"""paradisehill.cc — direct mp4 extractor.

Paradisehill embed strony renderują video.js z `og:video` meta tagiem wskazującym
na `/player/<id>/` iframe. Ten iframe zawiera inline JS:

    var videoList = [
        {"sources":[{"src":"https://v1.paradisehill.cc/video/<hash>_part1.mp4","type":"video/mp4"}]},
        {"sources":[{"src":"...part2.mp4",...}]},
        ...
    ];

Wieloczęściowe filmy są dzielone na part1..partN (~20-30 min każda). v1.paradisehill.cc
serwuje direct mp4 z Referer = paradisehill scene page; nie ma session auth ani token
bind (zweryfikowane 2026-05-15 z VPS Hetzner, status 200, ISO Media MP4).

Zwracamy listę StreamSource — jeden per part. Mobile player uznaje pierwszy element
(`best`) za główny; jeśli kiedyś potrzebowalibyśmy chapter switching, parts są w
`raw["parts"]` jako URL-e.
"""
from __future__ import annotations

import json
import logging
import re
from urllib.parse import urljoin

from app.extractors._fetch import browser_get, _DEFAULT_UA
from app.extractors._models import HosterDead, StreamSource

log = logging.getLogger(__name__)

_OG_VIDEO_RE = re.compile(r'<meta\s+property=["\']og:video["\']\s+content=["\']([^"\']+)["\']', re.IGNORECASE)
_VIDEOLIST_RE = re.compile(r'var\s+videoList\s*=\s*(\[[^;]+\]);', re.DOTALL)
_MP4_RE = re.compile(r'https?://[^\s"\'<>]+\.mp4(?:\?[^\s"\'<>]*)?', re.IGNORECASE)


def extract(page_url: str, *, timeout: float = 60.0) -> list[StreamSource] | None:
    headers = {
        "User-Agent": _DEFAULT_UA,
        "Accept": "text/html,application/xhtml+xml",
        "Accept-Language": "en-US,en;q=0.9",
    }
    r = browser_get(page_url, headers=headers, timeout=timeout)
    if r.status_code == 404 or r.status_code == 410:
        raise HosterDead(f"paradisehill {page_url}: HTTP {r.status_code}")
    if r.status_code != 200 or not r.text:
        log.info("paradisehill: page fetch fail %s status=%s", page_url, r.status_code)
        return None

    m = _OG_VIDEO_RE.search(r.text)
    if not m:
        log.info("paradisehill: no og:video meta in %s", page_url)
        return None
    player_url = urljoin(page_url, m.group(1))

    r2 = browser_get(player_url, headers={**headers, "Referer": page_url}, timeout=timeout)
    if r2.status_code != 200 or not r2.text:
        log.info("paradisehill: player iframe fail %s status=%s", player_url, r2.status_code)
        return None

    vl = _VIDEOLIST_RE.search(r2.text)
    parts: list[str] = []
    if vl:
        try:
            data = json.loads(vl.group(1))
            for item in data:
                for src in (item.get("sources") or []):
                    u = src.get("src")
                    if u and u not in parts:
                        parts.append(u)
        except json.JSONDecodeError as e:
            log.info("paradisehill: videoList JSON decode fail in %s: %s", player_url, e)

    if not parts:
        for m in _MP4_RE.finditer(r2.text):
            u = m.group(0)
            if u not in parts:
                parts.append(u)

    if not parts:
        log.info("paradisehill: no mp4 in player iframe %s", player_url)
        return None

    referer = page_url
    sources: list[StreamSource] = []
    for i, url in enumerate(parts):
        sources.append(
            StreamSource(
                link=url,
                quality=None,
                type="mp4",
                referer=referer,
                raw={"part_index": i, "total_parts": len(parts), "parts": parts} if len(parts) > 1 else None,
            )
        )
    return sources