Wwise name cracker

Crack Wwise FNV-1 hashes (32-bit and 64-bit)

Build

To compile cracker (you need LDC (LLVM D compiler), you can use others but they might not be as performant)

$ ./build.sh

You can also compile with Profile-Guided Optimizations (PGO) by specifying input arguments that will be used for optimization:

$ ./build.sh sample-hashes.txt sample-names.txt

(You might want to pass --expand 0 to save time)

Usage

1. Collect hashes from .pck files

   $ find GamePath -iname '*.pck' -print0 | xargs -0 -I!!! -n 1 quickbms -l wwiser-utils/scripts/wwise_pck_extractor.bms !!! | tr -s '[:blank:]' | cut -d ' ' -f 4 | cut -d '/' -f 2 | cut -d '.' -f 1 | sort -u > hashes.txt

2. Filter name string list - this will drastically reduce search space. You can skip this step if you're using names from wwiser-utils.

   $ ./crackNames --filter hashes.txt strings.txt > hash_names.txt

3. Convert from hash_names.txt (it contains hash:name pairs) to only names.txt. This step is required only if you did previous (2nd) step.

Need to select only most suitable names, but most likely case is there aren't any duplicates.

$ cut -d ':' -f 2 hash_names.txt | sort > names.txt
$ uniq -d names.txt
(should be empty, no output)

4. Crack hashes (you need atleast some names since those will be used as patterns)

   $ ./crackNames hashes.txt names.txt > hash_cnames.txt

(You can use wwiser-utils/wwnames/GameName (External).txt)

4. Remove duplicates and update names

You can inspect them with

$ cat hash_cnames.txt | cut -d ':' -f 1 | sort | uniq -d | xargs -I!!! grep !!! hash_cnames.txt

Then if there aren't duplicates

cat hash_cnames.txt | cut -d ':' -f 2 | sort -u > "wwiser-utils/wwnames/GameName (External).txt"`

Options

There are several options you can adjust:

$ ./crackNames --help
Usage: ./crackNames [options] [hashes.txt] names.txt [names2.txt ...]
-f    --filter Filter provided names, listing ones that matched hashes
-p  --patterns Show name patterns
-g --generated List all generated names
-j      --json Show internal raw info in JSON
-e    --expand Increase name search by this amount (default 5)
-n       --new List only new names without those that already matched
-r     --roman Recognize Roman Numerals
-b     --32bit Look for names for 32-bit FNV-1 hashes
-t   --threads Number of threads to use (default number of CPU cores)
-h      --help This help information.

Use --expand to reduce or increase search space.

And if you think there might be Roman Numerals used then pass --roman - it will try names with III, IV, V, VI and so on.

You can also inspect recognized patterns which can be useful for diff'ing:

$ ./crackNames --patterns names.txt
%s/voice/chapter%d_%d_soldier%c_%d.wem:{English;Japanese}[0-3][0-15]{A;AB;AC;B;V;W}[201-209]
%s/voice/archive_%s_%d.wem:{English;Japanese}{guinaifen;player}[1-22]

Crack event names (32-bit FNV-1 hashes)

// TODO - Need to document it and decide on best workflow