scale_convert.py

import glob
import itertools
import os
import subprocess
from argparse import ArgumentParser
from collections import Counter

from joblib import Parallel, delayed

parser = ArgumentParser()
parser.add_argument("in_dir")
parser.add_argument("out_dir")
args = parser.parse_args()

os.makedirs(args.out_dir, exist_ok=True)

files = itertools.chain(
    glob.iglob(f"{args.in_dir}/*/*.jpg"),
    glob.iglob(f"{args.in_dir}/*/*.JGPG"),
    glob.iglob(f"{args.in_dir}/*/*.jpeg"),
    glob.iglob(f"{args.in_dir}/*/*.JPEG"),
    glob.iglob(f"{args.in_dir}/*/*.png"),
    glob.iglob(f"{args.in_dir}/*/*.PNG"),
    glob.iglob(f"{args.in_dir}/*/*.svg"),
    glob.iglob(f"{args.in_dir}/*/*.SVG"),
)


def process_file(path):
    basename = os.path.basename(path)
    name = os.path.splitext(basename)[0]

    try:
        r = subprocess.run(
            f'convert {path} -resize "224^>" -colorspace RGB -density 1200 {args.out_dir}/{name}.jpg',
            shell=True,
            timeout=10,
        )
        rcode = r.returncode
    except subprocess.TimeoutExpired:
        print("conversion timeout expired")
        rcode = -1

    if rcode == 0:
        os.remove(path)

    return rcode


codes = Parallel(n_jobs=32, prefer="threads", verbose=1)(
    delayed(process_file)(f) for f in files
)
print(Counter(codes))