# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/ledomaine.py : connecteur Les Habitations Le Domaine
#   (ledomaine.ca — grand ensemble locatif du quartier Mercier,
#    arrondissement Mercier–Hochelaga-Maisonneuve, Montréal).
#   Site WordPress rendu serveur : une page par typologie
#   (/appartement/appartement-3-et-demi/, etc.) avec prix « à partir de »,
#   description, galerie photos et plan. Une annonce par typologie.
#   Les alias (ex. /appartement-2/) sont dédupliqués via <link rel=canonical>.
# -----------------------------------------------------------------------------
from __future__ import annotations

import re

from bs4 import BeautifulSoup

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

BASE = "https://www.ledomaine.ca"

ADDRESS = "2990, Avenue de Granby, Montréal"
SECTOR = "Mercier"
CITY = "Montréal"

_APT_LINK_RE = re.compile(r"https?://www\.ledomaine\.ca/appartement/([a-z0-9-]+)/?")
_IMG_HREF_RE = re.compile(
    r'href="(https://www\.ledomaine\.ca/wp-content/uploads/[^"]+\.(?:jpg|jpeg|png|webp))"',
    re.I)
_PRICE_RE = re.compile(r"[àa] partir de\s*([\d\s ]+)\s*\$", re.I)


class LeDomaineConnector(BaseConnector):
    source_id = "ledomaine"
    request_delay = 0.6
    max_pages = 12               # garde-fou de crawl

    def fetch(self) -> list[Listing]:
        listings: dict[str, Listing] = {}
        try:
            home = self.get(BASE + "/").text
        except Exception:
            return []

        slugs = list(dict.fromkeys(_APT_LINK_RE.findall(home)))
        for slug in slugs[: self.max_pages]:
            try:
                lst = self._parse_page(slug)
            except Exception:
                continue
            if lst and lst.external_id not in listings:
                listings[lst.external_id] = lst
        return list(listings.values())

    def _parse_page(self, slug: str) -> Listing | None:
        url = f"{BASE}/appartement/{slug}/"
        html = self.get(url).text
        soup = BeautifulSoup(html, "html.parser")

        # Déduplication des alias (/appartement-2/ -> /appartement-3-et-demi/)
        canon = soup.find("link", rel="canonical")
        if canon and canon.get("href"):
            m = _APT_LINK_RE.search(canon["href"])
            if m:
                slug = m.group(1)
                url = f"{BASE}/appartement/{slug}/"

        h1 = soup.find("h1")
        title = h1.get_text(" ", strip=True) if h1 else slug.replace("-", " ")

        # « Appartement 4 et demi sous-sol » -> 4½ (mention conservée au titre)
        tm = re.search(r"(\d)\s*et\s*demi", title, re.I)
        unit_type = f"{tm.group(1)}½" if tm else normalize_unit_type(title)

        # Prix « à partir de NNN $ par mois » (meta description ou corps)
        price = None
        price_label = ""
        meta = soup.find("meta", attrs={"name": "description"})
        sources = [meta.get("content", "") if meta else "",
                   soup.get_text(" ", strip=True)]
        for txt in sources:
            m = _PRICE_RE.search(txt or "")
            if m:
                amount = re.sub(r"[\s ]", "", m.group(1))
                price_label = f"À partir de {amount} $/mois"
                price = parse_price(f"{amount}$")
                break

        # Description : bloc .text-wrapper sous le h2 « Description de
        # l'appartement » (contient inclusions, balcon, politique animaux…)
        description = ""
        for h2 in soup.find_all("h2"):
            if "description" in h2.get_text(" ", strip=True).lower():
                wrapper = h2.find_parent(class_="text-wrapper") or h2.parent
                parts = []
                for p in wrapper.find_all("p"):
                    t = " ".join(p.get_text(" ", strip=True).split())
                    if t and not t.lower().startswith("consulter"):
                        parts.append(t)
                description = " ".join(parts).strip()[:600]
                break
        if not description:   # repli : ancien découpage textuel
            body_txt = soup.get_text("|", strip=True)
            dm = re.search(
                r"Description\|de l'appartement\|(.{20,900}?)\|Consulter",
                body_txt, re.S)
            if dm:
                description = re.sub(r"\s*\|\s*", " ", dm.group(1))
                description = re.sub(r"\s+", " ", description).strip()[:600]

        # Toutes les images (galerie + plan) de la page
        images = [u for u in dict.fromkeys(_IMG_HREF_RE.findall(html))
                  if not re.search(r"logo|icon|favicon", u, re.I)]
        if not images:
            return None

        return Listing(
            source=self.source_id,
            external_id=slug,
            url=url,
            title=f"{title} — Les Habitations Le Domaine",
            address=ADDRESS,
            sector=SECTOR,
            city=CITY,
            unit_type=unit_type,
            price=price,
            price_label=price_label,
            availability="",
            description=description,
            amenities=[],
            images=images,
        )