Python - image metadata remover and date fixer.

img_metadata_strip_fix.py

#!/usr/bin/env python
#
# image metadata remover and date fixer.
# Corey Goldberg, 2014


"""Recursively scan a directory tree for image files, and fix metadata.

  * removes all metadata (Exif, IPTC, XMP, GPS Info, comment, thumbnail)
  * sets metadata and file timestamps to oldest datetime found.
  * works with .jpg and .png formats
  * modifications are done in-place.

Requires: gexiv2 (GObject-based wrapper around the Exiv2 library)
Install gexiv2 on Ubuntu/Debian: sudo apt-get install gir1.2-gexiv2-0.4
"""


import os
import time

from gi.repository import GExiv2


DATE_PATTERN = '%Y:%m:%d %H:%M:%S'


def fix_image_dates(img_path):
    ct = os.path.getctime(img_path)
    ctime = time.strftime(DATE_PATTERN, time.localtime(ct))
    mt = os.path.getmtime(img_path)
    mtime = time.strftime(DATE_PATTERN, time.localtime(mt))
    exif = GExiv2.Metadata(img_path)
    try:
        dtime = exif['Exif.Image.DateTime']
    except KeyError:
        dtime = None
    try:
        dtime_digitized = exif['Exif.Photo.DateTimeDigitized']
    except KeyError:
        dtime_digitized = None
    try:
        dtime_original = exif['Exif.Photo.DateTimeOriginal']
    except KeyError:
        dtime_original = None
    oldest_time = sorted(
        [x for x in (ctime, mtime, dtime, dtime_digitized, dtime_original)
            if x is not None]
        )[0]
    exif.clear_comment()
    exif.clear_exif()
    exif.clear_iptc()
    exif.clear_xmp()
    exif.delete_gps_info()
    exif.erase_exif_thumbnail()
    exif.clear()
    exif['Exif.Image.DateTime'] = oldest_time
    exif['Exif.Photo.DateTimeDigitized'] = oldest_time
    exif['Exif.Photo.DateTimeOriginal'] = oldest_time
    exif.save_file()
    epoch = int(time.mktime(time.strptime(oldest_time, DATE_PATTERN)))
    os.utime(img_path, (epoch, epoch))
    return oldest_time
    

if __name__ == '__main__':
    dir = '.'
    for root, dirs, file_names in os.walk(dir):
        for file_name in file_names:
            if file_name.lower().endswith(('jpg', 'png')):
                img_path = os.path.join(root, file_name)
                print('-' * 40)
                print(img_path)
                print(fix_image_dates(img_path))