OqeeRewind/utils/stream.py

"""Utility module for streaming and manifest parsing."""

import xml.etree.ElementTree as ET
import base64
import os
import asyncio
import time
from typing import Dict, Any

import requests
import aiohttp
from tqdm.asyncio import tqdm
from utils.logging_config import logger


def parse_mpd_manifest(mpd_content: str) -> Dict[str, Any]:
    """Parse an MPD manifest and extract metadata.

    Args:
        mpd_content: The MPD manifest content as a string.

    Returns:
        A dictionary containing parsed manifest information.
    """
    root = ET.fromstring(mpd_content)
    namespaces = {"mpd": "urn:mpeg:dash:schema:mpd:2011", "cenc": "urn:mpeg:cenc:2013"}

    manifest_info = {
        "type": root.get("type"),
        "profiles": root.get("profiles"),
        "publishTime": root.get("publishTime"),
        "availabilityStartTime": root.get("availabilityStartTime"),
        "minimumUpdatePeriod": root.get("minimumUpdatePeriod"),
        "minBufferTime": root.get("minBufferTime"),
        "timeShiftBufferDepth": root.get("timeShiftBufferDepth"),
        "suggestedPresentationDelay": root.get("suggestedPresentationDelay"),
        "periods": [],
    }

    for period in root.findall("mpd:Period", namespaces):
        period_info = {
            "id": period.get("id"),
            "start": period.get("start"),
            "adaptation_sets": [],
        }
        for adaptation_set in period.findall("mpd:AdaptationSet", namespaces):
            adaptation_info = parse_adaptation_set(adaptation_set, namespaces)
            period_info["adaptation_sets"].append(adaptation_info)
        manifest_info["periods"].append(period_info)
    return manifest_info


def parse_adaptation_set(
    adaptation_set: ET.Element, namespaces: Dict[str, str]
) -> Dict[str, Any]:
    """Parse an AdaptationSet element from MPD manifest.

    Args:
        adaptation_set: The AdaptationSet XML element.
        namespaces: XML namespaces dictionary.

    Returns:
        A dictionary containing parsed adaptation set information.
    """
    adaptation_info = {
        "id": adaptation_set.get("id"),
        "group": adaptation_set.get("group"),
        "contentType": adaptation_set.get("contentType"),
        "lang": adaptation_set.get("lang"),
        "segmentAlignment": adaptation_set.get("segmentAlignment"),
        "startWithSAP": adaptation_set.get("startWithSAP"),
        "drm_info": [],
        "representations": [],
    }

    # Parse ContentProtection
    for content_protection in adaptation_set.findall(
        "mpd:ContentProtection", namespaces
    ):
        drm_info = parse_content_protection(content_protection, namespaces)
        adaptation_info["drm_info"].append(drm_info)

    # Parse Role
    role = adaptation_set.find("mpd:Role", namespaces)
    if role is not None:
        adaptation_info["role"] = role.get("value")

    # Parse Representations
    for representation in adaptation_set.findall("mpd:Representation", namespaces):
        rep_info = parse_representation(representation, namespaces)
        adaptation_info["representations"].append(rep_info)

    return adaptation_info


def parse_content_protection(
    content_protection: ET.Element, namespaces: Dict[str, str]
) -> Dict[str, Any]:
    """Parse ContentProtection element for DRM information.

    Args:
        content_protection: The ContentProtection XML element.
        namespaces: XML namespaces dictionary.

    Returns:
        A dictionary containing DRM information.
    """
    drm_info = {
        "schemeIdUri": content_protection.get("schemeIdUri"),
        "value": content_protection.get("value"),
    }

    default_kid = content_protection.get("{urn:mpeg:cenc:2013}default_KID")
    if default_kid:
        drm_info["default_KID"] = default_kid

    pssh_element = content_protection.find("cenc:pssh", namespaces)
    if pssh_element is not None and pssh_element.text:
        drm_info["pssh"] = pssh_element.text.strip()
        try:
            pssh_decoded = base64.b64decode(drm_info["pssh"])
            drm_info["pssh_hex"] = pssh_decoded.hex()
        except (ValueError, base64.binascii.Error):
            pass

    return drm_info


def parse_representation(
    representation: ET.Element, namespaces: Dict[str, str]
) -> Dict[str, Any]:
    """Parse Representation element from MPD manifest.

    Args:
        representation: The Representation XML element.
        namespaces: XML namespaces dictionary.

    Returns:
        A dictionary containing parsed representation information.
    """
    rep_info = {
        "id": representation.get("id"),
        "bandwidth": representation.get("bandwidth"),
        "codecs": representation.get("codecs"),
        "mimeType": representation.get("mimeType"),
        "width": representation.get("width"),
        "height": representation.get("height"),
        "frameRate": representation.get("frameRate"),
        "segments": {},
    }

    segment_template = representation.find("mpd:SegmentTemplate", namespaces)
    if segment_template is not None:
        rep_info["segments"] = {
            "timescale": segment_template.get("timescale"),
            "initialization": segment_template.get("initialization"),
            "media": segment_template.get("media"),
            "timeline": [],
        }

        segment_timeline = segment_template.find("mpd:SegmentTimeline", namespaces)
        if segment_timeline is not None:
            for s_element in segment_timeline.findall("mpd:S", namespaces):
                timeline_info = {
                    "t": (
                        int(s_element.get("t")) if s_element.get("t") is not None else 0
                    ),  # start time
                    "d": (
                        int(s_element.get("d")) if s_element.get("d") is not None else 0
                    ),  # duration
                    "r": (
                        int(s_element.get("r")) if s_element.get("r") is not None else 0
                    ),  # repeat count
                }
                rep_info["segments"]["timeline"].append(timeline_info)

    return rep_info


# pylint: disable=too-many-locals,too-many-branches
def organize_by_content_type(manifest_info: Dict[str, Any]) -> Dict[str, Any]:
    """Organize manifest information by content type.

    Args:
        manifest_info: Parsed manifest information dictionary.

    Returns:
        A dictionary organized by content type (video, audio, text).
    """
    organized = {
        "video": {},
        "audio": {},
        # 'text': {},
        # 'manifest_metadata': {
        #     'type': manifest_info.get('type'),
        #     'publishTime': manifest_info.get('publishTime'),
        #     'minBufferTime': manifest_info.get('minBufferTime'),
        # }
    }

    for period in manifest_info.get("periods", []):
        for adaptation_set in period.get("adaptation_sets", []):
            content_type = adaptation_set.get("contentType")

            if not content_type:
                continue

            for rep in adaptation_set.get("representations", []):
                track_info = {
                    "track_id": rep.get("id"),
                    "adaptation_set_id": adaptation_set.get("id"),
                    "bandwidth": int(rep.get("bandwidth", 0)),
                    "bitrate_kbps": int(rep.get("bandwidth", 0)) // 1000,
                    "codec": rep.get("codecs"),
                    "mime_type": rep.get("mimeType"),
                    "drm_info": adaptation_set.get("drm_info", []),
                    "segments": rep.get("segments", {}),
                }

                if content_type == "video":
                    width = rep.get("width")
                    height = rep.get("height")
                    frame_rate = rep.get("frameRate")

                    track_info.update(
                        {
                            "resolution": (
                                f"{width}x{height}" if width and height else "unknown"
                            ),
                            "width": int(width) if width else None,
                            "height": int(height) if height else None,
                            "frame_rate": frame_rate,
                        }
                    )

                    resolution_key = track_info["resolution"]
                    if resolution_key not in organized["video"]:
                        organized["video"][resolution_key] = []
                    organized["video"][resolution_key].append(track_info)

                elif content_type == "audio":
                    lang = adaptation_set.get("lang", "unknown")
                    role = adaptation_set.get("role", "main")

                    track_info.update(
                        {
                            "language": lang,
                            "role": role,
                        }
                    )

                    lang_key = f"{lang}_{role}"
                    if lang_key not in organized["audio"]:
                        organized["audio"][lang_key] = []
                    organized["audio"][lang_key].append(track_info)

                # elif content_type == 'text':
                #     lang = adaptation_set.get('lang', 'unknown')
                #     role = adaptation_set.get('role', 'caption')

                #     track_info.update({
                #         'language': lang,
                #         'role': role,
                #     })

                #     lang_key = f"{lang}_{role}"
                #     if lang_key not in organized['text']:
                #         organized['text'][lang_key] = []
                #     organized['text'][lang_key].append(track_info)

    # Sort video tracks by resolution (descending) and then by bitrate (descending)
    for resolution in organized["video"]:
        organized["video"][resolution].sort(key=lambda x: x["bandwidth"], reverse=True)

    # Sort audio tracks by bitrate (descending)
    for lang in organized["audio"]:
        organized["audio"][lang].sort(key=lambda x: x["bandwidth"], reverse=True)

    # Sort video resolutions by pixel count (descending)
    sorted_video = {}
    for resolution in sorted(
        organized["video"].keys(),
        key=lambda r: (
            int(r.split("x")[0]) * int(r.split("x")[1])
            if "x" in r and r.split("x")[0].isdigit()
            else 0
        ),
        reverse=True,
    ):
        sorted_video[resolution] = organized["video"][resolution]
    organized["video"] = sorted_video

    return organized


def get_manifest(manifest_id):
    """Fetch the MPD manifest for a given channel ID.

    Args:
        manifest_id: The channel/manifest identifier.

    Returns:
        The manifest content as text.
    """
    headers = {
        "accept": "*/*",
        "accept-language": "en-GB,en-US;q=0.9,en;q=0.8",
        "cache-control": "no-cache",
        "origin": "https://tv.free.fr",
        "pragma": "no-cache",
        "priority": "u=1, i",
        "referer": "https://tv.free.fr/",
        "sec-ch-ua": '"Google Chrome";v="143", "Chromium";v="143", "Not A(Brand";v="24"',
        "sec-ch-ua-mobile": "?0",
        "sec-ch-ua-platform": '"macOS"',
        "sec-fetch-dest": "empty",
        "sec-fetch-mode": "cors",
        "sec-fetch-site": "cross-site",
        "user-agent": (
            "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
            "AppleWebKit/537.36 (KHTML, like Gecko) "
            "Chrome/143.0.0.0 Safari/537.36"
        ),
    }

    format_id = 1
    url = (
        "https://api-proxad.dc2.oqee.net/playlist/v1/live/"
        f"{manifest_id}/{format_id}/live.mpd"
    )
    response = requests.get(url, headers=headers, timeout=10)
    return response.text


async def fetch_segment(session, ticks, track_id):
    """Fetch a media segment asynchronously.

    Args:
        session: The aiohttp ClientSession.
        ticks: The tick value for the segment.
        track_id: The track identifier.

    Returns:
        The tick value if successful, None otherwise.
    """
    url = f"https://media.stream.proxad.net/media/{track_id}_{ticks}"
    headers = {
        "Accept": "*/*",
        "Referer": "https://tv.free.fr/",
        "User-Agent": (
            "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
            "AppleWebKit/537.36 (KHTML, like Gecko) "
            "Chrome/143.0.0.0 Safari/537.36"
        ),
    }
    try:
        async with session.get(url, headers=headers) as resp:
            if resp.status == 200:
                return ticks
            return None
    except aiohttp.ClientError:
        return None


def get_init(output_folder, track_id):
    """Download and save the initialization segment for a track.

    Args:
        output_folder: The output folder path.
        track_id: The track identifier.
    """
    url = f"https://media.stream.proxad.net/media/{track_id}_init"
    headers = {
        "Accept": "*/*",
        "Referer": "https://tv.free.fr/",
        "User-Agent": (
            "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
            "AppleWebKit/537.36 (KHTML, like Gecko) "
            "Chrome/143.0.0.0 Safari/537.36"
        ),
    }
    response = requests.get(url, headers=headers, timeout=10)
    if response.status_code == 200:
        os.makedirs(f"{output_folder}/segments_{track_id}", exist_ok=True)
        init_path = f"{output_folder}/segments_{track_id}/init.mp4"
        with open(init_path, "wb") as f:
            f.write(response.content)
        logger.debug("Saved initialization segment to %s", init_path)
    return init_path


async def save_segments(output_folder, track_id, start_tick, rep_nb, duration, batch_size=64):
    """Download and save multiple media segments in batches.

    Args:
        output_folder: The output folder path.
        track_id: The track identifier.
        start_tick: The starting tick value.
        rep_nb: The number of segments to download.
        duration: The duration per segment.
        batch_size: Number of concurrent downloads per batch (default: 16).
    """
    os.makedirs(f"{output_folder}/segments_{track_id}", exist_ok=True)

    async def download_segment(session, tick, rep):
        """Download a single segment.
        
        Returns:
            Tuple of (success: bool, tick: int, rep: int)
        """
        url = f"https://media.stream.proxad.net/media/{track_id}_{tick}"
        headers = {
            "Accept": "*/*",
            "Referer": "https://tv.free.fr/",
            "User-Agent": (
                "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
                "AppleWebKit/537.36 (KHTML, like Gecko) "
                "Chrome/143.0.0.0 Safari/537.36"
            ),
        }
        try:
            async with session.get(url, headers=headers) as resp:
                if resp.status == 200:
                    content = await resp.read()
                    filename = f"{output_folder}/segments_{track_id}/{tick}.m4s"
                    with open(filename, "wb") as f:
                        f.write(content)
                    return (True, tick, rep)
                logger.error(
                    "Failed to download segment %d (tick %d): HTTP %d",
                    rep, tick, resp.status
                )
                return (False, tick, rep)
        except aiohttp.ClientError as e:
            logger.warning("Error downloading segment %d (tick %d): %s", rep, tick, e)
            return (False, tick, rep)

    logger.info("Starting download of %d segments in batches of %d...", rep_nb, batch_size)
    logger.debug("Track ID: %s", track_id)
    logger.debug("Base tick: %d", start_tick)

    start_time = time.time()

    # Build list of all segments to download
    segments_to_download = [(start_tick + i * duration, i) for i in range(rep_nb)]
    retry_list = []
    successful = 0

    async with aiohttp.ClientSession() as session:
        # Process segments in batches
        with tqdm(total=len(segments_to_download), desc="Downloading segments", unit="seg") as pbar:
            for batch_start in range(0, len(segments_to_download), batch_size):
                batch = segments_to_download[batch_start:batch_start + batch_size]
                tasks = [download_segment(session, tick, rep) for tick, rep in batch]
                
                results = await asyncio.gather(*tasks)
                
                for success, tick, rep in results:
                    if success:
                        successful += 1
                    else:
                        retry_list.append((tick, rep))
                    pbar.update(1)

        # Retry failed segments
        if retry_list:
            logger.info("Retrying %d failed segments...", len(retry_list))
            retry_successful = 0
            final_failures = []
            
            with tqdm(total=len(retry_list), desc="Retrying segments", unit="seg") as pbar:
                for batch_start in range(0, len(retry_list), batch_size):
                    batch = retry_list[batch_start:batch_start + batch_size]
                    tasks = [download_segment(session, tick, rep) for tick, rep in batch]
                    
                    results = await asyncio.gather(*tasks)
                    
                    for success, tick, rep in results:
                        if success:
                            retry_successful += 1
                            successful += 1
                        else:
                            final_failures.append((tick, rep))
                        pbar.update(1)
            
            if final_failures:
                logger.warning(
                    "Failed to download %d segments after retry: %s",
                    len(final_failures),
                    [tick for tick, _ in final_failures]
                )
            else:
                logger.info("All %d retried segments downloaded successfully", retry_successful)

    end_time = time.time()
    elapsed = end_time - start_time

    logger.debug("Download completed in %.2fs", elapsed)
    logger.info("Successfully downloaded %d/%d segments", successful, rep_nb)
    logger.info("Files saved to %s/segments_%s/", output_folder, track_id)


def get_kid(output_folder, track_id):
    """Extract the Key ID (KID) from downloaded segments.

    Args:
        output_folder: The output folder path.
        track_id: The track identifier.

    Returns:
        The KID as a hex string if found, None otherwise.
    """
    folder = f"{output_folder}/segments_{track_id}"
    for filename in os.listdir(folder):
        if filename.endswith(".m4s"):
            filepath = os.path.join(folder, filename)
            logger.debug("Checking file: %s", filepath)
            with open(filepath, "rb") as f:
                data = f.read()
                # Pattern before KID
                index = data.find(
                    b"\x73\x65\x69\x67\x00\x00\x00\x14"
                    b"\x00\x00\x00\x01\x00\x00\x01\x10"
                )
                if index != -1:
                    kid_bytes = data[index + 16 : index + 16 + 16]
                    kid = kid_bytes.hex()
                    return kid
    return None