From 8dff75939f1ba1d3cd6c03579710698becb9e3e4 Mon Sep 17 00:00:00 2001
From: Joel Wetzell <jwetzell@yahoo.com>
Date: Mon, 1 Mar 2021 19:49:44 -0600
Subject: [PATCH] Continue work on SBK parser

---
 sbk_header.txt |  47 ----------------
 sbkparser.py   | 145 +++++++++++++++++++++++++++++++++++++++++--------
 2 files changed, 123 insertions(+), 69 deletions(-)
 delete mode 100644 sbk_header.txt

diff --git a/sbk_header.txt b/sbk_header.txt
deleted file mode 100644
index 1635c85..0000000
--- a/sbk_header.txt
+++ /dev/null
@@ -1,47 +0,0 @@
-This is the start of an SBK file specifically BEACH.SBK from JAK1
-
-
-------------??????????????????------------------
-626561636800000000000000000000000000000028000000
-
-------------NAME---------------- --????--
-776F726D2D69646C6500000000000000 32000F10
-776F726D2D7269736531000000000000 32000F10
-776F726D2D73696E6B00000000000000 32000F10
-7368656C6C2D75700000000000000000 05800710
-7368656C6C2D646F776E000000000000 05800710
-736E6170000000000000000000000000 05800710
-637261622D736C696465000000000000 05800710
-6C75726B6572637261622D6469657300 05800710
-70656C6963616E2D666C617000000000 64801B10
-70656C6963616E2D67756C7000000000 64801B10
-70757070792D6261726B000000000000 19C00810
-6C75726B6572646F672D69646C650000 19C00810
-6C75726B6572646F672D646965730000 19C00810
-73656167756C6C2D74616B656F666600 3C001910
-67656172732D72756D626C6500000000 0F000F10
-726F70652D7374726574636800000000 05800710
-6D6F6E6B657900000000000000000000 05800710
-62697264000000000000000000000000 05800710
-66757365000000000000000000000000 0A800C10
-7361636B2D6C616E6400000000000000 1E001910
-6C75726B6572646F672D626974650000 05800710
-7361636B2D696E636F6D696E67000000 1E001910
-63616E6E6F6E2D73686F740000000000 96006410
-646972742D6372756D626C6500000000 19800710
-67726F74746F2D706F6C652D68697400 19800710
-6567672D637261636B00000000000000 96003210
-76656E742D726F636B2D627265616B00 50001910
-6567672D686974000000000000000000 05800710
-63616E6E6F6E2D636861726765000000 05800710
-62656163682D616D6232000000000000 32001410
-746F7765722D77696E64730000000000 05800710
-746F7765722D77696E64320000000000 32001410
-746F7765722D77696E64330000000000 05800710
-77617465722D6C617000000000000000 32001410
-776F726D2D6269746500000000000000 1E000A10
-776F726D2D7461756E74000000000000 05800710
-74656C6573636F706500000000000000 05800710
-66616C6C696E672D6567670000000000 05800710
-64726970000000000000000000000000 05800710
-776F726D2D6469657300000000000000 05800710
\ No newline at end of file
diff --git a/sbkparser.py b/sbkparser.py
index 963fb7d..56beb1a 100644
--- a/sbkparser.py
+++ b/sbkparser.py
@@ -1,6 +1,7 @@
+#UNFINISHED
 import os
 import argparse
-
+import struct
 def is_valid_file(parser, arg):
     if not os.path.exists(arg):
         parser.error("The file %s does not exist!" % arg)
@@ -23,26 +24,126 @@ sounds_start = 24
 sounds_length = 20
 
 with open(args.filepath, "rb") as f:
-    soundbank_name = f.read(20).decode('utf-8').strip()
-    f.seek(24)
-    soundsFound = 0
-    while 1:
-        empty = b'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00'
-        sounds_chunk = f.read(sounds_length)
-        if sounds_chunk == empty:
-            break
-        else:
-            sound_name = sounds_chunk[:-4].decode('utf-8').strip()
-            sound_data = sounds_chunk[-4:]
-            soundsFound += 1
-            print(f'Found {sound_name} with some data {sound_data[0:2].hex()} {sound_data[2:].hex()}')
-        
-
-
-
-
-        
-
-        
+    sbkInfo = struct.unpack("20sI",f.read(24))
+    print(sbkInfo)
+    soundDatas = []
+    for i in range(sbkInfo[1]):
+        soundData = struct.unpack("16sHH",f.read(20))
+        print(soundData)
+    f.seek(2048)
+
+    '''
+        seq
+        I version
+        I subversion
+        I minorversion
+        I something?
+        I soundbankStart
+        I soundBankSize
+    '''
+
+    seqStart = f.tell()
+    seq = struct.unpack("6I",f.read(24))
+    print(seq)
+
+    soundbankStart = seq[4] + seqStart
+    print(f"SoundBank Offset: {soundbankStart}")
+
+    '''
+        sb1k
+
+        4s signature 'SB1k'
+        I
+        I
+        4s
+        I
+        H
+        H
+        H
+        H
+        I instrumentOffset from beingging of this struct
+        I regionOffset from beginning of this struct
+        I
+        I
+        I
+        I
+        I
+        12s
+        H
+        H
+        H
+        H
+        I
+    '''
+    
+
+    soundBankHeaderStart = f.tell()
+    sb1k = struct.unpack("4sII4sIHHHHIIIIIII12sHHHHI",f.read(80))
+
+    instrumentStart = sb1k[8] + soundBankHeaderStart
+    print(instrumentStart)
+    f.seek(instrumentStart)
+    regionStart = sb1k[10] + soundBankHeaderStart
+    print(sb1k)
+
+    instruments = []
+    while(f.tell() < regionStart):
+        '''
+            struct Instrument {
+                uint8_t nRegion; // usually 1
+                uint8_t volume; // ? maybe... like 7f is 1.0?
+                uint16_t something; // always 0
+                uint32_t oRegion; // relative to start of SBv2 block I think, points to the first region for instrument
+            } __attribute__((packed));
+        '''
+        instrument = struct.unpack("BBBBHI",f.read(12))
+        instruments.append(instrument)
+
+    print("INSTRUMENTS")
+    offsets = []
+    for instrument in instruments:
+        if instrument[-1] not in offsets:
+            offsets.append(instrument[-1])
+        print(instrument)
+    print(len(offsets))
+
+    f.seek(regionStart)
+
+    regions = []
+    while(f.tell()<soundbankStart):
+        #format defintely not right but 
+        region = struct.unpack("10I", f.read(40))
+        regions.append(region)
+    print("REGIONS")
+    sampleOffsets = []
+    for region in regions:
+        print(region)
+        sampleOffset = soundbankStart + region[-4]
+        if sampleOffset not in sampleOffsets:
+            sampleOffsets.append(sampleOffset)
+        #print(f"Region Sample Audio  offset: {soundbankStart + region[10]}")
+    # this should align the index of the sample offset with it's sample ID
+    sampleOffsets.sort()
+
+    count = 0
+    for offset in sampleOffsets:
+        f.seek(offset)
+        soundData = f.read(1)
+        while(f.tell() not in sampleOffsets):
+            character = f.read(1)
+            if not character:
+                break
+            soundData += character
+
+        #VAGp header data length and sample rate are packed in bytes 11-15 and 16-20 respectively
+        header = b'\x56\x41\x47\x70\x00\x00\x00\x20\x00\x00\x00\x00'
+        header += struct.pack(">II", len(soundData), 22050)
+        header += b'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x61\x73\x73\x69\x73\x74\x61\x6E\x74\x2D\x76\x69\x6C\x6C\x61\x67\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00'
+        
+        with open(f'./OUT/VAGp/{count}.VAGp','wb') as out:
+            out.write(header)
+            out.write(soundData)
+            out.close()
+        count += 1