A WIP Rust library for flexibly parsing and extracting assets from DELTARUNE's data.win.
This is not just a dumping tool, everything is loaded into memory and can be accessed directly from your Rust code.
You must provide your own data.win & audiogroup1.dat file. I will not host them myself, nor will I host any extracted assets here.
It might work for other GameMaker 2 games, but I have not tested anything other than the DELTARUNE Chapter 1&2 Demo.
Based on the documentation here: https://pcy.ulyssis.be/undertale/unpacking-corrected
(+ some adjustments for the newer version of GM)
Currently the game metadata, sounds, fonts, spritesheets, sprites, and backgrounds/tilesets are fully extractable.
More specifically, these chunks are supported right now:
- GEN8
- OPTN
- SOND
- SPRT
- TPAG
- TXTR
- AUDO
- FONT
- BGND
Not supported right now:
- EXTN (unused)
- AGRP
- PATH
- SCPT
- SHDR (unused)
- TMLN
- OBJT
- ROOM
- DAFL (unused)
- CODE
- VARI
- FUNC
- STRG
The script examples/dump.rs is a simple example binary that uses the library to dump assets from a provided data.win & audiogroup1.dat.
To run it, do cargo run --release --example dump and it will dump from ./data.win & ./audiogroup1.dat into ./extract/.
While this is neat and all, this is a library, not just a tool for dumping to files.
I want this library to be very controllable: you should be able to tell it exactly what to load and when to do it.
This goal is a WIP: currently you control when to parse each individual chunk, and when to load assets (ie. image/audio data) for individual chunks that have assets (currently TXTR, SPRT, SOND, FONT, BGND; eventually more?). For sprites, sounds, and backgrounds/tilesets, you can also choose to load the image/audio data for only certain sprites/sounds/tilesets (by name).
After a chunk is parsed, you can access the parsed data as a pretty simple set of structs.
When you load the assets for TXTR/SPRT/FONT/BGND, the texture(s) are loaded into memory as image::DynamicImage from the image crate, and can be used by your program.
When you load the assets for SOND/AUDO, the audio data is loaded into memory as a Vec<u8> if the audio is embedded in the data.win, otherwise you can use the sound's file field to locate the external file. The Vec<u8> is the raw file data for the embedded file, so you can literally just dump the bytes directly to an .ogg file, or you can use a library to parse the audio in-memory.
See examples/simple.rs for an example of the logic flow.
Licensed under either of
- Apache License, Version 2.0 (LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0)
- MIT license (LICENSE-MIT or http://opensource.org/licenses/MIT)
at your option.
Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.
