import streamlit as st
import pandas as pd
import requests
from bs4 import BeautifulSoup
from google.oauth2.service_account import Credentials
import gspread
import plotly.express as px
import json  # 用於解析 JSON 文件

# Streamlit UI 標題
st.title("Booking.com 多項目數據抓取和視覺化展示")

# 合併的 URL 列表
urls = [
    "https://www.booking.com/hotel/it/appartamento-via-genova-roma.zh-tw.html",
    "https://www.booking.com/hotel/it/giulia-39-s-coliseum.zh-tw.html",
    "https://www.booking.com/hotel/it/le-stelle-dell-esquilino.zh-tw.html",
    "https://www.booking.com/hotel/it/retro-flat-termini.zh-tw.html",
    "https://www.booking.com/hotel/it/domus-fanti.zh-tw.html",
    "https://www.booking.com/hotel/it/radiant-retreat-in-esquilino.zh-tw.html",
    "https://www.booking.com/hotel/it/over-the-roof-top-roma.zh-tw.html",
    "https://www.booking.com/hotel/it/rome-sweet-home-roma1234.zh-tw.html",
    "https://www.booking.com/hotel/it/visione-guest-house-via-delle-fratte-36.zh-tw.html",
    "https://www.booking.com/hotel/it/appartamento-roma-centro-roma5.zh-tw.html",
    "https://www.booking.com/hotel/us/loews-royal-pacific-resort-at-universal.zh-tw.html",
    "https://www.booking.com/hotel/us/buena-vista-suites.zh-tw.html",
    "https://www.booking.com/hotel/us/ramada-international-drive.zh-tw.html",
    "https://www.booking.com/hotel/us/lake-buena-vista-14651-chelonia-parkway.zh-tw.html",
    "https://www.booking.com/hotel/it/trevi-apartment-roma1.zh-tw.html",
    "https://www.booking.com/hotel/it/white-flat-colosseo.zh-tw.html",
    "https://www.booking.com/hotel/it/via-cavour-238.zh-tw.html",
    "https://www.booking.com/hotel/it/modern-apartment-near-the-vatican-roma.zh-tw.html",
    "https://www.booking.com/hotel/it/apt-prati-lt-vatican-and-center.zh-tw.html",
    "https://www.booking.com/hotel/it/lion-99.zh-tw.html",
    "https://www.booking.com/hotel/it/sweet-dream-cavour-roma1.zh-tw.html",
    "https://www.booking.com/hotel/it/cavour-1.zh-tw.html",
    "https://www.booking.com/hotel/it/eufonia-vinyl-apartment.zh-tw.html",
    "https://www.booking.com/hotel/it/panisperna-apartment.zh-tw.html",
    "https://www.booking.com/hotel/it/navona-panoramic-penthouse.zh-tw.html"
]

# 設定 User-Agent
headers = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/87.0.4280.88 Safari/537.36"
}

# 定義抓取數據函數
def fetch_booking_data(url, headers):
    try:
        response = requests.get(url, headers=headers)
        soup = BeautifulSoup(response.content, "html.parser")
        title = soup.find("h2", class_="d2fee87262 pp-header__title").text.strip()
        rating = soup.find("div", class_="bcdcb105b3 f45d8e4c32 df64fda51b").text.strip()
        return {"標題": title, "評分": rating}
    except Exception as e:
        st.error(f"Error fetching data from {url}: {e}")
        return None

# 抓取數據並顯示
st.header("抓取數據")
data = []
for url in urls:
    result = fetch_booking_data(url, headers)
    if result:
        data.append(result)

# 將數據轉換為 DataFrame 並顯示
if data:
    df = pd.DataFrame(data)
    st.write("抓取到的數據：", df)
else:
    st.warning("無法從任何 URL 抓取數據。")

# 定義上傳到 Google Sheets 的函數
def upload_to_google_sheet(df, spreadsheet_url, creds_file_content):
    try:
        # 將金鑰內容轉換為字典
        creds_dict = json.loads(creds_file_content)
        # 創建憑據
        scope = ['https://www.googleapis.com/auth/spreadsheets']
        creds = Credentials.from_service_account_info(creds_dict, scopes=scope)
        # 授權並打開 Google Sheets
        gs = gspread.authorize(creds)
        sheet = gs.open_by_url(spreadsheet_url)
        worksheet = sheet.get_worksheet(0)
        # 清除並更新數據
        worksheet.clear()
        worksheet.update([df.columns.values.tolist()] + df.values.tolist())
        st.success("數據已成功上傳到 Google Sheets！")
    except Exception as e:
        st.error(f"數據上傳失敗：{e}")

# 上傳到 Google Sheets
st.header("上傳到 Google Sheets")
spreadsheet_url = st.text_input("Google Sheets URL", "https://docs.google.com/spreadsheets/d/1iOzoii9bVAmqlcqnseoqjZBkBuaFcpbIvUxZeRJ2kmk/edit?gid=0#gid=0")
creds_file = st.file_uploader("上傳 Google API 金鑰檔案", type=["json"])

if st.button("上傳數據至 Google Sheets") and data:
    if creds_file is not None:
        creds_content = creds_file.read()  # 讀取上傳的文件內容
        upload_to_google_sheet(df, spreadsheet_url, creds_content)
    else:
        st.error("請上傳 Google API 金鑰檔案")

# 數據視覺化
st.header("數據視覺化")
if data:
    # 提取數字評分
    df['數字評分'] = df['評分'].str.extract(r'評分：(\d+\.\d+)').astype(float)
    # 使用 Plotly 繪製條形圖
    fig = px.bar(df, x='標題', y='數字評分', title="Booking.com 評分比較", 
                 labels={'標題': '酒店標題', '數字評分': '評分'},
                 text='數字評分')
    fig.update_layout(xaxis_title="酒店標題", yaxis_title="評分", xaxis_tickangle=-45)
    st.plotly_chart(fig)
else:
    st.warning("無可用數據繪製圖表")