Update watcher.py

watcher.py

@@ -4,6 +4,7 @@ import time
 import json
 import requests
 from datetime import datetime, timezone
+from bs4 import BeautifulSoup
 
 # ===== Channel.io 設定 =====
 GET_URL = "https://desk-api.channel.io/desk/channels/200605/groups/519217/messages"
@@ -32,8 +33,8 @@ HEADERS_POST = {
     "x-account": X_ACCOUNT,
 }
 
-# ===== Vidfly API =====
-VIDFLY_API = "https://api.vidfly.ai/api/media/youtube/download?url="
+# ===== YouTube Downloader API =====
+YT_DETAIL_URL = "https://ssyoutube.online/yt-video-detail/"
 
 # ===== Utils =====
 def parse_updated_at(value):
@@ -59,26 +60,120 @@ def extract_youtube_id(text):
             return m.group(1)
     return None
 
+def get_video_download_links(youtube_url):
+    """YouTubeのダウンロードリンクを取得する"""
+    headers = {
+        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36",
+        "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8",
+        "Accept-Language": "ja,en-US;q=0.7,en;q=0.3",
+        "Accept-Encoding": "gzip, deflate, br",
+        "Content-Type": "application/x-www-form-urlencoded",
+        "Origin": "https://ssyoutube.online",
+        "Referer": "https://ssyoutube.online/",
+        "Upgrade-Insecure-Requests": "1"
+    }
+    
+    # フォームデータ
+    form_data = {
+        "videoURL": youtube_url
+    }
+    
+    try:
+        # HTMLを取得
+        response = requests.post(YT_DETAIL_URL, data=form_data, headers=headers, timeout=30)
+        response.raise_for_status()
+        
+        # BeautifulSoupで解析
+        soup = BeautifulSoup(response.text, 'html.parser')
+        
+        # 動画品質情報を抽出
+        video_links = []
+        
+        # テーブル行を検索
+        rows = soup.select('table.list tr')
+        for row in rows:
+            # ボタンからデータを抽出
+            button = row.select_one('button[data-url]')
+            if button:
+                data_url = button.get('data-url', '')
+                data_quality = button.get('data-quality', '')
+                data_has_audio = button.get('data-has-audio', 'false').lower() == 'true'
+                
+                # 品質ラベルを取得
+                quality_cell = row.select_one('td:first-child')
+                if quality_cell:
+                    quality_text = quality_cell.get_text(strip=True)
+                    # 数値の品質を抽出
+                    quality_match = re.search(r'(\d+)p', quality_text)
+                    if quality_match:
+                        quality = int(quality_match.group(1))
+                    else:
+                        # M4Aなどのオーディオ形式
+                        quality = 0
+                        if 'M4A' in quality_text:
+                            data_has_audio = True
+                else:
+                    quality = 0
+                
+                # サイズ情報を取得
+                size_cell = row.select('td')[1] if len(row.select('td')) > 1 else None
+                size_text = size_cell.get_text(strip=True) if size_cell else ""
+                
+                if data_url:
+                    video_links.append({
+                        "url": data_url,
+                        "ext": "mp4" if data_quality else "m4a",
+                        "height": quality,
+                        "type": "video_with_audio" if data_has_audio else "video",
+                        "quality": data_quality,
+                        "has_audio": data_has_audio,
+                        "size": size_text
+                    })
+        
+        return video_links
+        
+    except Exception as e:
+        print(f"HTML解析エラー: {e}")
+        return []
+
 def build_links(items):
+    """Channel.io用のリンクテキストを構築"""
     lines = []
     for item in items:
         url = item.get("url")
         ext = item.get("ext")
         height = item.get("height")
         media_type = item.get("type")
-
-        if not url or not ext or not height:
+        quality = item.get("quality", "")
+        has_audio = item.get("has_audio", False)
+        size = item.get("size", "")
+        
+        if not url:
             continue
-
-        label_type = "動画" if ext in ("mp4", "webm") else "音声"
-        audio_suffix = " (音声付き)" if media_type == "video_with_audio" else ""
-
-        line = f'<link type="url" value="{url}">{label_type}{audio_suffix} {height}p</link>'
+        
+        if ext == "m4a":
+            label_type = "音声"
+            audio_suffix = ""
+        else:
+            label_type = "動画"
+            audio_suffix = " (音声付き)" if has_audio else " (音声なし)"
+        
+        # 品質表示
+        if height > 0:
+            quality_display = f"{height}p"
+        else:
+            quality_display = quality if quality else "オーディオ"
+        
+        # サイズ情報があれば追加
+        size_display = f" [{size}]" if size else ""
+        
+        line = f'<link type="url" value="{url}">{label_type}{audio_suffix} {quality_display}{size_display}</link>'
         lines.append(line)
-
+    
     return "\n".join(lines)
 
 def send_to_channel(text):
+    """Channel.ioにメッセージを送信"""
     payload = {
         "requestId": f"desk-web-{int(time.time() * 1000)}",
         "blocks": [
@@ -93,19 +188,28 @@ def send_to_channel(text):
         "files": None,
         "customPayload": None
     }
-
-    res = requests.post(
-        POST_URL,
-        headers=HEADERS_POST,
-        data=json.dumps(payload),
-        timeout=30
-    )
-    res.raise_for_status()
+    
+    try:
+        res = requests.post(
+            POST_URL,
+            headers=HEADERS_POST,
+            data=json.dumps(payload),
+            timeout=30
+        )
+        res.raise_for_status()
+        print("Channel.ioに送信成功")
+        return True
+    except Exception as e:
+        print(f"Channel.io送信エラー: {e}")
+        return False
 
 # ===== Main =====
 def main():
+    processed_messages = set()  # 処理済みメッセージを追跡
+    
     while True:
         try:
+            # Channel.ioからメッセージを取得
             res = requests.get(
                 GET_URL,
                 headers=HEADERS_GET,
@@ -113,56 +217,62 @@ def main():
                 timeout=30,
             )
             res.raise_for_status()
-
+            
             messages = res.json().get("messages", [])
-
-            latest_msg = None
-            latest_time = None
-
+            
+            # 未処理のYouTubeリンクを含むメッセージを探す
             for msg in messages:
-                plain_text = msg.get("plainText")
-                updated_at = msg.get("updatedAt")
-
-                if not plain_text or updated_at is None:
+                msg_id = msg.get("id")
+                if msg_id in processed_messages:
                     continue
-
-                t = parse_updated_at(updated_at)
-                if not t:
+                
+                plain_text = msg.get("plainText", "")
+                if not plain_text:
                     continue
-
-                if latest_time is None or t > latest_time:
-                    latest_time = t
-                    latest_msg = msg
-
-            if not latest_msg:
-                time.sleep(10)
-                continue
-
-            text = latest_msg["plainText"]
-            youtube_id = extract_youtube_id(text)
-
-            if not youtube_id:
-                time.sleep(10)
-                continue
-
-            youtube_url = f"https://www.youtube.com/watch?v={youtube_id}"
-            vidfly_res = requests.get(VIDFLY_API + youtube_url, timeout=30)
-            vidfly_res.raise_for_status()
-
-            items = vidfly_res.json().get("data", {}).get("items", [])
-            if not items:
-                print("Vidfly の items が空です")
-                time.sleep(10)
-                continue
-
-            message_text = build_links(items)
-            send_to_channel(message_text)
-            print("送信完了")
-
+                
+                youtube_id = extract_youtube_id(plain_text)
+                if youtube_id:
+                    print(f"YouTubeリンクを検出: {youtube_id}")
+                    
+                    # 処理済みとしてマーク
+                    processed_messages.add(msg_id)
+                    
+                    # YouTube URLを構築
+                    youtube_url = f"https://www.youtube.com/watch?v={youtube_id}"
+                    
+                    # ダウンロードリンクを取得
+                    print(f"ダウンロードリンクを取得中: {youtube_url}")
+                    items = get_video_download_links(youtube_url)
+                    
+                    if items:
+                        print(f"{len(items)}個のダウンロードオプションを取得")
+                        
+                        # メッセージを構築して送信
+                        message_text = f"🎬 YouTube動画ダウンロードリンク\n\n"
+                        message_text += build_links(items)
+                        message_text += f"\n\n🔗 元の動画: {youtube_url}"
+                        
+                        if send_to_channel(message_text):
+                            print(f"ダウンロードリンクを送信完了: {youtube_id}")
+                        else:
+                            print(f"送信失敗: {youtube_id}")
+                    else:
+                        print(f"ダウンロードリンクの取得に失敗: {youtube_url}")
+                        error_msg = f"⚠️ ダウンロードリンクの取得に失敗しました\n\n🔗 {youtube_url}"
+                        send_to_channel(error_msg)
+            
+            # 処理済みメッセージが多くなりすぎたら古いものを削除
+            if len(processed_messages) > 100:
+                # 最新の50件だけ保持
+                processed_messages = set(list(processed_messages)[:50])
+            
+        except requests.exceptions.RequestException as e:
+            print(f"ネットワークエラー: {e}")
         except Exception as e:
-            print("エラー:", e)
-
+            print(f"予期せぬエラー: {e}")
+        
+        # 15秒間隔でチェック
         time.sleep(15)
 
 if __name__ == "__main__":
-    main()
+    main()