Get data from DWG

main.py

import sys
import itertools
import pdb


def find_sentinel(file, sentinel):
    counter = 0
    pattern = []
    for i in itertools.count():
        buffer = file.read(1)
        # NOTE: buffer[0] is 48 where buffer is equal to b'0\x00'.
        if buffer[0] == sentinel[counter]:
            # pdb.set_trace()
            pattern.append(buffer)
            counter += 1
            if counter >= len(sentinel):
                print('Index:', i)
                break
        else:
            if 0 < len(pattern) < len(sentinel):
                pattern.clear()
                counter = 0


def get_data(file, size):
    pattern = []
    for i in range(size):
        buffer = file.read(1)
        pattern.append(buffer)
    return b''.join(pattern)


def main():
    try:
        raw = sys.argv[1]
    except IndexError:
        print('Usage:', 'python main.py draw.dwg')
        sys.exit(1)

    try:
        with open(raw, "rb") as file:

            # NOTE: Data section AcDb:Preview (R13C3 AND LATER) [pag. 93]
            sentinel_start = [
                0x1F, 0x25, 0x6D, 0x07, 0xD4, 0x36, 0x28, 0x28,
                0x9D, 0x57, 0xCA, 0x3F, 0x9D, 0x44, 0x10, 0x2B]
            sentinel_end = [
                0xE0, 0xDA, 0x92, 0xF8, 0x2B, 0xc9, 0xD7, 0xD7,
                0x62, 0xA8, 0x35, 0xC0, 0x62, 0xBB, 0xEF, 0xD4]
            find_sentinel(file, sentinel_start)
            # overall size (overall size of image area)
            data = get_data(file, 4)
            # imagespresent (counter indicating what is present here)
            data = get_data(file, 1)
            # code (code indicating what follows)
            data = get_data(file, 1)
            # header data start if code == 1 (start of header data)
            data = get_data(file, 4)
            # header data size if code == 1 (size of header data)
            data = get_data(file, 4)
            print(data.hex())

            # TODO: Summary info Address in stream
            # R2007 Meta Data [pag. 38]
            # Data section AcDb:SummaryInfo Section [pag. 91]
            # buffer = file.seek(int('20', 16))

    except IOError:
        print('Could not open {}'.format(raw))
        sys.exit(1)


if __name__ == '__main__':
    main()

Following the specification from OpenDesign.