import os 
import hashlib 
import logging 
import sys 
 
def logger(): 
 """ 獲取logger""" 
 logger = logging.getLogger() 
 if not logger.handlers: 
 # 指定logger輸出格式 
 formatter = logging.Formatter('%(asctime)s %(levelname)-8s: %(message)s') 
 # 文件日志 
 file_handler = logging.FileHandler("test.log") 
 file_handler.setFormatter(formatter) # 可以通過setFormatter指定輸出格式 
 # 控制臺(tái)日志 
 console_handler = logging.StreamHandler(sys.stdout) 
 console_handler.formatter = formatter # 也可以直接給formatter賦值 
 # 為logger添加的日志處理器 
 logger.addHandler(file_handler) 
 logger.addHandler(console_handler) 
 # 指定日志的最低輸出級(jí)別，默認(rèn)為WARN級(jí)別 
 logger.setLevel(logging.INFO) 
 return logger 
 
def get_md5(filename): 
 m = hashlib.md5() 
 mfile = open(filename, "rb") 
 m.update(mfile.read()) 
 mfile.close() 
 md5_value = m.hexdigest() 
 return md5_value 
 
def get_urllist(): 
 #替換指定的文件夾路徑即可 
 base = ("F:\\pythonFile\\煎蛋網(wǎng)\\無聊圖\\jpg\\") 
 list = os.listdir(base) 
 urlList=[] 
 for i in list: 
 url = base + i 
 urlList.append(url) 
 return urlList 
 
if __name__ == '__main__': 
 log = logger() 
 md5List =[] 
 urlList =get_urllist() 
 for a in urlList: 
 md5 =get_md5(a) 
 if (md5 in md5List): 
  os.remove(a) 
  print("重復(fù)：%s"%a) 
  log.info("重復(fù)：%s"%a) 
 else: 
  md5List.append(md5) 
  # print(md5List) 
  print("一共%s張照片"%len(md5List))