goon/app/connectors/direct_scrapers/fullmovies.py

"""fullmovies.xxx — latest-vids browse scraper.

Identyczny engine co hdporn.gg (KVS sponsor_groups stack): `/videos/<slug>/`,
`/networks/<slug>/`, `/models/<slug>/`, `/tags/<slug>/`. og:image to `img.fullmovies.xxx/...`
— **prawdopodobnie auto-screenshot** (jak hdporn.gg → 8% match). Probe potwierdzi.
"""
from __future__ import annotations

import re

from app.connectors.base import (
    RawFingerprint,
    RawPerformer,
    RawPlaybackSource,
    RawScene,
    RawStudio,
    RawTag,
)
from app.connectors.direct_scrapers._browse_base import (
    BaseBrowseScraper,
    compute_thumbnail_phash,
    meta_content,
)
from app.normalize.text import slugify

_BASE = "https://www.fullmovies.xxx"
_SCENE_URL_RE = re.compile(r'href="(https://www\.fullmovies\.xxx/videos/[a-z0-9\-]+/)"', re.IGNORECASE)
_NETWORK_LINK_RE = re.compile(
    r'href="https://www\.fullmovies\.xxx/networks/([a-z0-9\-]+)/"[^>]*>([^<]+)', re.IGNORECASE
)
_MODEL_LINK_RE = re.compile(
    r'href="https://www\.fullmovies\.xxx/models/([a-z0-9\-]+)/"[^>]*>([^<]+)', re.IGNORECASE
)
_TAG_LINK_RE = re.compile(
    r'href="https://www\.fullmovies\.xxx/tags/([a-z0-9\-]+)/"[^>]*>([^<]+)', re.IGNORECASE
)


class FullmoviesScraper(BaseBrowseScraper):
    sitetag = "fullmoviesxxx"

    def _listing_url(self, page: int) -> str:
        if page <= 1:
            return f"{_BASE}/latest-updates/"
        return f"{_BASE}/latest-updates/{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 = 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:
        title = meta_content(detail_html, property="og:title")
        if not title:
            return None
        title = re.sub(r":\s*Free HD Porn\s*$|^Watch\s+|\s+Full XXX\s*$", "", title, flags=re.IGNORECASE).strip()

        description = meta_content(detail_html, property="og:description")
        thumbnail_url = meta_content(detail_html, property="og:image")

        duration_sec: int | None = None
        dur_meta = meta_content(detail_html, property="video:duration")
        if dur_meta and dur_meta.isdigit():
            duration_sec = int(dur_meta)

        # Studio z PREFIKSU tytułu ("Studio - Scene Title"), nie z sidebara /networks/.
        # Sidebar listuje WSZYSTKIE sieci → `_NETWORK_LINK_RE.finditer().first()` zawsze
        # zwracał pierwszą z listy (Brazzers) dla każdej sceny — mis-attribution. Tytuł
        # po oczyszczeniu ma format "Studio - Opis" (np. "Fake Hostel - ...").
        studio: RawStudio | None = None
        if " - " in title:
            studio_name = title.split(" - ", 1)[0].strip()
            if studio_name and len(studio_name) <= 50:
                studio = RawStudio(
                    external_id=f"fullmoviesxxx:studio:{slugify(studio_name)}",
                    name=studio_name,
                    slug=slugify(studio_name),
                )

        performers: list[RawPerformer] = []
        seen_perf: set[str] = set()
        for m in _MODEL_LINK_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 ("pornstars", "models"):
                continue
            seen_perf.add(slug)
            performers.append(
                RawPerformer(external_id=f"fullmoviesxxx:model:{slug}", name=name)
            )

        tags: list[RawTag] = []
        seen_tag: set[str] = set()
        for m in _TAG_LINK_RE.finditer(detail_html):
            slug, name = m.group(1), m.group(2).strip()
            if slug in seen_tag:
                continue
            seen_tag.add(slug)
            tags.append(RawTag(external_id=f"fullmoviesxxx:tag:{slug}", name=name, slug=slug))

        # Phash WYŁĄCZONY (pilot 2026-06-01: 0% trafień ≤5, mediana Hamming 14 do
        # canonical — auto-screenshoty img.fullmovies.xxx, nie hot-linkowane studio
        # thumbnaile). Matching trzyma się na title+performer+duration (seed: 92% tagged),
        # więc download thumbnaila pod phash to czysty narzut. thumbnail_url zostaje (display).
        fingerprints: list[RawFingerprint] = []

        playback_sources = [
            RawPlaybackSource(
                origin=f"tube:{self.sitetag}",
                page_url=scene_url,
                duration_sec=duration_sec,
                thumbnail_url=thumbnail_url,
            )
        ]

        return RawScene(
            external_id=f"{self.sitetag}:{scene_url}",
            title=title,
            description=description,
            duration_sec=duration_sec,
            url=scene_url,
            studio=studio,
            performers=performers,
            tags=tags,
            fingerprints=fingerprints,
            playback_sources=playback_sources,
        )