"""台股本益比 / 殖利率 / 股價淨值比查詢(TWSE 每日 BWIBBU_d)。 原始版本(phycomp/vwrStock pages/本益比II.py)為 Colab notebook 匯出, 含 google.colab / !pip 指令無法在 Streamlit 執行 — 此為乾淨重寫版。 """ from datetime import date, timedelta from io import StringIO import pandas as pd import requests import streamlit as st BASE_URL = "https://www.twse.com.tw/exchangeReport/BWIBBU_d" METRICS = ["本益比", "殖利率(%)", "股價淨值比"] _HEADERS = { "User-Agent": "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 " "(KHTML, like Gecko) Chrome/120.0 Safari/537.36" } @st.cache_data(ttl=3600, show_spinner=False) def fetch_bwibbu(day: date): """抓取 TWSE 個股日本益比資訊;該日無資料時自動往前找(最多 7 天)。 回傳 (DataFrame, 實際使用日期)。 """ for _ in range(7): datestr = day.strftime("%Y%m%d") try: resp = requests.get( BASE_URL, params={"response": "csv", "date": datestr, "selectType": "ALL"}, headers=_HEADERS, timeout=15, ) resp.raise_for_status() except requests.RequestException as e: raise RuntimeError(f"TWSE 連線失敗:{e}") from e try: df = pd.read_csv(StringIO(resp.text), header=1) except Exception: df = pd.DataFrame() if len(df) > 0 and "證券代號" in df.columns: return _clean(df), day day = day - timedelta(days=1) return pd.DataFrame(), day def _clean(df: pd.DataFrame) -> pd.DataFrame: df = df.copy() # 過濾說明列 / 合計列:證券代號須為純數字 df = df[df["證券代號"].astype(str).str.strip().str.isdigit()] for col in METRICS: if col in df.columns: df[col] = pd.to_numeric(df[col], errors="coerce") keep = ( ["證券代號", "證券名稱"] + [c for c in METRICS if c in df.columns] + (["財報日期"] if "財報日期" in df.columns else []) ) return df[keep].reset_index(drop=True) def render_stock_vwr(): st.title("📈 台股本益比 / 殖利率 / 股價淨值比") st.caption("資料來源:TWSE 每日個股日成交資訊(BWIBBU_d)") day = st.date_input("日期", value=date.today()) try: df, used = fetch_bwibbu(day) except RuntimeError as e: st.error(str(e)) return if df.empty: st.info(f"{used} 與其前 7 天皆無資料(假日或 TWSE 尚未更新),請換個日期試試。") return if used != day: st.info(f"{day} 當日無資料,已自動改用最近交易日:{used}") c1, c2, c3, c4 = st.columns([1.4, 1, 1, 1.6]) with c1: metric = st.selectbox("指標", METRICS, index=0) with c2: n = st.slider("Top N", 5, 50, 10) with c3: direction = st.radio("排序", ["最高", "最低"], horizontal=True) with c4: query = st.text_input("搜尋(代號或名稱)", "") st.divider() if query.strip(): q = query.strip() mask = df["證券代號"].astype(str).str.contains(q, na=False) | df[ "證券名稱" ].astype(str).str.contains(q, na=False) shown = df[mask] st.markdown(f"符合「{q}」:**{len(shown)} 檔**") else: valid = df.dropna(subset=[metric]) shown = ( valid.nlargest(n, metric) if direction == "最高" else valid.nsmallest(n, metric) ) st.markdown(f"{metric} **{direction} {n} 檔**({used},全市場共 {len(df)} 檔)") if shown.empty: st.info("沒有符合的資料。") return st.dataframe(shown, use_container_width=True)