paco_extract.py

paco_extract.py

#!/usr/bin/env python3

import struct
import sys

from dataclasses import dataclass, field

@dataclass
class PACoHeader:
    # _ stands for unkown usage
    head: int               # 0x0001
    head2: int              # 0x0026
    width: int
    height:int 
    fps: int                # negative is indicative of no audio, but not definitive
    flags: int              # flags is 0x100: audio
    max_chunk_size: int
    _zero: int              # always zero?
    _audio_related: int     # unkown, audio related
    frames: int             # number of chunks, i.e. frames
    frames2: int            # number of chunks repeated
    _eight: int             # always 8?
    _six: int               # always 0x600
    _flags2: int            # some flags
    _zero2: int             # always 0?

    has_audio: bool = field(init=False)
    framerate: int = field(init=False)
    
    def __post_init__(self):
        self.has_audio = (self.flags & 0x100) == 0x100
        self.framerate = abs(self.fps)

    def check(self):
        assert self.head == 0x1, f"head is not 0x1: {self.head}"
        assert self.head2 == 0x26, f"head2 is not 0x26: {self.head2}"
        assert self._zero == 0, f"_zero is not 0: {self._zero}"
        assert self._eight == 8, f"_eight is not 8: {self._eight}"
        assert self._six == 0x600, f"_six is not 0x600: {self._six}"
        assert self._zero2 == 0, f"_zero2 is not 0: {self._zero2}"

def main():
    filename = sys.argv[1]
    print(f"opening: {filename}")
    f = open(filename, "rb")
    audio_out = open(f"{filename}.pcm", "wb")
    get8 = lambda: struct.unpack(">B", f.read(1))[0]
    get16 = lambda: struct.unpack(">H", f.read(2))[0]
    get16_s = lambda: struct.unpack(">h", f.read(2))[0]
    get24 = lambda: struct.unpack(">I", b'\x00' + f.read(3))[0]
    get32 = lambda: struct.unpack(">I", f.read(4))[0]
    header = PACoHeader(get16(), get16(), get16(), get16(), get16_s(), get16(), get32(), get32(), get32(), get16(), get16(), get16(), get16(), get32(), get16())
    header.check()
    print(header)
    frame_sizes = [get32() for i in range(header.frames)]
    for i in range(header.frames):
        next_frame = f.tell() + frame_sizes[i]
        print("new frame")
        while (f.tell() < next_frame):
            chunk_type = get8()
            chunk_size = get24()
            if chunk_type in CHUNK_TYPES:
                chunk_name = CHUNK_TYPES[chunk_type]
            else:
                chunk_name = f"{chunk_type}: unknown"
            print(f"chunk type: {chunk_name} size: {chunk_size}")
            if chunk_type == 2:
                palette_data = f.read(chunk_size-4)
                quicktime_palette = b'\x30\x00\x00\x00\x00\x00\x00\x00' 
                custom_palette = b'\x10\x00\x00\x00\x00\x00\x00\x00'
                if palette_data.startswith(quicktime_palette):
                    print("QuicktimePalette")
                elif palette_data.startswith(custom_palette):
                    print("Defines custom palette")
                else:
                    print("unkown palette")
            elif chunk_type == 4:
                # 8 bit unsigned pcm
                sampling_rates = [5563, 7418, 11127, 22254]
                header = get16()
                get16()
                index = (header >> 10) & 7
                print(f"\t sampling rate: {sampling_rates[index]}")
                audio_out.write(f.read(chunk_size-8))
            else:
                f.seek(f.tell() + chunk_size - 4)
    audio_out.close()
    
if __name__ == "__main__":
    main()