import re def extract_date(url): return re.findall(r'/(\d{4})/(\d{1,2})/(\d{1,2})/', url) url1= "https://www.washingtonpost.com/news/football-insider/wp/2016/09/02/odell-beckhams-fame-rests-on-one-stupid-little-ball-josh-norman-tells-author/" print(extract_date(url1))