metube/app/mark_watched.py

"""
Mark videos as watched on various websites after successful download.
Uses the same cookies that were used for downloading.
"""

import os
import re
import logging
from urllib.parse import urlparse, parse_qs
from typing import Optional, Dict, Callable

from headless_watcher import headless_mark_watched, HAS_PLAYWRIGHT

log = logging.getLogger("mark_watched")

# Try to import curl_cffi for better impersonation, fallback to requests
try:
    from curl_cffi import requests as curl_requests
    HAS_CURL_CFFI = True
except ImportError:
    HAS_CURL_CFFI = False
    import requests


class MarkWatchedHandler:
    """Base class for mark-as-watched handlers."""
    
    def __init__(self, cookie_file: str):
        self.cookie_file = cookie_file
    
    def can_handle(self, url: str) -> bool:
        """Check if this handler can handle the given URL."""
        raise NotImplementedError
    
    async def mark_watched(self, url: str) -> bool:
        """Mark the video as watched. Returns True on success."""
        raise NotImplementedError
    
    def _load_cookies(self) -> Dict[str, str]:
        """Load cookies from the Netscape cookie file."""
        cookies = {}
        if not os.path.exists(self.cookie_file):
            log.warning(f"Cookie file not found: {self.cookie_file}")
            return cookies
        
        try:
            with open(self.cookie_file, "r") as f:
                for line in f:
                    line = line.strip()
                    if not line or line.startswith("#"):
                        continue
                    # Netscape format: domain flag path secure expiration name value
                    parts = line.split("\t")
                    if len(parts) >= 7:
                        name = parts[5]
                        value = parts[6]
                        cookies[name] = value
            log.debug(f"Loaded {len(cookies)} cookies from {self.cookie_file}")
        except Exception as e:
            log.error(f"Error loading cookies: {e}")
        
        return cookies
    
    def _make_request(self, method: str, url: str, **kwargs) -> Optional[object]:
        """Make HTTP request using curl_cffi if available, else requests."""
        try:
            if HAS_CURL_CFFI:
                # Use curl_cffi for better impersonation
                session = curl_requests.Session()
                # Set browser impersonation
                session.impersonate = "chrome124"
                response = session.request(method, url, **kwargs)
                return response
            else:
                return requests.request(method, url, **kwargs)
        except Exception as e:
            log.error(f"Request failed: {e}")
            return None


class PHHandler(MarkWatchedHandler):
    """Handler for PornHub."""
    
    DOMAINS = ["pornhub.com", "www.pornhub.com", "de.pornhub.com", "fr.pornhub.com", "es.pornhub.com", "it.pornhub.com", "rt.pornhub.com"]

    def can_handle(self, url):
        parsed = urlparse(url)
        domain = parsed.netloc.lower()
        return any(d in domain for d in self.DOMAINS)
    
    async def mark_watched(self, url):
        # Extract viewkey from URL
        parsed = urlparse(url)
        params = parse_qs(parsed.query)
        viewkey = params.get("viewkey", [None])[0]
        
        if not viewkey:
            # Try to extract from path
            match = re.search(r"viewkey=([^&]+)", url)
            if match:
                viewkey = match.group(1)
        
        if not viewkey:
            log.warning(f"Could not extract viewkey from URL: {url}")
            return False
        
        log.info(f"Marking video as watched, viewkey: {viewkey}")
        
        # Load cookies
        cookies = self._load_cookies()
        if not cookies:
            log.warning("No cookies available, cannot mark as watched")
            return False
        
        # Try multiple endpoints as PH may use different URLs
        # First try the AJAX endpoint which is most commonly used
        endpoints = [
            # AJAX endpoint (most common)
            ("POST", f"https://www.pornhub.com/user/watched/add/video/{viewkey}"),
            # Alternative format
            ("GET", f"https://www.pornhub.com/user/watched/add/video/{viewkey}"),
            # Legacy endpoint
            ("POST", f"https://www.pornhub.com/user/watched/video/viewkey/{viewkey}"),
        ]
        
        headers = {
            "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
            "Accept": "application/json, text/javascript, */*; q=0.01",
            "Accept-Language": "en-US,en;q=0.9",
            "X-Requested-With": "XMLHttpRequest",
            "Referer": url,
        }
        
        for method, api_url in endpoints:
            log.debug(f"Trying {method} {api_url}")
            response = self._make_request(method, api_url, headers=headers, cookies=cookies, allow_redirects=True)
            
            if response is None:
                continue
                
            log.debug(f"Response status: {response.status_code}")
            
            if response.status_code == 200:
                log.info(f"Successfully marked video as watched using {api_url}")
                return True
            elif response.status_code == 302:
                # Redirect often means success on PH
                log.info(f"Successfully marked video as watched (redirect from {api_url})")
                return True
        
        # API methods failed, try headless browser as fallback
        log.info("API methods failed, trying headless browser approach")
        if HAS_PLAYWRIGHT:
            return await headless_mark_watched(url, self.cookie_file, wait_seconds=5)
        else:
            log.warning("Playwright not available for headless browser fallback")
        
        return False


class YouTubeHandler(MarkWatchedHandler):
    """Handler for YouTube - uses YouTube API or internal endpoints."""
    
    DOMAINS = ["youtube.com", "www.youtube.com", "youtu.be", "m.youtube.com"]
    
    def can_handle(self, url):
        parsed = urlparse(url)
        domain = parsed.netloc.lower()
        return any(d in domain for d in self.DOMAINS)
    
    async def mark_watched(self, url):
        # YouTube marking as watched requires more complex handling
        # typically done through the browse endpoint with protobuf
        # This is a simplified implementation
        log.info("YouTube mark-as-watched not yet fully implemented")
        return False


# Registry of all handlers
HANDLERS = [
    PHHandler,
    YouTubeHandler,
]


def get_handler(url: str, cookie_file: str) -> Optional[MarkWatchedHandler]:
    """
    Get the appropriate handler for a URL.
    
    Args:
        url: The video URL
        cookie_file: Path to the Netscape cookie file
    
    Returns:
        Handler instance if found, None otherwise
    """
    if not cookie_file or not os.path.exists(cookie_file):
        return None
    
    for handler_class in HANDLERS:
        handler = handler_class(cookie_file)
        if handler.can_handle(url):
            return handler
    
    return None


async def mark_watched(url: str, cookie_file: str) -> bool:
    """
    Mark a video as watched on its respective site.
    
    Args:
        url: The video URL
        cookie_file: Path to the Netscape cookie file
    
    Returns:
        True if successfully marked as watched, False otherwise
    """
    handler = get_handler(url, cookie_file)
    if not handler:
        log.debug(f"No mark-watched handler available for URL: {url}")
        return False
    
    try:
        return await handler.mark_watched(url)
    except Exception as e:
        log.error(f"Error marking video as watched: {e}")
        return False