[python] 파일 리스트 읽고, 특정 파일 외 나머지 삭제하기

reads the file list.txt in which the file names are listed and deletes the rest leaving only the files that exist in it

 

import os
import glob
from tqdm import tqdm
import pdb

# Read the list of filenames from list.txt

for step in ["train","val", "test"]:
    print("=="+step+"==")
    with open(step+".txt", 'r') as f:
        filenames = f.read().splitlines()
    filenames = [f_n.split("/")[-1].rstrip("\n") for f_n in filenames]
    
    
    # Get a list of all the files in the directory
    imgs_in_dir = sorted(glob.glob(step+"/images/*"))
    gts_in_dir = sorted(glob.glob(step+"/labels/*"))
    
    # Loop through the files in the directory and delete any that aren't in the list
    for img_f, gts_f in tqdm(zip(imgs_in_dir, gts_in_dir)):
        file_nm = img_f.split("/")[-1].rstrip("\n")
        if file_nm not in filenames:
            # Check file names
            if os.path.splitext(os.path.split(img_f)[-1])[0]!=os.path.splitext(os.path.split(gts_f)[-1])[0]:
                pdb.set_trace()
            print("Delete: %s, %s"%(img_f, gts_f))
            os.remove(img_f)
            os.remove(gts_f)