# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/picard.py : connecteur Picard Immobilier (picardimmobilier.net)
#   La page /recherche/ liste tous les logements disponibles (rendu serveur),
#   cartes .property-box-2 avec lien /logement/<secteur>/ref=<id>/.
#   Fiches détaillées : photos S3, disponibilité, description, commodités.
# -----------------------------------------------------------------------------
from __future__ import annotations

import hashlib
import re

from bs4 import BeautifulSoup

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

BASE = "https://picardimmobilier.net"
SEARCH_URL = f"{BASE}/recherche/"

SECTOR_LABELS = {
    "charny": "Charny",
    "les-saules": "Les Saules",
    "sainte-foy": "Sainte-Foy",
    "beauport": "Beauport",
    "limoilou": "Limoilou",
    "loretteville": "Loretteville",
    "vieux-quebec": "Vieux-Québec",
    "montcalm": "Montcalm",
}
IMG_RE = re.compile(r"https://picardimmobilier\.s3\.amazonaws\.com/media/"
                    r'[^"\s\\)]+\.(?:jpe?g|png|webp)', re.I)
LATLNG_RE = re.compile(r'data-lat="(-?[\d.]+)"\s+data-lng="(-?[\d.]+)"')


class _DetailBudget(Exception):
    """Plafond de requêtes de fiches atteint pour cette synchronisation."""


class PicardConnector(BaseConnector):
    source_id = "picard"
    request_delay = 0.6
    max_details = 80        # garde-fou de fetch des fiches

    def fetch(self) -> list[Listing]:
        listings: dict[str, Listing] = {}
        try:
            html = self.get(SEARCH_URL).text
        except Exception:
            return []
        soup = BeautifulSoup(html, "html.parser")

        for card in soup.select("div.property-box-2"):
            try:
                a = card.select_one('a[href*="ref="]')
                if not a:
                    continue
                m = re.search(r"/logement/([a-z\-]+)/ref=(\d+)/", a.get("href", ""))
                if not m:
                    continue
                sector_slug, ext_id = m.group(1), m.group(2)
                if ext_id in listings:
                    continue
                sector = SECTOR_LABELS.get(sector_slug,
                                           sector_slug.replace("-", " ").title())
                title_el = card.select_one("h3.title a")
                title = title_el.get_text(" ", strip=True) if title_el else ""
                # Exclure commercial / stationnement / rangement
                if re.search(r"Commercial|Stationnement|Rangement|Parking",
                             title, re.I):
                    continue
                loc_el = card.select_one("h5.location a")
                address = loc_el.get_text(" ", strip=True) if loc_el else ""
                price_el = card.select_one(".price-box")
                price_label = (re.sub(r"\s+", " ", price_el.get_text(" ", strip=True))
                               if price_el else "")
                # ex. "$1055.00 Par mois" -> normaliser pour parse_price
                pm = re.search(r"\$\s*([\d\s,\.]+)", price_label)
                price = None
                if pm:
                    try:
                        price = float(pm.group(1).replace(",", "").replace(" ", ""))
                        if not (100 <= price <= 20000):
                            price = None
                    except ValueError:
                        price = None
                unit_type = ""
                for li in card.select("ul.facilities-list li"):
                    t = li.get_text(" ", strip=True)
                    if "Chambre" in t:
                        unit_type = t.replace("Chambres", "").strip()
                        break
                img = card.select_one(".property-photo img")
                images = [img["src"]] if img and img.get("src") else []
                listings[ext_id] = Listing(
                    source=self.source_id,
                    external_id=ext_id,
                    url=f"{BASE}/logement/{sector_slug}/ref={ext_id}/",
                    title=title or (address or f"Logement {ext_id}"),
                    address=address,
                    sector=sector,
                    city=infer_city(sector),
                    unit_type=normalize_unit_type(unit_type),
                    price=price,
                    price_label=price_label,
                    images=images,
                )
            except Exception:
                continue

        # Fiches détaillées (via cache self.detail : une vraie requête
        # seulement si la carte liste a changé) : photos, disponibilité,
        # adresse complète, description, commodités, lat/lng.
        budget = {"n": 0}
        for lst in listings.values():
            key = hashlib.sha1("|".join(
                (lst.title, lst.price_label, lst.address, lst.unit_type,
                 lst.images[0] if lst.images else "")).encode()).hexdigest()

            def _fetch(url=lst.url):
                if budget["n"] >= self.max_details:
                    raise _DetailBudget(url)
                budget["n"] += 1
                return self._parse_detail(self.get(url).text)

            try:
                payload = self.detail(lst.external_id, key, _fetch)
            except Exception:
                continue
            if payload.get("images"):
                lst.images = payload["images"][:25]
            if payload.get("availability"):
                lst.availability = payload["availability"]
            if payload.get("address"):
                lst.address = payload["address"]
            if payload.get("description"):
                lst.description = payload["description"]
            if payload.get("amenities"):
                lst.amenities = payload["amenities"]
            if payload.get("lat") is not None:
                lst.lat, lst.lng = payload["lat"], payload["lng"]

        return list(listings.values())

    # -- fiche /logement/<secteur>/ref=<id>/ ----------------------------------
    @staticmethod
    def _parse_detail(detail: str) -> dict:
        payload: dict = {}
        imgs = [u for u in dict.fromkeys(IMG_RE.findall(detail))
                if not re.search(r"logo|icon|favicon|static", u, re.I)]
        if imgs:
            payload["images"] = imgs
        dsoup = BeautifulSoup(detail, "html.parser")
        text = re.sub(r"\s+", " ", dsoup.get_text(" ", strip=True))
        # Adresse complète + disponibilité en tête de fiche
        am = re.search(r"(Disponible[^#]{0,40}?)\s*#?\s*Référence", text)
        if am:
            payload["availability"] = am.group(1).strip()
        adm = re.search(
            r"par mois\s+(.{5,90}?)\s+(?:Disponible|#\s*Référence)", text)
        if adm:
            payload["address"] = adm.group(1).strip()
        dm = re.search(r"Description\s+(.{10,600}?)\s+Caractéristiques", text)
        if dm:
            payload["description"] = dm.group(1).strip()
        amenities = [re.sub(r"\s+", " ", d.get_text(" ", strip=True))
                     for d in dsoup.select("div.col-6.text-dark")
                     if d.get_text(strip=True)][:15]
        if amenities:
            payload["amenities"] = amenities
        # Coordonnées : carte Google `#map-custom` (attributs data-lat/lng)
        mm = LATLNG_RE.search(detail)
        if mm:
            try:
                payload["lat"] = float(mm.group(1))
                payload["lng"] = float(mm.group(2))
            except ValueError:
                pass
        return payload