# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/souleymane.py : connecteur Gestion Souleymane (gestionsouleymane.com)
#   Gestionnaire local de Gatineau (secteurs est : Masson-Angers, Buckingham,
#   + Hull/Aylmer). WordPress + plugin immobilier ESTATIK : la page /a-louer
#   liste toutes les annonces (« 14 results », pas de pagination), chaque
#   carte `div.js-es-listing` portant data-post-id (external_id stable),
#   l'adresse civique en titre, le prix, chambres/salles de bain et la
#   galerie du carrousel (data-lazy). La fiche /property/<slug> (cache BD)
#   ajoute la description longue rédigée par l'agence — qui contient
#   « 📅 Disponible immédiatement », « secteur Masson-Angers », inclusions —
#   exploitée pour availability et le secteur, le reste par textmine.
# -----------------------------------------------------------------------------
from __future__ import annotations

import hashlib
import re

from bs4 import BeautifulSoup

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

BASE = "https://gestionsouleymane.com"
LIST_URL = f"{BASE}/a-louer"

_SIZE_SUFFIX = re.compile(r"-\d{2,4}x\d{2,4}(?=\.(?:jpg|jpeg|png|webp)$)", re.I)
# secteurs de Gatineau (fusion 2002) repérables dans l'adresse/description
_SECTORS = ["Masson-Angers", "Masson", "Buckingham", "Aylmer", "Hull",
            "Templeton", "Pointe-Gatineau", "Limbour", "Plateau"]


class SouleymaneConnector(BaseConnector):
    source_id = "souleymane"
    request_delay = 1.0
    max_details = 25
    max_images = 15

    # -- helpers ---------------------------------------------------------------
    @staticmethod
    def _sector(*texts: str) -> str:
        for txt in texts:
            for s in _SECTORS:
                if re.search(rf"(?i)\b{re.escape(s)}\b", txt or ""):
                    return "Masson-Angers" if s == "Masson" else s
        return ""

    # -- fiche détail ------------------------------------------------------------
    def _fetch_detail(self, url: str) -> dict:
        self._fetched += 1
        soup = BeautifulSoup(self.get(url).text, "html.parser")
        out: dict = {}
        desc_el = soup.select_one(".es-description, [itemprop='description']")
        # la description complète vit dans la section « Description » ; repli
        # sur le texte principal de la fiche
        block = None
        for h in soup.find_all(["h2", "h3", "h4"]):
            if h.get_text(strip=True).lower().startswith("description"):
                block = h.parent
                break
        el = block or desc_el
        if el:
            txt = el.get_text("\n", strip=True)
            txt = re.sub(r"^(?:Description\s*:?\s*\n?)+", "", txt)
            txt = re.sub(r"\n{2,}", "\n", txt)
            out["description"] = txt.strip()[:2500]
        return out

    # -- fetch -----------------------------------------------------------------
    def fetch(self) -> list[Listing]:
        soup = BeautifulSoup(self.get(LIST_URL).text, "html.parser")
        self._fetched = 0
        listings: dict[str, Listing] = {}
        for card in soup.select("div.js-es-listing"):
            try:
                self._parse_card(card, listings)
            except Exception:
                continue
        return list(listings.values())

    def _parse_card(self, card, listings: dict[str, Listing]) -> None:
        ext = str(card.get("data-post-id") or "")
        link = card.select_one("h3.es-listing__title a[href]")
        if not link:
            return
        url = link["href"]
        if not ext:
            m = re.search(r"/property/([^/]+)/?", url)
            ext = m.group(1) if m else ""
        if not ext or ext in listings:
            return

        address = re.sub(r"\s+", " ", link.get_text(" ", strip=True))
        price_el = card.select_one(".es-price")
        price_label = price_el.get_text(" ", strip=True) if price_el else ""

        excerpt_el = card.select_one("p.es-excerpt")
        excerpt = (re.sub(r"\s+", " ", excerpt_el.get_text(" ", strip=True))
                   if excerpt_el else "")

        beds = ""
        beds_el = card.select_one(".es-listing__meta-bedrooms b")
        if beds_el:
            beds = beds_el.get_text(strip=True)
        baths_el = card.select_one(".es-listing__meta-bathrooms b")
        baths = baths_el.get_text(strip=True) if baths_el else ""
        unit_type = (normalize_unit_type(f"{beds} chambres")
                     if beds.isdigit() else "")
        # les maisons restent des maisons, peu importe le compte de pièces
        if re.search(r"(?i)\bmaison\b", excerpt + " " + address):
            unit_type = "Maison"

        images: list[str] = []
        for img in card.select(".es-listing__image img"):
            u = img.get("data-lazy") or img.get("src") or ""
            if u.startswith("http"):
                u = _SIZE_SUFFIX.sub("", u)
                if u not in images:
                    images.append(u)

        # fiche détail : description complète (cache BD)
        payload: dict = {}
        key = hashlib.sha1(f"{address}|{price_label}|{excerpt}"
                           .encode("utf-8")).hexdigest()[:20]
        if self._fetched < self.max_details:
            try:
                payload = self.detail(ext, key,
                                      lambda u=url: self._fetch_detail(u))
            except Exception:
                payload = {}
        description = payload.get("description") or excerpt

        # non résidentiel : garages/entreposage/locaux annoncés sur la même page
        head = f"{address} {excerpt} {description[:200]}"
        if re.search(r"(?i)garage à louer|stationnement à louer|entreposage"
                     r"|local commercial", head):
            return

        # « 📅 Disponible immédiatement » / « Disponible le 1er septembre »
        availability = ""
        m = re.search(r"(?i)disponible[^\n.!]{0,50}", description)
        if m:
            availability = m.group(0).strip()

        details: dict = {}
        if baths.isdigit():
            details["bathrooms"] = int(baths)

        listings[ext] = Listing(
            source=self.source_id,
            external_id=ext,
            url=url,
            title=address,
            address=address,
            sector=self._sector(address, description),
            city="Gatineau",
            unit_type=unit_type,
            price=parse_price(price_label.replace(",", "")),
            price_label=price_label,
            availability=availability,
            description=description,
            details=details,
            images=images[: self.max_images],
        )