goon/app/connectors/direct_scrapers/hdporngg.py

"""hdporn.gg, latest-vids browse scraper.

Engine podobny do freshporno: `/videos/<slug>/` URL, `/networks/<slug>/` = studio,
`/models/<slug>/` = performer, `/tags/<slug>/` = tag.

Quirk: og:image to internal CDN `img.hdporn.gg/...`, przed merging do prod
sprawdzamy phash distance (gate-keeper: jeśli Hamming >5 dla >70% scen → orphan
factory, wyłącz; analogia do shyfap).
"""
from __future__ import annotations

import re
from datetime import date
from urllib.parse import urljoin

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.hdporn.gg"
_SCENE_URL_RE = re.compile(r'href="(https://www\.hdporn\.gg/videos/[a-z0-9\-]+/)"', re.IGNORECASE)
_NETWORK_LINK_RE = re.compile(
    r'href="https://www\.hdporn\.gg/networks/([a-z0-9\-]+)/"[^>]*>([^<]+)', re.IGNORECASE
)
_UPLOAD_DATE_RE = re.compile(r'"uploadDate"\s*:\s*"(\d{4})-(\d{2})-(\d{2})')
_MODEL_LINK_RE = re.compile(
    r'href="https://www\.hdporn\.gg/models/([a-z0-9\-]+)/"[^>]*>([^<]+)', re.IGNORECASE
)
_TAG_LINK_RE = re.compile(
    r'href="https://www\.hdporn\.gg/tags/([a-z0-9\-]+)/"[^>]*>([^<]+)', re.IGNORECASE
)


class HDPornGGScraper(BaseBrowseScraper):
    sitetag = "hdporngg"

    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
        # og:title typowo zawiera ": Free HD Porn" suffix, usuń
        title = re.sub(r":\s*Free HD Porn\s*$", "", title, flags=re.IGNORECASE).strip()
        # I "Brazzers - " prefix często też w title, zostaw, bo studio name w title
        # to silny sygnał dla fuzzy match.

        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)
        # Data z JSON-LD `"uploadDate": "2026-10-13T16:51:32Z"` (audyt 2026-10-14: 0 dat).
        release_date = None
        um = _UPLOAD_DATE_RE.search(detail_html)
        if um:
            try:
                release_date = date(int(um.group(1)), int(um.group(2)), int(um.group(3)))
            except ValueError:
                release_date = None

        # Studio z PREFIKSU tytułu ("Studio - Scene Title"), nie z sidebara /networks/.
        # Sidebar listuje WSZYSTKIE sieci → pierwszy match zawsze ten sam (Brazzers) dla
        # każdej sceny. Tytuł po oczyszczeniu ma format "Studio - Opis" (np. "Dad Crush - ...").
        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"hdporngg: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"hdporngg: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"hdporngg:tag:{slug}", name=name, slug=slug)
            )

        # Phash WYŁĄCZONY (pilot 2026-06-01: 0% trafień ≤5, mediana Hamming 14 do
        # canonical, auto-screenshoty img.hdporn.gg, 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,
            release_date=release_date,
            duration_sec=duration_sec,
            url=scene_url,
            studio=studio,
            performers=performers,
            tags=tags,
            fingerprints=fingerprints,
            playback_sources=playback_sources,
        )