Mdisk (WORK IN PROGRESS)

A tool for the game Beyond Good & Evil.

Note that I'm using Beyond Good & Evil - Steam & GOG - PC (sally_clean.bf), this tool is compatible with both versions.

Technical Summary

You can find all the researches I've done on game files here.

I'm trying to not use any library to decompress/compress the files as I'm also using this project to improve my skills, I'm only using lzo for the Bin files and iconv-lite for String <=> Uint8Array conversions as I'm using a lot of different string encodings.

Notes:

• Implemented a file reading system with chunks.
• Mostly using Uint8Arrays for hex data.
• Helper functions for bytes, files & numbers.
• Complete BMP export implementation.
• Classes for pointers & cache.

I'm using the implemented BMP generation to preview data from files such as mtx trailers or texture Bin files, in the goal of reverse engineering them.

Usage

The goal here was to build a tool that can extract and build the game files. The tool is separated into two main parts: the extraction and the building.

Extraction

The extraction is done by providing the path to the Big File and the path to the output directory. Note that the output directory will be created if it does not exist.

The extraction process will create multiple directories and one file:

• Bin: Contains the extracted .bin files.
• EngineDatas: Contains all the other type of files.
• ExtractedFiles: Contains the extracted files from the Bin & EngineDatas directories.
• metadata.json: Contains all the information about the files inside the Big File (for building).

Note that ExtractedFiles is basically where you will modify the files.

Inside of ExtractedFiles, you'll find a folder named for each extracted file, once modified, you can rebuild them, it will generate a build folder inside of it, this folder will contain the rebuilt file that will be automatically used to build the Big File.

Building

The building is done by providing the path to the extracted Big File (directory) and the path to the output directory where the new Big File will be created. Note that it verifies that all needed files are present in the extracted Big File.

Basically, all files that are not built inside the ExtractedFiles directory will be replaced by the original ones using the original Big File.

So if there's nothing inside the ExtractedFiles directory, the Big File will be the same as the original one, without the unused directories (the Big File contains a lot of unused dirs..).

WARNING: NEVER OVERWRITE THE ORIGINAL BIG FILE BY MINE, I PREFER TO SAY IT..

Extraction / Building Support

File	Description	Status
.bf	Big File	EXTRACTION / BUILDING
.bik	Video	EXTRACTION / BUILDING
.wa*	Audio	EXTRACTION
.mtx	Trailer videos ?	-----------------------
.bin	Binary files	WORK IN PROGRESS
.wol	List for ff0	-----------------------
.ofc	Binarized actions	-----------------------
.oin	Binarized actions	-----------------------
.ova	Binarized actions	-----------------------
.omd	Binarized actions	-----------------------

Bin extraction / building Support

File	Description	Status
fe*	Sound headers	-----------------------
ff4*	Sound effects	-----------------------
fd*	Texts	EXTRACTION
ff8*	Textures	-----------------------
ff0*	Miscellaneous	-----------------------

Credits

• Most of the information about general parsing comes from this documentation made by Kapouett.
• Another good source of information comes from this repository made by panzi.
• There's also this mysterious quickBMS script wrote by AnonBaiter
• Some info here that helped me to complete the list made by Kapouett.
• I'm also using ChatGPT to interpret some file structures.
• I used this file as a reference to extract audio files.