Youtube preview now use yt-dlp

2025-10-15 22:45:20 -05:00
parent 2d4ff9c1e2
commit dbd93583c1
2 changed files with 93 additions and 31 deletions
--- a/plugins/youtube-preview.py
+++ b/plugins/youtube-preview.py
@@ -7,11 +7,11 @@ import re
 import logging
 import asyncio
 import aiohttp
-from pytubefix import YouTube
+import yt_dlp
 import simplematrixbotlib as botlib
 from youtube_title_parse import get_artist_title

-LYRICIST_API_URL = "https://lyrist.vercel.app/api/{}"
+LYRICIST_API_URL = "https://lyrist.vercel.app/api/{}/{}"


 def seconds_to_minutes_seconds(seconds):
@@ -43,18 +43,26 @@ async def fetch_lyrics(song, artist):
    """
    try:
        async with aiohttp.ClientSession() as session:
-            async with session.get(LYRICIST_API_URL.format(song, artist)) as response:
-                data = await response.json()
-                return data.get("lyrics")
+            url = LYRICIST_API_URL.format(artist, song)
+            logging.info(f"Fetching lyrics from: {url}")
+            async with session.get(url, timeout=aiohttp.ClientTimeout(total=10)) as response:
+                if response.status == 200:
+                    data = await response.json()
+                    return data.get("lyrics")
+                else:
+                    logging.warning(f"Lyrics API returned status {response.status}")
+                    return None
+    except asyncio.TimeoutError:
+        logging.error("Timeout fetching lyrics")
+        return None
    except Exception as e:
        logging.error(f"Error fetching lyrics: {str(e)}")
        return None


-
 async def fetch_youtube_info(youtube_url):
    """
-    Asynchronously fetches information about a YouTube video.
+    Asynchronously fetches information about a YouTube video using yt-dlp.

    Args:
        youtube_url (str): The URL of the YouTube video.
@@ -64,26 +72,71 @@ async def fetch_youtube_info(youtube_url):
            None if an error occurs during fetching.
    """
    try:
-        video = YouTube(youtube_url)
-        title = video.title
-        artist, song = get_artist_title(title)
+        logging.info(f"Fetching YouTube info for: {youtube_url}")
+
+        # Configure yt-dlp options
+        ydl_opts = {
+            'quiet': True,
+            'no_warnings': True,
+            'extract_flat': False,
+            'skip_download': True,
+        }
+
+        # Run yt-dlp in thread pool to avoid blocking
+        loop = asyncio.get_event_loop()
+
+        def extract_info():
+            with yt_dlp.YoutubeDL(ydl_opts) as ydl:
+                return ydl.extract_info(youtube_url, download=False)
+
+        info = await loop.run_in_executor(None, extract_info)
+
+        if not info:
+            logging.error("No info returned from yt-dlp")
+            return None
+
+        # Extract video information
+        title = info.get('title', 'Unknown Title')
+        description = info.get('description', 'No description available')
+        duration = info.get('duration', 0)
+        view_count = info.get('view_count', 0)
+        uploader = info.get('uploader', 'Unknown')
+
+        logging.info(f"Video title: {title}")
+
+        length = seconds_to_minutes_seconds(duration)
+
+        # Parse artist and song from title
+        artist, song = get_artist_title(title)
+        logging.info(f"Parsed artist: {artist}, song: {song}")
+
+        # Limit description length to avoid huge messages
+        if len(description) > 500:
+            description = description[:500] + "..."

-        description = video.description
-        length = seconds_to_minutes_seconds(video.length)
-        views = video.views
-        author = video.author
        description_with_breaks = description.replace('\n', '<br>')

-        # Fetching lyrics
-        lyrics = await fetch_lyrics(song, artist)
-        lyrics = lyrics.replace('\n', "<br>")
+        # Build basic info message
+        info_message = f"""<strong>🎬🎝 Title:</strong> {title}<br><strong>Length:</strong> {length} | <strong>Views:</strong> {view_count:,} | <strong>Uploader:</strong> {uploader}<br><details><summary><strong>⤵︎Description⤵︎</strong></summary>{description_with_breaks}</details>"""
+
+        # Try to fetch lyrics if artist and song were parsed
+        if artist and song:
+            logging.info("Attempting to fetch lyrics...")
+            lyrics = await fetch_lyrics(song, artist)
+            if lyrics:
+                lyrics = lyrics.replace('\n', "<br>")
+                # Limit lyrics length
+                if len(lyrics) > 3000:
+                    lyrics = lyrics[:3000] + "<br>...(truncated)"
+                info_message += f"<br><details><summary><strong>🎵 Lyrics:</strong></summary><br>{lyrics}</details>"
+            else:
+                logging.info("No lyrics found")
+        else:
+            logging.info("Could not parse artist/song from title, skipping lyrics")

-        info_message = f"""<strong>🎬🎝 Title:</strong> {title} | <strong>Length</strong>: {length} minutes | <strong>Views</strong>: {views}\n<details><summary><strong>⤵︎Description⤵︎</strong></summary>{description_with_breaks}</details>"""
-        if lyrics:
-            info_message += f"<br><details><summary><strong>🎵 Lyrics:</strong></summary><br>{lyrics}</details>"
        return info_message
    except Exception as e:
-        logging.error(f"Error fetching YouTube video information: {str(e)}")
+        logging.error(f"Error fetching YouTube video information: {str(e)}", exc_info=True)
        return None


@@ -102,14 +155,20 @@ async def handle_command(room, message, bot, prefix, config):
        None
    """
    match = botlib.MessageMatch(room, message, bot, prefix)
-    if match.is_not_from_this_bot() and re.search(r'youtube\.com/watch\?v=', message.body):
-        logging.info("YouTube link detected")
-        video_id_match = re.search(r'youtube\.com/watch\?v=([^\s]+)', message.body)
+
+    # Check if message contains a YouTube link
+    if match.is_not_from_this_bot() and re.search(r'(youtube\.com/watch\?v=|youtu\.be/)', message.body):
+        logging.info(f"YouTube link detected in message: {message.body}")
+
+        # Match both youtube.com and youtu.be formats
+        video_id_match = re.search(r'(?:youtube\.com/watch\?v=|youtu\.be/)([a-zA-Z0-9_-]{11})', message.body)
+
        if video_id_match:
            video_id = video_id_match.group(1)
            youtube_url = f"https://www.youtube.com/watch?v={video_id}"
-            logging.info(f"Fetching information for YouTube video: {youtube_url}")
-            retry_count = 3
+            logging.info(f"Fetching information for YouTube video ID: {video_id}")
+
+            retry_count = 2  # Reduced retries since yt-dlp is more reliable
            while retry_count > 0:
                info_message = await fetch_youtube_info(youtube_url)
                if info_message:
@@ -117,8 +176,12 @@ async def handle_command(room, message, bot, prefix, config):
                    logging.info("Sent YouTube video information to the room")
                    break
                else:
-                    logging.info("Retrying...")
+                    logging.warning(f"Failed to fetch info, retrying... ({retry_count-1} attempts left)")
                    retry_count -= 1
-                    await asyncio.sleep(1)  # wait for 1 second before retrying
+                    if retry_count > 0:
+                        await asyncio.sleep(2)  # wait for 2 seconds before retrying
            else:
-                logging.error("Failed to fetch YouTube video information after retries")
+                logging.error("Failed to fetch YouTube video information after all retries")
+                await bot.api.send_text_message(room.room_id, "Failed to fetch YouTube video information. The video may be unavailable or age-restricted.")
+        else:
+            logging.warning("Could not extract video ID from YouTube URL")
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,6 +1,5 @@
 python-dotenv
 requests
-pytubefix
 duckduckgo_search
 nio
 markdown2
@@ -11,4 +10,4 @@ youtube_title_parse
 dnspython
 croniter
 schedule
-
+yt-dlp