Create app.py

app.py

@@ -0,0 +1,628 @@
+import gradio as gr
+import requests
+import pandas as pd
+import time
+import json
+from urllib.parse import urlencode
+import resend
+from datetime import datetime
+import io
+from linebot.v3.messaging import Configuration, MessagingApi
+from linebot.v3.messaging.models import TextMessage
+from linebot import LineBotApi
+from linebot.models import TextSendMessage
+
+class MeetTaiwanAPIScraper:
+    def __init__(self):
+        self.base_url = "https://service.meettaiwan.com"
+        self.api_base = "https://service.meettaiwan.com/gpa/api/v2/events"
+
+        # 設定session
+        self.session = requests.Session()
+        self.session.headers.update({
+            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
+            'Accept': 'application/json, text/plain, */*',
+            'Accept-Language': 'zh-TW,zh;q=0.9,en;q=0.8',
+            'Accept-Encoding': 'gzip, deflate, br',
+            'Connection': 'keep-alive',
+            'Referer': 'https://service.meettaiwan.com/gpa/zh/events/list;type=all',
+            'Origin': 'https://service.meettaiwan.com'
+        })
+
+    def get_events_by_page(self, page=1, page_size=10, event_type=None, progress=None):
+        """調用API獲取指定頁面的活動資料"""
+        try:
+            # 構建API URL
+            params = {
+                'page': page,
+                'pageSize': page_size
+            }
+
+            if event_type:
+                params['type'] = event_type
+
+            # 嘗試不同的API端點
+            api_endpoints = [
+                f"{self.api_base}/",  # 主要API
+                f"{self.api_base}/tt-events"  # TT events API
+            ]
+
+            for api_url in api_endpoints:
+                try:
+                    response = self.session.get(api_url, params=params, timeout=30)
+
+                    if response.status_code == 200:
+                        try:
+                            data = response.json()
+
+                            # 檢查資料結構
+                            if isinstance(data, dict):
+                                # 常見的資料結構
+                                if 'data' in data:
+                                    events = data['data']
+                                elif 'items' in data:
+                                    events = data['items']
+                                elif 'results' in data:
+                                    events = data['results']
+                                elif 'events' in data:
+                                    events = data['events']
+                                else:
+                                    events = data
+
+                                if isinstance(events, list) and events:
+                                    return events, data
+                                else:
+                                    continue
+
+                            elif isinstance(data, list):
+                                return data, data
+
+                        except json.JSONDecodeError as e:
+                            continue
+
+                except Exception as e:
+                    continue
+
+            return None, None
+
+        except Exception as e:
+            return None, None
+
+    def get_all_events(self, progress_callback=None):
+        """獲取所有類型的活動資料"""
+        all_events = []
+        page_size_options = [50, 30, 20]
+        max_pages = 20
+
+        for page_size in page_size_options:
+            for page in range(1, max_pages + 1):
+                if progress_callback:
+                    progress_callback((page-1) * 5, f"正在獲取第 {page} 頁資料 (頁面大小: {page_size})")
+
+                events, raw_data = self.get_events_by_page(page=page, page_size=page_size, event_type=None)
+
+                if not events or len(events) == 0:
+                    if page == 1:
+                        break
+                    else:
+                        return all_events
+
+                # 處理事件資料
+                for event in events:
+                    processed_event = self.process_event_data(event, page, 'All')
+                    if processed_event:
+                        # 檢查是否已存在（避免重複）
+                        is_duplicate = False
+                        for existing_event in all_events:
+                            if (existing_event['name'] == processed_event['name'] and
+                                existing_event['event_date'] == processed_event['event_date']):
+                                is_duplicate = True
+                                break
+
+                        if not is_duplicate:
+                            all_events.append(processed_event)
+
+                # 如果這一頁的資料少於頁面大小，可能是最後一頁
+                if len(events) < page_size:
+                    return all_events
+
+                time.sleep(0.5)  # 減少延遲
+
+            # 如果成功獲取到資料，就不需要嘗試其他頁面大小
+            if all_events:
+                break
+
+        return all_events
+
+    def process_event_data(self, event, page_num, event_type):
+        """處理單個活動資料"""
+        try:
+            if isinstance(event, dict):
+                name = event.get('name', event.get('title', event.get('eventName', '')))
+                form = event.get('type', event.get('category', event.get('eventType', event_type or '')))
+                event_date = event.get('eventDate', event.get('startDate', event.get('date', '')))
+                upload_date = event.get('createdAt', event.get('uploadDate', event.get('publishDate', '')))
+
+                # 構建連結
+                event_id = event.get('id', event.get('eventId', ''))
+                if event_id:
+                    link = f"{self.base_url}/gpa/zh/events/{form}/{event_id}"
+                else:
+                    link = ""
+
+                return {
+                    'name': str(name),
+                    'link': link,
+                    'form': str(form),
+                    'event_date': str(event_date),
+                    'upload_date': str(upload_date),
+                    'page_num': page_num
+                }
+
+        except Exception as e:
+            pass
+
+        return None
+
+def create_html_table(df, max_display=10):
+    """將活動資料轉換為 HTML 表格格式"""
+    if df is None or df.empty:
+        return "<p>沒有找到活動資料</p>"
+
+    display_count = min(len(df), max_display)
+    df_display = df.head(display_count)
+    
+    # 創建 HTML 表格
+    html_content = f"""
+    <html>
+    <head>
+        <meta charset="UTF-8">
+        <style>
+            body {{ font-family: Arial, sans-serif; margin: 20px; line-height: 1.6; }}
+            .greeting {{ 
+                font-size: 18px; 
+                color: #2c3e50; 
+                margin-bottom: 20px; 
+                font-weight: bold;
+            }}
+            h2 {{ color: #2c3e50; margin-top: 20px; }}
+            .data-source {{
+                background-color: #e8f4f8;
+                padding: 15px;
+                border-left: 4px solid #3498db;
+                margin: 20px 0;
+                font-weight: bold;
+                color: #2c3e50;
+            }}
+            table {{ 
+                border-collapse: collapse; 
+                width: 100%; 
+                margin-top: 20px;
+                box-shadow: 0 2px 8px rgba(0,0,0,0.1);
+            }}
+            th, td {{ 
+                border: 1px solid #ddd; 
+                padding: 12px; 
+                text-align: left; 
+            }}
+            th {{ 
+                background-color: #3498db; 
+                color: white;
+                font-weight: bold;
+            }}
+            tr:nth-child(even) {{ background-color: #f2f2f2; }}
+            tr:hover {{ background-color: #e8f4f8; }}
+            a {{ color: #3498db; text-decoration: none; }}
+            a:hover {{ text-decoration: underline; }}
+            .summary {{ 
+                background-color: #ecf0f1; 
+                padding: 15px; 
+                border-radius: 5px; 
+                margin-bottom: 20px;
+            }}
+            .copyright {{
+                text-align: center;
+                margin-top: 30px;
+                padding: 20px;
+                background-color: #34495e;
+                color: white;
+                border-radius: 5px;
+                font-size: 14px;
+            }}
+        </style>
+    </head>
+    <body>
+        <div class="greeting">
+            親愛的會員您好：
+        </div>
+        
+        <div class="data-source">
+            📋 資料來源：全球政府採購商機網
+        </div>
+        
+        <h2>🎯 最新活動資訊</h2>
+        <div class="summary">
+            <strong>📊 資料統計：</strong>顯示前 {display_count} 筆，共 {len(df)} 筆活動
+        </div>
+        <table>
+            <thead>
+                <tr>
+                    <th>序號</th>
+                    <th>名稱</th>
+                    <th>形式</th>
+                    <th>活動日期</th>
+                    <th>上載日期</th>
+                    <th>網址</th>
+                </tr>
+            </thead>
+            <tbody>
+    """
+    
+    for idx, row in df_display.iterrows():
+        link_html = f'<a href="{row["超連結網址"]}" target="_blank">查看詳情</a>' if row["超連結網址"] else "無連結"
+        html_content += f"""
+                <tr>
+                    <td>{idx + 1}</td>
+                    <td><strong>{row['名稱']}</strong></td>
+                    <td>{row['形式']}</td>
+                    <td>{row['活動日期']}</td>
+                    <td>{row['上載日期']}</td>
+                    <td>{link_html}</td>
+                </tr>
+        """
+    
+    html_content += """
+            </tbody>
+        </table>
+    """
+    
+    if len(df) > max_display:
+        html_content += f"""
+        <div class="summary" style="margin-top: 20px;">
+            <strong>📝 提醒：</strong>還有 {len(df) - max_display} 筆資料未顯示，
+            請查看附加的 CSV 檔案獲取完整資料。
+        </div>
+        """
+    
+    html_content += """
+        <div class="summary" style="margin-top: 20px;">
+            <strong>🤖 自動爬蟲系統</strong><br>
+            此郵件由 MeetTaiwan API 爬蟲系統自動產生並發送
+        </div>
+        
+        <div class="copyright">
+            2025 © Copyright robert_studio
+        </div>
+    </body>
+    </html>
+    """
+    
+    return html_content
+
+def send_events_email(df_events, recipient_email, api_key, max_display=5):
+    """發送活動資料到指定郵箱"""
+    if df_events is None or df_events.empty:
+        return False, "沒有資料可發送"
+    
+    try:
+        # 設定 Resend API Key
+        resend.api_key = api_key
+        
+        # 取前N筆資料用於顯示
+        df_display = df_events.head(max_display)
+        
+        # 建立 HTML 內容
+        html_content = create_html_table(df_display, max_display=max_display)
+        
+        # 準備郵件主題
+        current_time = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+        subject = f"📊 MeetTaiwan 最新活動資訊 - {len(df_events)}筆活動 ({current_time})"
+        
+        # 發送郵件
+        r = resend.Emails.send({
+            "from": "onboarding@resend.dev",
+            "to": recipient_email,
+            "subject": subject,
+            "html": html_content
+        })
+        
+        return True, f"郵件發送成功！郵件 ID: {r.get('id', 'N/A')}"
+        
+    except Exception as e:
+        return False, f"郵件發送失敗: {str(e)}"
+
+def send_line_message(message, channel_access_token):
+    """
+    Send Line broadcast message to all users who have added the official account
+    """
+    try:
+        # Initialize LineBotApi
+        line_bot_api = LineBotApi(channel_access_token)
+
+        # Create text message
+        text_message = TextSendMessage(text=message)
+
+        # Broadcast message
+        line_bot_api.broadcast(text_message)
+
+        return True, "訊息成功發送到所有用戶！"
+
+    except Exception as e:
+        return False, f"發送訊息時發生錯誤: {e}"
+
+def scrape_events(progress=gr.Progress()):
+    """爬取活動資料的主函數"""
+    try:
+        progress(0, desc="初始化爬蟲...")
+        scraper = MeetTaiwanAPIScraper()
+        
+        def progress_callback(percent, message):
+            progress(percent/100, desc=message)
+        
+        progress(20, desc="開始爬取活動資料...")
+        events_data = scraper.get_all_events(progress_callback=progress_callback)
+        
+        if events_data and len(events_data) > 0:
+            progress(80, desc="處理資料...")
+            # 轉換為 DataFrame
+            df_events = pd.DataFrame(events_data)
+            df_events.columns = ["名稱", "超連結網址", "形式", "活動日期", "上載日期", "頁數"]
+            
+            # 去除重複資料
+            original_count = len(df_events)
+            df_events = df_events.drop_duplicates(subset=['名稱', '活動日期'])
+            deduplicated_count = len(df_events)
+            
+            progress(100, desc="完成！")
+            
+            success_msg = f"✅ 成功獲取 {deduplicated_count} 筆活動資料！"
+            if original_count != deduplicated_count:
+                success_msg += f"\n📝 去除了 {original_count - deduplicated_count} 筆重複資料"
+            
+            # 創建下載用的CSV
+            csv_data = df_events.to_csv(index=False, encoding='utf-8-sig')
+            
+            return (
+                success_msg,
+                df_events,
+                csv_data,
+                f"meettaiwan_events_{datetime.now().strftime('%Y%m%d_%H%M%S')}.csv"
+            )
+        else:
+            return "❌ 無法獲取活動資料，可能原因：API端點變更、需要認證或網路問題", None, None, None
+    
+    except Exception as e:
+        return f"❌ 爬取過程發生錯誤: {str(e)}", None, None, None
+
+def send_email_report(df_events, recipient_email, api_key, max_display):
+    """發送郵件報告"""
+    if df_events is None or df_events.empty:
+        return "❌ 沒有資料可發送，請先爬取資料"
+    
+    if not api_key.strip():
+        return "❌ 請輸入 Resend API Key"
+    
+    if not recipient_email.strip():
+        return "❌ 請輸入收件人郵箱"
+    
+    success, message = send_events_email(df_events, recipient_email, api_key, max_display)
+    
+    if success:
+        return f"✅ {message}"
+    else:
+        return f"❌ {message}"
+
+def send_line_notification(df_events, channel_access_token, message_template):
+    """發送 LINE 通知"""
+    if df_events is None or df_events.empty:
+        return "❌ 沒有資料可發送，請先爬取資料"
+    
+    if not channel_access_token.strip():
+        return "❌ 請輸入 LINE Channel Access Token"
+    
+    try:
+        # 準備訊息內容
+        current_time = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+        
+        if message_template.strip():
+            message = message_template
+        else:
+            # 默認訊息模板
+            top_5_events = df_events.head(5)
+            events_list = ""
+            for idx, row in top_5_events.iterrows():
+                events_list += f"{idx+1}. {row['名稱']} ({row['活動日期']})\n"
+            
+            message = f"""🎯 MeetTaiwan 最新活動通知
+
+📊 共找到 {len(df_events)} 筆活動資料
+⏰ 更新時間：{current_time}
+
+📋 最新 5 筆活動：
+{events_list}
+
+🤖 此訊息由自動爬蟲系統發送
+資料來源：全球政府採購商機網"""
+        
+        success, result_message = send_line_message(message, channel_access_token)
+        
+        if success:
+            return f"✅ {result_message}"
+        else:
+            return f"❌ {result_message}"
+    
+    except Exception as e:
+        return f"❌ 發送 LINE 通知時發生錯誤: {str(e)}"
+
+# 創建 Gradio 界面
+def create_interface():
+    with gr.Blocks(
+        title="🎯 MeetTaiwan 活動爬蟲系統", 
+        theme=gr.themes.Soft(),
+        css="""
+        .gradio-container {
+            font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif !important;
+        }
+        .gr-button-primary {
+            background: linear-gradient(45deg, #3498db, #2ecc71) !important;
+            border: none !important;
+        }
+        .gr-button-secondary {
+            background: linear-gradient(45deg, #e74c3c, #f39c12) !important;
+            border: none !important;
+        }
+        """
+    ) as app:
+        
+        # 標題區域
+        gr.Markdown("""
+        # 🎯 MeetTaiwan 活動爬蟲系統
+        **全球政府採購商機網活動資訊自動抓取與通知系統**
+        
+        本系統整合了資料爬取、郵件發送和 LINE 通知功能，讓您輕鬆獲取最新的活動資訊。
+        """)
+        
+        # 狀態變量
+        scraped_data = gr.State(None)
+        
+        with gr.Tabs():
+            # Tab 1: 資料爬取
+            with gr.Tab("📊 資料爬取", id="scraping"):
+                with gr.Row():
+                    with gr.Column(scale=2):
+                        gr.Markdown("### 🚀 開始爬取活動資料")
+                        scrape_btn = gr.Button(
+                            "開始爬取", 
+                            variant="primary", 
+                            size="lg"
+                        )
+                        
+                        scrape_status = gr.Textbox(
+                            label="爬取狀態", 
+                            interactive=False,
+                            placeholder="等待開始爬取..."
+                        )
+                    
+                    with gr.Column(scale=1):
+                        gr.Markdown("### 📥 資料下載")
+                        download_file = gr.File(
+                            label="下載 CSV 檔案",
+                            interactive=False
+                        )
+                
+                # 資料預覽
+                gr.Markdown("### 📋 資料預覽")
+                data_preview = gr.Dataframe(
+                    label="活動資料",
+                    wrap=True,
+                    interactive=False
+                )
+            
+            # Tab 2: 郵件發送
+            with gr.Tab("📧 郵件通知", id="email"):
+                with gr.Row():
+                    with gr.Column():
+                        gr.Markdown("### ⚙️ 郵件設定")
+                        email_api_key = gr.Textbox(
+                            label="Resend API Key",
+                            type="password",
+                            value="re_ZGacBiDw_HFEBpuCbaJ2S3NThPWiMU7Ex",
+                            placeholder="請輸入您的 Resend API Key"
+                        )
+                        recipient_email = gr.Textbox(
+                            label="收件人郵箱",
+                            value="cjhuang38@gmail.com",
+                            placeholder="請輸入收件人郵箱地址"
+                        )
+                        max_display_email = gr.Slider(
+                            label="郵件顯示筆數",
+                            minimum=5,
+                            maximum=20,
+                            value=10,
+                            step=1
+                        )
+                    
+                    with gr.Column():
+                        gr.Markdown("### 📨 發送郵件")
+                        send_email_btn = gr.Button(
+                            "發送郵件報告", 
+                            variant="secondary",
+                            size="lg"
+                        )
+                        email_status = gr.Textbox(
+                            label="郵件發送狀態",
+                            interactive=False,
+                            placeholder="等待發送..."
+                        )
+            
+            # Tab 3: LINE 通知
+            with gr.Tab("📱 LINE 通知", id="line"):
+                with gr.Row():
+                    with gr.Column():
+                        gr.Markdown("### ⚙️ LINE 設定")
+                        line_token = gr.Textbox(
+                            label="LINE Channel Access Token",
+                            type="password",
+                            value="FuM3pGmpqyOldcMltKVxkzBuy32o6mpkWv/gVfrR3sm9VxFUxTVzLlKU9C1ssOi2l/om2JkpKIdB/R+VLAyCvQA2o4pTD757kpN4GmUUq68FKuWwEaQXG376pR8hhyqUvElGn4rEYA7oxJDgsm4EBAdB04t89/1O/w1cDnyilFU=",
+                            placeholder="請輸入您的 LINE Channel Access Token"
+                        )
+                        message_template = gr.Textbox(
+                            label="自訂訊息內容 (選填)",
+                            placeholder="留空將使用默認模板...",
+                            lines=5
+                        )
+                    
+                    with gr.Column():
+                        gr.Markdown("### 📲 發送通知")
+                        send_line_btn = gr.Button(
+                            "發送 LINE 通知", 
+                            variant="secondary",
+                            size="lg"
+                        )
+                        line_status = gr.Textbox(
+                            label="LINE 發送狀態",
+                            interactive=False,
+                            placeholder="等待發送..."
+                        )
+        
+        # 頁尾
+        gr.Markdown("""
+        ---
+        <div style='text-align: center; color: #666; margin-top: 20px;'>
+            <p>🤖 MeetTaiwan API 爬蟲系統 | 2025 © Copyright robert_studio</p>
+            <p>資料來源：全球政府採購商機網</p>
+        </div>
+        """)
+        
+        # 事件綁定
+        scrape_btn.click(
+            fn=scrape_events,
+            outputs=[scrape_status, scraped_data, download_file, gr.State()]
+        ).then(
+            fn=lambda data: data if data is not None else gr.DataFrame(),
+            inputs=[scraped_data],
+            outputs=[data_preview]
+        )
+        
+        send_email_btn.click(
+            fn=send_email_report,
+            inputs=[scraped_data, recipient_email, email_api_key, max_display_email],
+            outputs=[email_status]
+        )
+        
+        send_line_btn.click(
+            fn=send_line_notification,
+            inputs=[scraped_data, line_token, message_template],
+            outputs=[line_status]
+        )
+    
+    return app
+
+if __name__ == "__main__":
+    app = create_interface()
+    app.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        share=True,
+        show_error=True
+    )