find_wh_jp2k.py

import logging

log = logging.getLogger(__name__)

def get_dimensions(img_path: str) -> Tuple[int, int]:
    """
    Returns a tuple (w, h) of an image with a given ID. Reads the w/h data directly
    from the JPEG 2000 header.

    :param img_path: a path to a JPEG 2000 image
    :return: A tuple containing the width and height of the image. 0,0 if there was a problem indexing the file
    """
    log.debug("Getting dimensions for %s", img_path)
    width: int = 0
    height: int = 0

    try:
        f = open(img_path, 'rb')
    except FileNotFoundError:
        # Log this as a debug message, since we will catch this up the call stack and also include the master file
        # from which we found it.
        log.debug("%s was not found.", img_path)
        return width, height

    d = f.read(100)
    start_header = d.find(b'ihdr')

    if start_header == -1:
        log.error("Index error with %s. Malformed JPEG 2000 image. Setting to 0, 0", img_path)
        return width, height

    hs = start_header + 4
    ws = start_header + 8
    try:
        height = d[hs] * 256 ** 3 + d[hs + 1] * 256 ** 2 + d[hs + 2] * 256 + d[hs + 3]
        width = d[ws] * 256 ** 3 + d[ws + 1] * 256 ** 2 + d[ws + 2] * 256 + d[ws + 3]
    except IndexError:
        # If there was a problem reading the JP2, allow the process to continue. It should be easy to identify
        # failed images later as they will be the ones with w,h as 0,0.
        log.error("Index error with %s. Could not find width and height. Setting to 0, 0.", img_path)
    finally:
        f.close()

    return width, height