import datetime import pathlib import re import pandas as pd def str2date(s): y = datetime.date.today().year m, d = map(int, re.findall("\d{1,2}", s.split("~")[-1].strip())) return pd.Timestamp(y, m, d) df = pd.read_html("https://www.pref.aichi.jp/site/covid19-aichi/kansensya-kensa.html", match="検査日", na_values="-")[0].iloc[:-1] df["備考"] = df["検査日"] df.loc[df["検査日"].str.contains("~"), "合算"] = "○" df["検査日"] = df["検査日"].apply(str2date) p = pathlib.Path("data", "inspections_summary.csv") p.parent.mkdir(parents=True, exist_ok=True) df.to_csv(p, index=False, header=True)