Version 0.3.6

rla2json: add `--dump-audio` option
rla2json: add `--no_parallel` option

.vscode/launch.json

@@ -163,8 +163,9 @@
             "module": "sssekai",
             "args": [
                 "rla2json",
+                "--dump-audio",
                 "D:\\proseka_reverse\\assets\\streaming_live\\archive\\1st_live_vbs-1",
-                "D:\\proseka_reverse\\assets\\streaming_live\\archive\\1st_live_vbs-1_src"
+                "D:\\proseka_reverse\\assets\\streaming_live\\archive\\1st_live_vbs-1_hca"
             ],
             "justMyCode": true
         },

sssekai/__init__.py

@@ -1,5 +1,5 @@
 __VERSION_MAJOR__ = 0
 __VERSION_MINOR__ = 3
-__VERSION_PATCH__ = 5
+__VERSION_PATCH__ = 6
 
 __version__ = '%s.%s.%s' % (__VERSION_MAJOR__,__VERSION_MINOR__,__VERSION_PATCH__)

sssekai/__main__.py

@@ -82,8 +82,10 @@ def write(__s):
     spineextract_parser.set_defaults(func=main_spineextract)
     # rla2json
     rla2json_parser = subparsers.add_parser('rla2json', help='''Read streaming_live/archive files and dump their information to JSON''')
-    rla2json_parser.add_argument('infile', type=str, help='input file')
-    rla2json_parser.add_argument('outdir', type=str, help='output directory. multiple json files may be produced')
+    rla2json_parser.add_argument('infile', type=str, help='input file. either a streaming_live bundle or a zip file with the same file hierarchy (i.e. containing sekai.rlh, sekai_xx_xxxxxx.rla files)')
+    rla2json_parser.add_argument('outdir', type=str, help='output directory. RLA files in JSON format will be saved, unless otherwise specified.')
+    rla2json_parser.add_argument('--dump-audio', action='store_true', help='dump raw HCA audio data instead of the JSON data. Use https://github.com/mos9527/sssekai_streaming_hca_decoder to decode them.')
+    rla2json_parser.add_argument('--no-parallel', action='store_true', help='disable parallel processing')
     rla2json_parser.set_defaults(func=main_rla2json)
     # apphash
     apphash_parser = subparsers.add_parser('apphash', help='''Download/extract game AppHash values''')

sssekai/entrypoint/rla2json.py

@@ -8,37 +8,67 @@
 from UnityPy.enums import ClassIDType
 from sssekai.fmt.rla import read_rla
 from tqdm import tqdm
+import zipfile, base64
 logger = getLogger(__name__)
 
-
-def worker_job(sname, version, script):
+def dump_json_job(sname, version, script):
     rla = read_rla(BytesIO(script), version)
     dump(rla, open(sname + '.json', 'w', encoding='utf-8'), indent=4, ensure_ascii=False)
 
+def dump_audio_job(sname, version, script):
+    rla = read_rla(BytesIO(script), version)
+    for tick, data in rla.items():
+        data = data.get('SoundData', None)
+        if data:
+            for i, b64data in enumerate(data):
+                fname = sname + '_%d_%d.hca' % (i, tick)
+                raw_data = base64.b64decode(b64data['data'])
+                with open(fname, 'wb') as f: f.write(raw_data)
+
 def main_rla2json(args):
-    with open(args.infile,'rb') as f:
-        env = load_assetbundle(f)
+    with open(args.infile,'rb') as f:        
         datas = dict()
-        for obj in env.objects:
-            if obj.type in {ClassIDType.TextAsset}:
-                data = obj.read()
-                datas[data.name] = data 
+        if f.read(2) == b'PK':
+            f.seek(0)            
+            with zipfile.ZipFile(f, 'r') as z:
+                for name in z.namelist():
+                    with z.open(name) as zf:
+                        datas[name] = zf.read()
+        else:
+            f.seek(0)
+            rla_env = load_assetbundle(f)            
+            for obj in rla_env.objects:
+                if obj.type in {ClassIDType.TextAsset}:
+                    data = obj.read()
+                    datas[data.name] = data.script.tobytes()
         header = datas.get('sekai.rlh', None)
         assert header, "RLH Header file not found!"
         makedirs(args.outdir, exist_ok=True)
-        header = loads(header.text)
+        header = loads(header.decode('utf-8'))
+        splitSeconds = header['splitSeconds']
+
         dump(header, open(path.join(args.outdir, 'sekai.rlh.json'), 'w', encoding='utf-8'), indent=4, ensure_ascii=False)
         version = tuple(map(int, header['version'].split('.')))
-        splitSeconds = header['splitSeconds']
         logger.info('Version: %d.%d' % version)
         logger.info('Count: %d' % len(header['splitFileIds']))        
+
+        worker_job = dump_json_job
+        if args.dump_audio:
+            worker_job = dump_audio_job
+            logger.info('Dumping RLA raw HCA audio data')
+        else:
+            logger.info('Dumping RLA data to JSON')
         with ProcessPoolExecutor() as executor:
-            logger.info('Dumping RLA data with %d processors' % executor._max_workers)
+            if not args.no_parallel:
+                logger.info('Dumping RLA data with %d processors' % executor._max_workers)
             futures = []
             for sid in header['splitFileIds']:
                 sname = 'sekai_%2d_%08d' % (splitSeconds, sid)
-                script = datas[sname + '.rla'].script.tobytes()                
-                futures.append(executor.submit(worker_job, path.join(args.outdir,sname), version, script))
+                script = datas[sname + '.rla']
+                if args.no_parallel:
+                    worker_job(path.join(args.outdir,sname), version, script)
+                else:
+                    futures.append(executor.submit(worker_job, path.join(args.outdir,sname), version, script))
             finsihed_futures = set()
             with tqdm(total=len(futures)) as pbar:
                 while len(finsihed_futures) < len(futures):