app.py

import requests
from bs4 import BeautifulSoup
import pandas as pd
import plotly.graph_objects as go
import streamlit as st

# 設定應用標題
st.title("餐廳資料抓取與分析")

# 從 Google 試算表中讀取 URLs
sheet_id = "1W20lawjiQtEpljUKoEaMVPDlSdnhvJLPUy2jk5xao_w"
urls_df = pd.read_csv(f"https://docs.google.com/spreadsheets/d/{sheet_id}/export?format=csv")

# 將 URLs 轉換為列表
urls = urls_df['網址'].tolist()

# 初始化一個空的 DataFrame 列表來儲存所有資料
df_list = []

# 迭代每個網址並抓取資料
for url in urls:
    response = requests.get(url)
    soup = BeautifulSoup(response.content, 'html.parser')

    # 解析並抓取所需資料
    title_tag = soup.find('h1', class_='restaurant-details__heading--title')
    title = title_tag.text.strip() if title_tag else 'N/A'
    
    address_tag = soup.find('li', class_='restaurant-details__heading--address')
    address = address_tag.text.strip() if address_tag else 'N/A'
    
    phone_tag = soup.find('a', {'data-event': 'CTA_tel'})
    phone = phone_tag['href'].replace('tel:', '') if phone_tag else 'N/A'
    
    description_tag = soup.find('div', class_='restaurant-details__description--text')
    description = description_tag.text.strip() if description_tag else 'N/A'

    # 將抓取的資料新增到列表中
    df_list.append({
        'Title': title,
        'Address': address,
        'Phone': phone,
        'Description': description
    })

# 使用 pd.DataFrame() 將所有資料合併成一個 DataFrame
df = pd.DataFrame(df_list)

# 顯示抓取的資料
st.subheader("抓取的餐廳資料")
st.dataframe(df)

# 統計每個區的商家數量
df['Area'] = df['Address'].str.extract(r'(\w+區)')  # 提取區域
area_counts = df['Area'].value_counts()  # 統計各區的商家數量

# 繪製柱狀圖
fig_bar = go.Figure(data=[go.Bar(x=area_counts.index, y=area_counts.values)])
fig_bar.update_layout(title='每個區的商家數量', xaxis_title='區域', yaxis_title='商家數量')

# 顯示柱狀圖
st.plotly_chart(fig_bar)

# 繪製圓餅圖
fig_pie = go.Figure(data=[go.Pie(labels=area_counts.index, values=area_counts.values)])
fig_pie.update_layout(title='每個區的商家數量比例')

# 按鈕來顯示圓餅圖
if st.button('顯示每個區的商家數量比例圓餅圖'):
    st.plotly_chart(fig_pie)