widberg / bff Goto Github PK

• Highlight previewable resources
• Sort by name, extension, and file size
• Filter by extension

Pass a context around during binrw and serde operations.

choose binary, JSON, or rich and it will try to do the one you selected or fall back to a more basic one if that fails.

Binary

The decompressed/depooled object from the BigFile without any post-processing.

JSON

A single JSON object representing the object

Rich

Common file formats representing the object. May be multiple files.

Directory Structure

USA1.DPC.d
|- manifest.json
|- objects
|- whatever_name_you_want.UserDefine_Z.d
| |- object.json
| |- data.txt
|- whatever_name_i_want.Mesh_Z.d
|- object.json
You can name the folders whatever you want since the object.json file has the object name.

Add a way to insert a message into padding somewhere near the top of the file when writing. By default, it could say something like "made with <3 by bff contributors (https://github.com/widberg/bff)". Optional of course.

Ubi uses BFXX. BS uses DBX.

Version Number Override: Version number to treat the BigFile as having.
Platform Override

Add support for Sound BigFiles (sdx) and aif (the Zouna one), joe, and vai. https://github.com/widberg/fmtk/wiki/Asobo-Audio-Formats.

• Sound BigFiles
• AIF
• JOE
• VAI

We need PS ADPCM and NGC DSP codec support

Shaun White: World Tour for the Wii has a SoundStreams_WII.bin file. Shaun White for the PS2 has a bunch of sound formats LM1+LS1, SM1+SS1, and SP1. Plague Tale + others use Wwise PCK files.

Store extra info for byte-for-byte recreation. maybe call it too much information --tmi could control a global setting cuz serde doesn't do args well. Also perfect is a strong word.

List of names used by an object

cli subcommands that accept files should use - as the file name for stdin/stdout when applicable.

De/serialize names as strings when possible. Allow loading name files and use the loaded names in output. Cli command for merging and generating name files. Allow dumping a name file when generating a BigFile using the deserialized names.

Support Ubisoft compiled scripts and compiled script packages

Auto-generate version info from git for the long version like dpc tool did.

The title says it all. Basic writing capabilities.

Ability to write nicknames for resources, save and load them as JSON

Make block offset optional and calculate it as needed. If it is supplied and insufficient then warn.

For some ungodly reason, different games use different algorithms for names. And newer games use 64 bit names.

Maybe make Name an enum of all the variants and have a global select which type of name is in use based on the version number. Using generics everywhere is going to get bad fast.

🦆

Since bff is becoming a Zouna swiss army knife we might as well handle .psc files

Avoid recompressing resource bodies when writing to the pool

Parallelize reading/writing the actual BigFile file by only scanning for resource offsets the first time through and not actually reading/decompressing anything

Store binary versions of resources in a hidden folder and use those to avoid redoing work when importing if none of the exported files were modified.

Use arrayfire or similar for reverse crc

Use rayon for sorting and filtering in the GUI, and potentially other places

Use base64 for large byte arrays when serializing (https://docs.rs/serde_with/latest/serde_with/base64/index.html ?).

At least they'll display as strings even if we cant parse them.

Try every available object parser and use the first success. Only an error if all fail. Some way to evaluate a BigFile to see if an existing object form works for all objects of that type in the BigFile. Possibly another subcommand that tries to determine if any of the existing parsers work for every resource of a class in a collection of bigfiles of the same version.

Improve performance by not copying the model all the time and load materials

Earlier games don't use the _Z suffix on class names