Remove chunks from a Minecraft world based on their not being any 'preserve' blocks in the chunk.
java -cp OBMapTrim-1.0.jar net.obmc.OBMapTrim.OBMapTrim -w [ -r <radius> ] [ -p <preserve blocks> ]
Where:
-w <world path> The path to the world folder
-r <radius> Preserve area around current chunk being checked
-p <preserve blocks> A list of 'preserve' blocks to look for in a chunk
-l Load chunk data from a file
You only need to provide the world folder path. The radius and preserve block list have defaults that will be used
in the event you do not provide them. While running it will output what it's doing and will write a lot of detail
to a log file. The log file will have the format 'obmaptrim-YYYYMMDD-HHMMSS.log' where YYYYMMDD are the year, month
and day, and HHMMSS are the hour, minutes and seconds when the program was run. The -l load option is for loading
chunk data from a save file instead of scanning all region files for chunk data, instead of scanning for it. If you
have a big map it will save you a lot of time in the event you need to run the program multiple times.
Examples:
Simplest example would be to run with the default values. This will use the default radius and default preserve blocks:
java -cp OBMapTrim-1.0.jar net.obmc.OBMapTrim.OBMapTrim -w ~/myserver/world
Specify the radius you want to preserve around a chunk by specifying it with the -r option:
java -cp OBMapTrim-1.0.jar net.obmc.OBMapTrim.OBMapTrim -w ~/myserver/world -r 10
Tell the program the radius and what blocks to look for in the chunks:
java -cp OBMapTrim-1.0.jar net.obmc.OBMapTrim.OBMapTrim -w ~/myserver/world -r 10
-p "redstone_lamp,lit_redstone_lamp,sea_lantern"
Use the chunk load option to load chunk state from the file "chunkData.dat":
java -cp OBMapTrim-1.0.jar net.obmc.OBMapTrim.OBMapTrim -w ~/myserver/world -r 10 -l
The default radius is 3 and the default block list is based on blocks we use for our city build server:
sponge,powered_rail,white_wool,smooth_stone_slab,anvil,redstone_lamp,lit_redstone_lamp,
sea_lantern,beacon,redstone_block,emerald_block,barrier,anvil
Here is an image of our build server before and after running the trim with a radius of 10. This was
a 186 region files, 160,000 chunks, 1.2GB map. After the clear down of the unnecessary chunks it was
reduced to 116 region files, 64,293 chunks deleted and the size now 515MB.
Before:
After:
The program will make three passes of the world region files in order to remove chunks. The first is to
get a count of chunks in each file and make an internal hash of the chunks and a unique identifier for
every chunk. The second pass processes the blocks of each chunk and looks for any of the preserve blocks.
If there are any, then the chunk in the hash is marked for keeping and all of the chunks in the specified
radius around it will also be marked for keeping, if not already marked. At this point the chunkData.dat
is created from the hashmap, which can be loaded with the -l option and therefore skipping the first two
passes. Very handy if you want to perform multiple runs of the program to get the map just right. The third
and final pass will remove the chunk if not marked for keeping based on the hash status. If there are no
chunks to be retained in the region file, then the file is removed along with any corresponding poi end entity
region files, otherwise the region file is saved to update it with any changes.
IMPORTANT: Make sure you back up your world before running this program!!
If any errors are encountered processing region files, then the program will terminate. This could leave your
world in an inconsistent state and corrupt. You need to maintain a backup of your world!!
All of this processing is recorded in the log file.
Compiled for 1.21 and Java 21.