048,2048]: minimum y value to generate at
    #registry.max_y chunk_scan.ores.data [-2048,2048]: maximum y value to generate at
    #registry.min_veins chunk_scan.ores.data [0,32]: minimum number of veins to generate per chunk
    #registry.max_veins chunk_scan.ores.data [0,32]: maximum number of veins to generate per chunk
    #registry.min_vein_size chunk_scan.ores.data [0,16]: minimum number of ores per vein
    #registry.max_vein_size chunk_scan.ores.data [0,16]: maximum number of ores per vein
    #registry.ignore_restrictions chunk_scan.ores.data [0,1]: if 0, ore will only generate in stone like blocks. 1 for anywhere (even air).

    Optional: add this line to adjust dimension and/or biome whitelist/blacklist.
    Defaults to overworld and no biome restrictions (note- 'minecraft:' prefix is required).
    data modify storage chunk_scan.ores:registry input set value {dimension:"minecraft:overworld", biome:"<biome or biome tag>", biome_blacklist:0b}
	
  Output:
    #registry.result_id chunk_scan.ores.data: Returns -1 if registering ore failed.
    Otherwise, returns generated ore reg ID num.
    Save this number to a score. You will need to later to generate your ore 
    Example: scoreboard players operation <my_ore> <my_objective> = #registry.result_id chunk_scan.ores.data 

Function Tags

Functions tags are called by ChunkScan.Ores to inform you an event has happened, like an ore needs to be placed. To use these calls, you must add a function to the tag list.

function #chunk_scan.ores:v1/place_ore
  Executed at the location the ore will be placed.
  Input:
    #gen.id chunk_scan.ores.data -> id of ore to generate. If the id you received
    from registering your ore matches, place it.
    Example: execute if score <my_ore> <my_objective> = #gen.id chunk_scan.ores.data run setblock ~ ~ ~ minecraft:dirt

How to use

  1. Install LanternLoad in your datapack, following its install directions.
  2. Install ChunkScan in your datapack, following its install directions
  3. Copy the ChunkScan.Ores/data/chunk_scan.ores folder into your data pack
  4. Merge the contents of ChunkScan.Ores/data/chunk_scan/tags/functions/v2/generate.json into the file <your_datapack>/data/chunk_scan/tags/functions/v2/generate.json
  5. Merge the file contents of ChunkScan.Ores/data/load/tags/functions/* into the files at <your_datapack>/data/load/tags/functions/*
  6. Implement the API as described above.

For easier mangament of dependencies, check out my project Datapack Build Manager.

Smithed

Copyright © 2024 Smithed
Not an official Minecraft product. Not approved by or associated with Mojang Studios