Jul-23-2018, 04:04 PM
Well, I couldn't walk away. The Python gods would not allow it. So, for those browsing and absorbing info, I found a nice bit of coding on GitHub (which has been acquired by Microsoft now--uh oh! The king of closed source buying a valuable open source resource? I'm scared) which helped immensely. Here's my extracted and modified code:
import struct, datetime
matroska_tags = {
0x1A45DFA3 : ('MASTER', 'EBML'),
0x4286 : ('UINT', 'EBML Version'),
0x42F7 : ('UINT', 'EBML Read Version'),
0x42F2 : ('UINT', 'EBML Max ID Length'),
0x42F3 : ('UINT', 'EBML Max Size Length'),
0x4282 : ('STRING', 'Doc Type'),
0x4287 : ('UINT', 'Doc Type Version'),
0x4285 : ('UINT', 'Doc Type Read Version'),
0xEC : ('BINARY', 'Void Data'),
0xBF : ('BINARY', 'CRC-32'),
0x18538067 : ('MASTER', 'Segment'),
0x114D9B74 : ('MASTER', 'Seek Head'),
0x4DBB : ('MASTER', 'Seek'),
0x53AB : ('BINARY', 'Seek ID'),
0x53AC : ('UINT', 'Seek Position'),
0x1549A966 : ('MASTER', 'Information'),
0x73A4 : ('BINARY', 'Segment UID'),
0x7384 : ('STRING-8', 'Segment Filename'),
0x3CB923 : ('BINARY', 'Previous UID'),
0x3C83AB : ('STRING-8', 'Previous Filename'),
0x3EB923 : ('BINARY', 'Next UID'),
0x3E83BB : ('STRING-8', 'Next Filename'),
0x4444 : ('BINARY', 'Segment Family'),
0x6924 : ('MASTER', 'Chapter Translate'),
0x69FC : ('UINT', 'Chapter Translate Edition UID'),
0x69BF : ('UINT', 'Chapter Translate Codec'),
0x69A5 : ('BINARY', 'Chapter Translate ID'),
0x2AD7B1 : ('UINT', 'Timecode Scale'),
0x4489 : ('FLOAT', 'Duration'),
0x4461 : ('DATE', 'Date UTC'),
0x7BA9 : ('STRING-8', 'Title'),
0x4D80 : ('STRING-8', 'Muxing App'),
0x5741 : ('STRING-8', 'Writing App'),
0x1F43B675 : ('MASTER', 'Cluster'),
0xE7 : ('UINT', 'Timecode'),
0x5854 : ('MASTER', 'Silent Tracks'),
0x58D7 : ('UINT', 'Silent Track Number'),
0xA7 : ('UINT', 'Position'),
0xAB : ('UINT', 'Previous Size'),
0xA3 : ('BINARY', 'Simple Block'),
0xA0 : ('MASTER', 'Block Group'),
0xA1 : ('BINARY', 'Block'),
0x75A1 : ('MASTER', 'Block Additions'),
0xA6 : ('MASTER', 'Block More'),
0xEE : ('UINT', 'Block Add ID'),
0xA5 : ('BINARY', 'Block Additional'),
0x9B : ('UINT', 'Block Duration'),
0xFA : ('UINT', 'Reference Priority'),
0xFB : ('SINT', 'Reference Block'),
0xA4 : ('BINARY', 'Codec State'),
0x75A2 : ('SINT', 'Discard Padding'),
0x8E : ('MASTER', 'Slices'),
0xE8 : ('MASTER', 'Time Slice'),
0xCC : ('UINT', 'Lace Number'),
0x1654AE6B : ('MASTER', 'Tracks'),
0xAE : ('MASTER', 'Track Entry'),
0xD7 : ('UINT', 'Track Number'),
0x73C5 : ('UINT', 'Track UID'),
0x83 : ('UINT', 'Track Type'),
0xB9 : ('UINT', 'Flag Enabled'),
0x88 : ('UINT', 'Flag Default'),
0x55AA : ('UINT', 'Flag Lacing'),
0x6DE7 : ('UINT', 'Min Cache'),
0x6DF8 : ('UINT', 'Max Cache'),
0x23E383 : ('UINT', 'Default Duration'),
0x234E7A : ('UINT', 'Default Decoded Field Duration'),
0x55EE : ('UINT', 'Max Block Addition ID'),
0x536E : ('STRING-8', 'Name'),
0x22B59C : ('STRING', 'Language'),
0x86 : ('STRING', 'Codec ID'),
0x63A2 : ('BINARY', 'Codec Private'),
0x258688 : ('STRING-8', 'Codec Name'),
0x7446 : ('UINT', 'Attachment Length'),
0xAA : ('UINT', 'Codec Decode All'),
0x6FAB : ('UINT', 'Track Overlay'),
0x56AA : ('UINT', 'Codec Delay'),
0x56BB : ('UINT', 'Seek Preroll'),
0x6624 : ('MASTER', 'Track Translate'),
0x66FC : ('UINT', 'Track Translate Edition UID'),
0x66BF : ('UINT', 'Track Translate Codec'),
0x66A5 : ('BINARY', 'Track Translate Track ID'),
0xE0 : ('MASTER', 'Video'),
0x9A : ('UINT', 'Interlaced'),
0x9D : ('UINT', 'Field Order'),
0x53B8 : ('UINT', 'Stereo Mode'),
0x53C0 : ('UINT', 'Alpha Mode'),
0xB0 : ('UINT', 'Pixel Height'),
0xBA : ('UINT', 'Pixel Width')
}
def get_ebml_id(file):
byte = file.read(1)
dec = ord(byte)
if dec & 0b10000000: #1 byte long EBML ID
return byte
elif dec & 0b01000000: #2 bytes long EBML ID
return struct.unpack('>H', byte + file.read(1))[0]
elif dec & 0b00100000: #3 bytes long EBML ID
return struct.unpack('>L', b'\0' + byte + file.read(2))[0]
elif dec & 0b00010000: #4 bytes long EBML ID
return struct.unpack('>L', byte + file.read(3))[0]
def get_size(file):
dec = ord(file.read(1))
if dec & 0b10000000: #1 byte long
return dec & 0b01111111
elif dec & 0b01000000: #2 bytes long
return struct.unpack('>H', bytes([dec & 0b00111111]) + file.read(1))[0]
elif dec & 0b00100000: #3 bytes long
return struct.unpack('>L', b'\0' + bytes([dec & 0b00011111]) + file.read(2))[0]
elif dec & 0b00010000: #4 bytes long
return struct.unpack('>L', bytes([dec & 0b00001111]) + file.read(3))[0]
elif dec & 0b00001000: #5 bytes long
return struct.unpack('>Q', b'\0\0\0' + bytes([dec & 0b00000111]) + file.read(4))[0]
elif dec & 0b00000100: #6 bytes long
return struct.unpack('>Q', b'\0\0' + bytes([dec & 0b00000011]) + file.read(5))[0]
elif dec & 0b00000010: #7 bytes long
return struct.unpack('>Q', b'\0' + bytes([dec & 0b00000001]) + file.read(6))[0]
elif dec & 0b00000001: #8 bytes long
return struct.unpack('>Q', b'\0' + file.read(7))[0]
def read_integer(file, length, signed=False):
if length == 1:
v = ord(file.read(1))
elif length == 2:
v = struct.unpack('>H', file.read(2))[0]
elif length == 3:
v = struct.unpack('>L', b'\0' + file.read(3))[0]
elif length == 4:
v = struct.unpack('>L', file.read(4))[0]
elif length == 5:
v = struct.unpack('>Q', b'\0\0\0' + file.read(5))[0]
elif length == 6:
v = struct.unpack('>Q', b'\0\0' + file.read(6))[0]
elif length == 7:
v = struct.unpack('>Q', b'\0' + file.read(7))[0]
elif length == 8:
v = struct.unpack('>Q', file.read(8))[0]
if signed:
nbits = (8 - length) + 8 * (length - 1)
if v >= (1 << (nbits - 1)):
v -= 1 << nbits
return v
def read_float(file, length):
if length == 4:
return struct.unpack('>f', file.read(4))[0]
elif length == 8:
return struct.unpack('>d', file.read(8))[0]
def parse(file, _from=0, _to=None):
while file.tell() < _to:
ID = get_ebml_id(file)
size = get_size(file)
_type, _name = matroska_tags[ID]
value = 0
if _type == 'SINT':
value = read_integer(file, size, True)
elif _type == 'UINT':
value = read_integer(file, size)
elif _type == 'FLOAT':
value = read_float(file, size)
elif _type == 'STRING-8':
value = file.read(size).decode('utf-8')
elif _type == 'STRING':
value = file.read(size).decode('ascii')
elif _type == 'DATE':
micro = read_integer(file, size, True) / 1000.0
value = datetime.datetime(2001, 1, 1) + datetime.timedelta(microseconds=micro)
elif _type == 'BINARY':
value = file.read(size)
return print(_name + ':', value)
with open('D:\\Rampage_t01.mkv', 'rb') as f:
parse(f, 0, 96)It partially works. I believe the MASTER type needs work. I saw what is done in the linked code for that type, but it plain doesn't work for me. As you can see from comparing the linked code and my code, mine is stripped. I'm after a specific thing here, not a general purpose parser. So, I need to figure out the how to handle the MASTER type and see where that lands me. Once again, any help would be appreciated.
