BGI/Ethornell .arc rebuild

A python script made to fix the annoying amount of files when making a translation patch for a BGI/Ethornell Visual Novel.

What is this?/What does it do?

This is a python script that takes all the BGI/Ethornell files in a directory to repack a .arc file.

(It also works if you create sub-folders with the name of the .arc file to create and put the files inside!)

How do I use it?

Simply run python rebuild.py in a terminal/cmd, a folder called "extracted" should be created along with some indications.

Now, you can take two routes (because you can do both at the same time):

1. Bulk .arc rebuilding

You can create sub-folders with the name of the .arc file you want to create (example: 'data01510') and drop all the extension-less files there.

When you run the script, a new .arc.new file with the name of the sub-folder will be created with the files inside it.

2. One-by-one .arc rebuilding

Just drop the extension-less files inside the 'extracted' folder. You only can make one .arc at a time.

Make sure you have the original .arc file, for naming purposes. (little life hack: create a new text file, call it whatever and make the extension .arc, the script just wants the name.)

Useful information

• Delete the .new extension in order to make BGI understand it. (And delete/backup the original .arc file, don't be a dummy)

• Most dialog scripts are almost always stored in data015x0.arc. (data01500.arc, data01510.arc, data01520.arc and so on)

• Sometimes, menu images and CGs you might want to edit are stored in data025x0.arc.

• I strongly recommend you use crskycode's fork of GARbro to unpack the .arc files as well as marcussacana's SacanaWrapper with the plugin for BGI files to work with the scenario scripts and make translation easier.

TODO

• Compression. Yeah, no. I tried remaking the huffman tree BGI uses and every time I tried it was like BGI said "Nuh UH" and crashed, so I won't keep trying.

In all seriousness, BGI has some internal checks for these kinds of stuff, and I'm no expert on reverse engineering to figure out where or how does it work for now.

Besides, I don't think the files would be that big, maybe compared to the originals it's gonna be like 500% bigger because of it having no compression, but going from 750kb to 5mb isn't that life changing, you know?

Found a problem?/Have a suggestion?

Create an Issue! I'll be happy to try and fix your problem or listen to your feedback.