import os import urllib import urllib.parse import youtube_transcript_api from langchain.tools import BaseTool from youtube_transcript_api.proxies import WebshareProxyConfig class GetYouTubeTranscript(BaseTool): name: str = "get_youtube_transcript" description: str = ( "Get the transcript of a YouTube video. Input should be the YouTube video URL." ) def _run(self, url: str) -> str: """ Fetches the transcript of a YouTube video given its URL. Args: url (str): The URL of the YouTube video. Returns: str: The full transcript of the video as a single string if successful, or an error message if the transcript could not be fetched. """ try: parsed_url = urllib.parse.urlparse(url) qparams = urllib.parse.parse_qs(parsed_url.query) video_id = qparams.get("v", [None])[0] if not video_id: return "Video ID not found in the URL" api = youtube_transcript_api._api.YouTubeTranscriptApi( proxy_config=WebshareProxyConfig( proxy_username=os.getenv( "YOUTUBE_PROXY_USERNAME", "your_proxy_username" ), proxy_password=os.getenv( "YOUTUBE_PROXY_PASSWORD", "your_proxy_password" ), ) ) transcript = api.fetch(video_id) full_transcript = "".join([t.text for t in transcript]) except Exception as e: return f"Error fetching transcript: {e}" return full_transcript