You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
96 lines
3.1 KiB
Python
96 lines
3.1 KiB
Python
#!/usr/bin/python3
|
|
|
|
import subprocess
|
|
import itertools
|
|
import os
|
|
import sys
|
|
import threading
|
|
|
|
|
|
GET_ALL_PICS_LIST_STRING="find . -iname '*.jpg' -o -iname '*.png' -o -iname '*.jpeg'"
|
|
COMPARE_PICS_CMD="compare -metric AE -fuzz 0.1% \"{}\" \"{}\" /dev/null"
|
|
|
|
PICS_MAP_SQUSHED = {}
|
|
|
|
def main():
|
|
out = subprocess.Popen(['sh', '-c', GET_ALL_PICS_LIST_STRING],
|
|
stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
|
|
all_pics_string = out.communicate()[0].decode('utf-8')
|
|
all_pics = all_pics_string.split('\n')
|
|
print(f"Gathered {len(all_pics)} photos.")
|
|
|
|
# I. basenames
|
|
pics_map = {}
|
|
for pic in all_pics:
|
|
key = os.path.basename('.'.join(pic.split('.')[:-1]))
|
|
if key not in pics_map:
|
|
pics_map.update({key: [pic]})
|
|
continue
|
|
pics_map[key].append(pic)
|
|
print(f"Hashtable for {len(pics_map)} keys.")
|
|
|
|
# II. sort
|
|
pics_map_keys = sorted(pics_map.keys(), key=len)
|
|
|
|
# III. squash
|
|
pics_map_squashed = {}
|
|
squash_marker = ['-', ' ', '(']
|
|
while pics_map_keys:
|
|
key = pics_map_keys.pop(0)
|
|
if len(key) < 7:
|
|
continue
|
|
if not key:
|
|
continue
|
|
squashed_keys = []
|
|
for other_key in pics_map_keys:
|
|
if other_key.startswith(key) and other_key[len(key)] in squash_marker:
|
|
squashed_keys.append(other_key)
|
|
if not squashed_keys:
|
|
continue
|
|
pics_map_squashed.update({key: pics_map[key]})
|
|
for squashed_key in squashed_keys:
|
|
pics_map_squashed[key] += pics_map[squashed_key]
|
|
pics_map_keys.remove(squashed_key)
|
|
|
|
print(f"Squashed hashtable to {len(pics_map_squashed)} keys.")
|
|
|
|
estimated = len(pics_map_squashed)
|
|
for pics in pics_map_squashed.values():
|
|
if estimated % 10 == 0:
|
|
print(f"Estimated: {estimated}")
|
|
print(f"Do {len(pics)} photos: '{pics}'")
|
|
estimated -= 1
|
|
delete_list = []
|
|
compars_count = 0
|
|
for pic_a, pic_b in itertools.combinations(pics, 2):
|
|
if pic_b in delete_list or pic_a in delete_list:
|
|
continue
|
|
if not os.path.exists(pic_a):
|
|
print(f"pic_a doesnt exist: '{pic_a}'")
|
|
continue
|
|
if not os.path.exists(pic_b):
|
|
print(f"pic_b doesnt exist: '{pic_b}'")
|
|
continue
|
|
compars_count += 1
|
|
out = subprocess.Popen(['sh', '-c', COMPARE_PICS_CMD.format(pic_a, pic_b)],
|
|
stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
|
|
value_string, stderr = out.communicate()
|
|
if out.returncode >= 2:
|
|
print(f"HERE!!! stdout: {value_string}\nstderr: {stderr}\nreturn code: {out.returncode}")
|
|
continue
|
|
#diff = float(value_string.decode('utf-8'))
|
|
#if diff == 0.:
|
|
if out.returncode == 0:
|
|
delete_list.append(pic_b)
|
|
|
|
print(f"Cmps: {compars_count}")
|
|
if not delete_list:
|
|
continue
|
|
print(f"Delete: {delete_list}")
|
|
for delete_elem in delete_list:
|
|
os.remove(delete_elem)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
main()
|