yt-local/youtube/ytdlp_proxy.py

#!/usr/bin/env python3
"""
Proxy for serving videos with specific audio using yt-dlp.

This module provides streaming functionality for unified formats
with specific audio languages.
"""
import logging
from flask import Response, request, stream_with_context
import urllib.request
import urllib.error
from youtube.ytdlp_service import find_best_unified_format

logger = logging.getLogger(__name__)


def stream_video_with_audio(video_id: str, audio_language: str = 'en', max_quality: int = 720):
    """
    Stream video with specific audio language.

    Args:
        video_id: YouTube video ID
        audio_language: Preferred audio language (default: 'en')
        max_quality: Maximum video height (default: 720)

    Returns:
        Flask Response with video stream, or 404 if not available
    """
    logger.info(f'Stream request: {video_id} | audio={audio_language} | quality={max_quality}p')

    # Find best unified format
    best_format = find_best_unified_format(video_id, audio_language, max_quality)

    if not best_format:
        logger.info(f'No suitable unified format found, returning 404 to trigger fallback')
        return Response('No suitable unified format available', status=404)

    url = best_format.get('url')
    if not url:
        logger.error('Format found but no URL available')
        return Response('Format URL not available', status=500)

    logger.debug(f'Streaming from: {url[:80]}...')

    # Stream the video
    try:
        req = urllib.request.Request(url)
        req.add_header('User-Agent', 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36')
        req.add_header('Accept', '*/*')

        # Add Range header if client requests it
        if 'Range' in request.headers:
            req.add_header('Range', request.headers['Range'])
            logger.debug(f'Range request: {request.headers["Range"]}')

        resp = urllib.request.urlopen(req, timeout=60)

        def generate():
            """Generator for streaming video chunks."""
            try:
                while True:
                    chunk = resp.read(65536)  # 64KB chunks
                    if not chunk:
                        break
                    yield chunk
            except Exception as e:
                logger.error(f'Stream error: {e}')
                raise

        # Build response headers
        response_headers = {
            'Content-Type': resp.headers.get('Content-Type', 'video/mp4'),
            'Access-Control-Allow-Origin': '*',
        }

        # Copy important headers
        for header in ['Content-Length', 'Content-Range', 'Accept-Ranges']:
            if header in resp.headers:
                response_headers[header] = resp.headers[header]

        status_code = resp.getcode()
        logger.info(f'Streaming started: {status_code}')

        return Response(
            stream_with_context(generate()),
            status=status_code,
            headers=response_headers,
            direct_passthrough=True
        )

    except urllib.error.HTTPError as e:
        logger.error(f'HTTP error streaming: {e.code} {e.reason}')
        return Response(f'Error: {e.code} {e.reason}', status=e.code)
    except urllib.error.URLError as e:
        logger.error(f'URL error streaming: {e.reason}')
        return Response(f'Network error: {e.reason}', status=502)
    except Exception as e:
        logger.error(f'Streaming error: {e}', exc_info=True)
        return Response(f'Error: {e}', status=500)