# ==============================================================================
# Author: Simon-Pierre Boucher <contact@spboucher.ai>
# File:   src/main.py (ka-patreon)
# Desc:   Pages Patreon publiques — la page /<vanity> embarque l'objet campagne
#         (patron_count, creation_count, is_monthly, pay_per_name, résumé) +
#         og:title/og:image. Aucune clé. Proxy résidentiel recommandé.
# ==============================================================================
from __future__ import annotations

import asyncio
import html as htmllib
import json
import re

from apify import Actor

from .net import Fetcher

PAGE_URL = "https://www.patreon.com/{u}"
_NUM = {
    "patrons": re.compile(r'"patron_count"\s*:\s*(\d+)'),
    "posts": re.compile(r'"creation_count"\s*:\s*(\d+)'),
    "paid_posts": re.compile(r'"paid_member_count"\s*:\s*(\d+)'),
}
_OG_TITLE = re.compile(r'<meta property="og:title" content="([^"]*)"')
_OG_IMG = re.compile(r'<meta property="og:image" content="([^"]*)"')
_OG_DESC = re.compile(r'<meta property="og:description" content="([^"]*)"')
_MONTHLY = re.compile(r'"is_monthly"\s*:\s*(true|false)')
_NSFW = re.compile(r'"is_nsfw"\s*:\s*(true|false)')
_COVER = re.compile(r'"cover_photo_url"\s*:\s*"((?:[^"\\]|\\.)*)"')
_CREATION = re.compile(r'"creation_name"\s*:\s*"((?:[^"\\]|\\.)*)"')
# liens sociaux auto-déclarés de la campagne (objets campaign de l'app Patreon)
_SOCIALS = {
    "facebook": re.compile(r'"facebook"\s*:\s*"(https?:(?:[^"\\]|\\.)+)"'),
    "twitter": re.compile(r'"twitter"\s*:\s*"(https?:(?:[^"\\]|\\.)+)"'),
    "youtube": re.compile(r'"youtube"\s*:\s*"(https?:(?:[^"\\]|\\.)+)"'),
    "instagram": re.compile(r'"instagram"\s*:\s*"(https?:(?:[^"\\]|\\.)+)"'),
    "tiktok": re.compile(r'"tiktok"\s*:\s*"(https?:(?:[^"\\]|\\.)+)"'),
    "twitch": re.compile(r'"twitch"\s*:\s*"(https?:(?:[^"\\]|\\.)+)"'),
    "discord": re.compile(r'"discord"\s*:\s*"(https?:(?:[^"\\]|\\.)+)"'),
}
_PUBLISHED_RE = re.compile(r'"published_at"\s*:\s*"([^"]+)"')
_PPN_RE = re.compile(r'"pay_per_name"\s*:\s*"([^"]+)"')


def _dec(raw: str) -> str:
    try:
        return htmllib.unescape(json.loads(f'"{raw}"'))
    except Exception:
        return raw


async def main() -> None:
    async with Actor:
        inp = await Actor.get_input() or {}
        usernames = [u.strip().lstrip("@").lower()
                     for u in (inp.get("usernames") or []) if u and u.strip()]
        proxy = await Actor.create_proxy_configuration(
            actor_proxy_input=inp.get("proxyConfiguration"))
        fetcher = Fetcher(proxy, delay=float(inp.get("requestDelay") or 1))
        sem = asyncio.Semaphore(int(inp.get("concurrency") or 4))

        async def one(u: str) -> None:
            async with sem:
                miss = {"kind": "profile", "platform": "patreon",
                        "username": u, "found": False}
                try:
                    resp = await fetcher.get(PAGE_URL.format(u=u),
                                             session_id=u)
                except Exception as exc:
                    await Actor.push_data({**miss, "error": str(exc)[:200]})
                    return
                text = resp.text or ""
                if resp.status_code == 404:
                    await Actor.push_data({**miss, "error": "http_404"})
                    return
                m = _NUM["patrons"].search(text)
                if resp.status_code != 200 or not m:
                    await Actor.push_data(
                        {**miss, "error": f"http_{resp.status_code}"})
                    return

                def num(key):
                    mm = _NUM[key].search(text)
                    return int(mm.group(1)) if mm else None
                title = _OG_TITLE.search(text)
                img = _OG_IMG.search(text)
                desc = _OG_DESC.search(text)
                monthly = _MONTHLY.search(text)
                nsfw = _NSFW.search(text)
                cover = _COVER.search(text)
                creation = _CREATION.search(text)
                socials = {k: _dec(m.group(1))
                           for k, rx in _SOCIALS.items()
                           if (m := rx.search(text))}
                name = htmllib.unescape(title.group(1)) if title else u
                name = re.sub(r"\s*\|\s*Patreon\s*$", "", name).strip()
                await Actor.push_data({
                    "kind": "profile",
                    "platform": "patreon",
                    "found": True,
                    "username": u,
                    "full_name": name,
                    "biography": (htmllib.unescape(desc.group(1))
                                  if desc else None),
                    "followers": num("patrons"),   # patrons = « abonnés » KA
                    "patrons": num("patrons"),
                    "posts_count": num("posts"),
                    "is_monthly": (monthly.group(1) == "true"
                                   if monthly else None),
                    "is_nsfw": nsfw.group(1) == "true" if nsfw else None,
                    "published_at": (pub.group(1) if
                                     (pub := _PUBLISHED_RE.search(text))
                                     else None),
                    "pay_per_name": (ppn.group(1)[:40] if
                                     (ppn := _PPN_RE.search(text))
                                     else None),
                    "creation_name": (_dec(creation.group(1))[:200]
                                      if creation else None),
                    "avatar": htmllib.unescape(img.group(1)) if img else None,
                    "banner": _dec(cover.group(1)) if cover else None,
                    "social_links": socials or None,
                })

        await asyncio.gather(*[one(u) for u in usernames])
        Actor.log.info(f"terminé : {len(usernames)} pages")