summarizer.py

import os
import yt_dlp
from faster_whisper import WhisperModel
from ollama import Client
import logging

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

def download_audio(url):
    """
    Simple audio download function with minimal options
    """
    try:
        ydl_opts = {
            'format': 'm4a/bestaudio/best',
            'paths': {'home': './'},
            'outtmpl': {
                'default': 'audio.%(ext)s'
            },
            'postprocessors': [{
                'key': 'FFmpegExtractAudio',
                'preferredcodec': 'wav',
            }],
            'cookiefile': 'private.txt',  # Added cookiefile option
            # Add proxy if needed
            # 'proxy': 'socks5://proxy-server:1080',
        }

        with yt_dlp.YoutubeDL(ydl_opts) as ydl:
            error_code = ydl.download([url])
            
            if os.path.exists('audio.wav'):
                return 'audio.wav'
            raise Exception("Download failed to produce audio file")
            
    except Exception as e:
        logger.error(f"Download failed: {str(e)}")
        raise

def transcribe_audio(audio_file):
    """
    Transcribe audio using Whisper
    """
    try:
        model = WhisperModel("base", device="cpu", compute_type="int8")
        segments, _ = model.transcribe(audio_file)
        return " ".join([segment.text for segment in segments])
    except Exception as e:
        logger.error(f"Transcription failed: {str(e)}")
        raise
    finally:
        if os.path.exists(audio_file):
            os.remove(audio_file)

def generate_notes_and_summary_stream(transcript):
    """
    Generate summary using Ollama
    """
    try:
        prompt = f"""Based on this transcript, create:
        1. A set of concise, informative notes
        2. A brief summary of the main points

        Transcript: {transcript}

        Notes and Summary:"""

        client = Client()
        stream = client.generate(model='llama2:latest', prompt=prompt, stream=True)
        for chunk in stream:
            yield chunk['response']
    except Exception as e:
        logger.error(f"Summary generation failed: {str(e)}")
        raise

def process_video(url):
    """
    Process the video URL
    """
    try:
        audio_file = download_audio(url)
        transcript = transcribe_audio(audio_file)
        return generate_notes_and_summary_stream(transcript)
    except Exception as e:
        raise Exception(f"Processing failed: {str(e)}")