goon/app/connectors/direct_scrapers/xvideos_browse.py

"""xvideos.com, deep-crawl browse scraper (JSON-LD + page-parse).

xvideos SSR-uje JSON-LD VideoObject (duration, name, uploadDate) ORAZ na detail-stronie
linki `/models/<slug>` (performerzy tej sceny) + `/tags/<slug>` (tagi). Sample 2026-06-03
(15 scen): median ~10.5min, 93% ≥3min, dobry full-scene content (nie trailery).

Mega-katalog (~13M) → deep_crawl z per-tube page-cap (xvideoscom w deep_crawl._PAGE_CAP),
żeby nie monopolizował round-robin ani nie zalał bazy. Listing: /new/<page> (newest).
Scene: /video.<hash>/<slug>. Playback: page_url + origin tube:xvideoscom (istniejący
extractor `xvideoscom` resolvuje stream mobile-side). Phash pominięty (xvideos robi
własne crop-thumbnaile, 0% hit do canonical, jak fullmovies/hdporn).
"""
from __future__ import annotations

import json
import logging
import re
from datetime import date, datetime

from app.connectors.base import RawPerformer, RawPlaybackSource, RawScene, RawTag
from app.connectors.direct_scrapers._browse_base import BaseBrowseScraper, meta_content
from app.normalize.text import slugify

log = logging.getLogger(__name__)

_BASE = "https://www.xvideos.com"
_SCENE_URL_RE = re.compile(r'href="(/video\.[0-9a-z]+/[a-z0-9_]+)"', re.IGNORECASE)
_JSONLD_RE = re.compile(
    r'<script[^>]+type=["\']application/ld\+json["\'][^>]*>(.*?)</script>', re.IGNORECASE | re.DOTALL
)
# Model anchor: `<a href="/models/<slug>" ...><span ...></span><span class="name">Display Name</span>...`.
# Nazwa siedzi w ZAGNIEŻDŻONYM <span class="name">, nie jako bezpośredni tekst anchora,
# poprzedni wzorzec `>([^<]{2,60})</a>` wymagał tekstu zaraz po `>` i NIGDY nie pasował do
# obecnego markupu xvideos → wszystkie browse-scraped sceny xvideos lądowały z 0 performerami
# (bug-report 2026-06-07 "czemu nie ma aktorów, są na stronie"). `(?:(?!</a>).)*?` trzyma match
# w obrębie jednego anchora (nie przecieka do nazwy następnego modela gdy span.name brak).
_MODEL_RE = re.compile(
    r'/models/([a-z0-9_-]+)"(?:(?!</a>).)*?<span class="name">\s*([^<]{2,60})</span>',
    re.IGNORECASE | re.DOTALL,
)
_TAG_RE = re.compile(r'href="/tags/([a-z0-9_-]+)"', re.IGNORECASE)
_SETTITLE_RE = re.compile(r"html5player\.setVideoTitle\('([^']+)'\)")
_ISO_DUR_RE = re.compile(r"PT(?:(\d+)H)?(?:(\d+)M)?(?:(\d+)S)?", re.IGNORECASE)


def _dur_to_sec(value: str | None) -> int | None:
    if not value:
        return None
    m = _ISO_DUR_RE.match(str(value).strip())
    if not m:
        return None
    total = int(m.group(1) or 0) * 3600 + int(m.group(2) or 0) * 60 + int(m.group(3) or 0)
    return total or None


def _iso_date(value: str | None) -> date | None:
    if not value:
        return None
    try:
        return datetime.fromisoformat(str(value).replace("Z", "+00:00")).date()
    except ValueError:
        m = re.match(r"(\d{4}-\d{2}-\d{2})", str(value))
        return date.fromisoformat(m.group(1)) if m else None


def _video_object(html: str) -> dict | None:
    for m in _JSONLD_RE.finditer(html):
        raw = m.group(1).strip()
        if not raw:
            continue
        try:
            data = json.loads(raw)
        except (json.JSONDecodeError, ValueError):
            continue
        items = data if isinstance(data, list) else (data.get("@graph", [data]) if isinstance(data, dict) else [])
        for obj in items:
            if isinstance(obj, dict) and obj.get("@type") == "VideoObject":
                return obj
    return None


class XVideosBrowseScraper(BaseBrowseScraper):
    sitetag = "xvideoscom"

    def _listing_url(self, page: int) -> str:
        return f"{_BASE}/new/{page}"

    def _extract_scene_urls(self, listing_html: str) -> list[str]:
        seen: set[str] = set()
        out: list[str] = []
        for m in _SCENE_URL_RE.finditer(listing_html):
            url = f"{_BASE}{m.group(1)}"
            if url in seen:
                continue
            seen.add(url)
            out.append(url)
        return out

    def _parse_detail(self, scene_url: str, detail_html: str) -> RawScene | None:
        video = _video_object(detail_html) or {}

        title = (video.get("name") or "").strip()
        if not title:
            m = _SETTITLE_RE.search(detail_html)
            title = m.group(1).strip() if m else (meta_content(detail_html, property="og:title") or "").strip()
        if not title:
            return None

        duration_sec = _dur_to_sec(video.get("duration"))
        release_date = _iso_date(video.get("uploadDate") or video.get("datePublished"))
        thumbnail_url = video.get("thumbnailUrl") or meta_content(detail_html, property="og:image")
        if isinstance(thumbnail_url, list):
            thumbnail_url = thumbnail_url[0] if thumbnail_url else None

        # Performerzy: linki /models/<slug> (scene-specific; nav xvideos używa innego patternu).
        performers: list[RawPerformer] = []
        seen_perf: set[str] = set()
        for m in _MODEL_RE.finditer(detail_html):
            slug, name = m.group(1), m.group(2).strip()
            if not name or slug in seen_perf or name.lower() in ("models", "pornstars"):
                continue
            seen_perf.add(slug)
            performers.append(RawPerformer(external_id=f"{self.sitetag}:model:{slug}", name=name))
            if len(performers) >= 8:
                break

        # Tagi: /tags/<slug>.
        tags: list[RawTag] = []
        seen_tag: set[str] = set()
        for m in _TAG_RE.finditer(detail_html):
            slug = m.group(1)
            if slug in seen_tag or len(slug) > 60:
                continue
            seen_tag.add(slug)
            tags.append(RawTag(external_id=f"{self.sitetag}:tag:{slug}", name=slug.replace("-", " "), slug=slug))
            if len(tags) >= 15:
                break

        return RawScene(
            external_id=f"{self.sitetag}:{scene_url}",
            title=title,
            duration_sec=duration_sec,
            release_date=release_date,
            url=scene_url,
            performers=performers,
            tags=tags,
            playback_sources=[
                RawPlaybackSource(
                    origin=f"tube:{self.sitetag}",
                    page_url=scene_url,
                    duration_sec=duration_sec,
                    thumbnail_url=thumbnail_url,
                )
            ],
            raw={"source": "xvideos_browse"},
        )