# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/hestia.py : connecteur Hestia Groupe immobilier (gestionhestia.com)
#   WordPress Elementor, CPT « apartments » non exposé en REST : la page
#   /location/ liste des cartes <article class="apartment-single"> (type,
#   secteur, prix/mois, photo, lien fiche). Les fiches détail (cache BD)
#   ajoutent titre, adresse civique complète, description, caractéristiques
#   (dont « Animaux interdits »), services de proximité et galerie.
# -----------------------------------------------------------------------------
from __future__ import annotations

import hashlib
import re

from bs4 import BeautifulSoup

from ..schema import Listing, normalize_unit_type, parse_price, strip_accents
from .base import BaseConnector

BASE = "https://www.gestionhestia.com"
LIST_URL = f"{BASE}/location/"

_BG_URL_RE = re.compile(r"url\((['\"]?)(https?://[^)'\"]+)\1\)")


def _pets_from_amenities(amenities: list[str]) -> str | None:
    """« Animaux interdits » / « Animaux acceptés » (caractéristique structurée)."""
    for a in amenities:
        k = strip_accents(a.lower())
        if "animaux" in k or "animal" in k:
            if "interdit" in k or "refus" in k:
                return "non"
            if "accept" in k or "permis" in k:
                return "oui"
    return None


class HestiaConnector(BaseConnector):
    source_id = "hestia"
    request_delay = 0.7
    max_details = 40    # garde-fou fiches détail (vraies requêtes)

    def fetch(self) -> list[Listing]:
        html = self.get(LIST_URL).text
        soup = BeautifulSoup(html, "html.parser")

        listings: dict[str, Listing] = {}
        for a in soup.select("ul.apartments a[href*='/apartments/']"):
            art = a.find("article", class_="apartment-single")
            if art is None:
                continue
            url = a["href"]
            m = re.search(r"/apartments/([^/]+)/?", url)
            if not m:
                continue
            ext_id = m.group(1)              # slug WP du CPT « apartments »
            if ext_id in listings:
                continue
            rooms_el = art.select_one(".apartment-single__rooms")
            rooms = rooms_el.get_text(" ", strip=True) if rooms_el else ""
            if re.search(r"commercial|stationnement|rangement|chambre",
                         rooms + " " + ext_id, re.I):
                continue
            sector_el = art.select_one(".apartment-single__sector")
            sector = sector_el.get_text(" ", strip=True) if sector_el else ""
            price_el = art.select_one(".apartment-single__price")
            price_label = re.sub(r"\s+", " ", price_el.get_text(" ", strip=True)) if price_el else ""

            images: list[str] = []
            img_el = art.select_one(".apartment-single__img[style]")
            if img_el:
                mm = _BG_URL_RE.search(img_el["style"])
                if mm:
                    images.append(mm.group(2))

            listings[ext_id] = Listing(
                source=self.source_id,
                external_id=ext_id,
                url=url,
                title=rooms,                       # remplacé par le titre de la fiche
                # tout le parc Hestia (Domaine Cartier, Boisé Nature 3R) est à
                # Trois-Rivières ; l'adresse de la fiche confirme/écrase
                city=sector or "Trois-Rivières",
                unit_type=normalize_unit_type(rooms),
                price=parse_price(price_label),
                price_label=price_label,
                images=images,
            )

        # Fiches détail (cache BD) : titre, adresse, description,
        # caractéristiques + services, galerie complète
        self._fetched = 0
        for lst in listings.values():
            card_key = hashlib.sha1(
                f"{lst.title}|{lst.price_label}|{lst.url}".encode("utf-8")
            ).hexdigest()
            try:
                payload = self.detail(lst.external_id, card_key,
                                      lambda u=lst.url: self._fetch_detail(u))
            except Exception:
                continue
            self._apply_detail(lst, payload)
        return list(listings.values())

    # -- fiche détail --------------------------------------------------------
    def _fetch_detail(self, url: str) -> dict:
        if self._fetched >= self.max_details:
            raise RuntimeError("budget de fiches détail atteint")
        self._fetched += 1
        soup = BeautifulSoup(self.get(url).text, "html.parser")
        out: dict = {}

        title_el = soup.select_one(".apartment__title")
        if title_el:
            out["title"] = re.sub(r"\s+", " ", title_el.get_text(" ", strip=True))
        addr_el = soup.select_one("address.apartment__address")
        if addr_el:
            out["address"] = re.sub(r"\s+", " ", addr_el.get_text(" ", strip=True))
        desc_el = soup.select_one(".apartment__description")
        if desc_el:
            out["description"] = desc_el.get_text("\n", strip=True)[:1200]
        # onglets « Caractéristiques » / « Services de proximité »
        out["amenities"] = [li.get_text(" ", strip=True)
                            for li in soup.select(".tabs .tabs__content li")
                            if li.get_text(strip=True)][:40]
        out["images"] = []
        for img in soup.select(".apartment__gallery img[src]"):
            src = img["src"]
            if src.startswith("http") and src not in out["images"]:
                out["images"].append(src)
        out["images"] = out["images"][:30]
        return out

    def _apply_detail(self, lst: Listing, d: dict) -> None:
        if not d:
            return
        if d.get("title"):
            lst.title = d["title"]
        addr = d.get("address", "")
        if addr:
            # « 6005 rue de la Mattawin, Trois-Rivières, QC G8Y 0M8, Canada »
            parts = [p.strip() for p in addr.split(",") if p.strip()]
            lst.address = parts[0] if parts else addr
            if len(parts) >= 2 and not re.match(r"(?i)qc|q[ué]ebec|canada|[A-Z]\d[A-Z]", parts[1]):
                lst.city = parts[1]
        if d.get("description"):
            lst.description = d["description"]
        if d.get("amenities"):
            lst.amenities = list(dict.fromkeys(lst.amenities + d["amenities"]))
            pets = _pets_from_amenities(d["amenities"])
            if pets:
                lst.pets = pets
        if d.get("images"):
            lst.images = list(dict.fromkeys(d["images"] + lst.images))[:30]