from tools.interfaces.interface_video_transcriptor_provider import TranscriptProviderInterface
from youtube_transcript_api import YouTubeTranscriptApi
from youtube_transcript_api._errors import TranscriptsDisabled, NoTranscriptFound
import re

class YouTubeTranscriptProvider(TranscriptProviderInterface):
    def get_transcript(self, video_url_or_id: str) -> str:
        try:
            if "youtube.com" in video_url_or_id or "youtu.be" in video_url_or_id:
                match = re.search(r"(?:v=|\/)([0-9A-Za-z_-]{11})", video_url_or_id)
                video_id = match.group(1) if match else video_url_or_id
            else:
                video_id = video_url_or_id

            transcript = YouTubeTranscriptApi.get_transcript(video_id, languages=['fr', 'en'])
            return " ".join([t["text"] for t in transcript]) or "(Transcription) Vide."
        except TranscriptsDisabled:
            return "(Transcription) Sous-titres désactivés."
        except NoTranscriptFound:
            return "(Transcription) Aucune transcription disponible."
        except Exception as e:
            return f"(Transcription) Erreur : {e}"