# -----------------------------------------------------------------------------
# Immo-Ka — Agrégateur de maisons à vendre (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/kw_urbain.py : Keller Williams Urbain (kwurbain.ca)
#   Page /inscriptions rendue serveur : toutes les inscriptions de l'agence en
#   cartes HTML (plateforme marketingwebsites.ca, commune aux bureaux KW Canada).
# -----------------------------------------------------------------------------
from __future__ import annotations

import html as _html
import os
import re

from .base import BaseConnector
from . import _detailutil as du
from ..normalize import parse_price
from ..schema import PropertyListing

BASE = "https://www.kwurbain.ca"
LISTINGS = f"{BASE}/inscriptions/"
IMG_TMPL = "https://realestate.marketingwebsites.ca/property-images/{id}/{id}-{n:02d}.jpg"
DETAIL_LIMIT = int(os.environ.get("IMMOKA_KW_DETAIL_LIMIT", "400"))
_KW_IMG_RE = re.compile(
    r'https://realestate\.marketingwebsites\.ca/property-images/\d+/[^"\'\\ ]+\.(?:jpg|jpeg|png|webp)',
    re.I)
_META_DESC_RE = re.compile(r'<meta\s+name="description"\s+content="([^"]{40,})"', re.I)

_CARD_RE = re.compile(r'property-card mix property-(\d+)"(.*?)(?=property-card mix property-\d+"|</div>\s*</div>\s*</div>\s*</section>|$)', re.S)
_H5_RE = re.compile(r'<h5>(.*?)</h5>', re.S)
_PRICE_RE = re.compile(r'<span>(.*?)</span>', re.S)
_SUB_RE = re.compile(r'card-subtitle[^>]*>(.*?)</h6>', re.S)
_BED_RE = re.compile(r'fa-bed"></i>\s*(\d+)')
_BATH_RE = re.compile(r'fa-bath"></i>\s*(\d+)')
_STATUS_RE = re.compile(r'card-status[^>]*>(.*?)</div>', re.S)


def _txt(s: str) -> str:
    return _html.unescape(re.sub(r"<[^>]+>", "", s or "")).strip()


class KwUrbainConnector(BaseConnector):
    source_id = "kw_urbain"
    request_delay = 0.6

    def fetch(self) -> list[PropertyListing]:
        html = self.get(LISTINGS).text
        listings: list[PropertyListing] = []
        for pid, block in _CARD_RE.findall(html):
            lst = self._to_listing(pid, block)
            if lst is not None:
                listings.append(lst)
        # fiche détail : galerie complète + description
        du.enrich(self, listings, DETAIL_LIMIT, parse_kw_detail, key="v2")
        return listings

    def _to_listing(self, pid: str, block: str) -> PropertyListing | None:
        addr = _txt((_H5_RE.search(block) or [None, ""])[1] if _H5_RE.search(block) else "")
        m_addr = _H5_RE.search(block)
        addr = _txt(m_addr.group(1)) if m_addr else ""
        m_price = _PRICE_RE.search(block)
        price_label = _txt(m_price.group(1)) if m_price else ""
        m_sub = _SUB_RE.search(block)
        sub = _txt(m_sub.group(1)) if m_sub else ""     # "Blainville, Laurentides J7B1M1"
        city = region = ""
        if sub:
            parts = [p.strip() for p in sub.split(",")]
            city = parts[0]
            if len(parts) > 1:
                # « Laurentides J7B1M1 » -> retirer le code postal
                region = re.sub(r"\s*[A-Z]\d[A-Z]\s?\d[A-Z]\d\s*$", "", parts[1]).strip()
        beds = _BED_RE.search(block)
        baths = _BATH_RE.search(block)
        return PropertyListing(
            source=self.source_id,
            external_id=pid,
            url=f"{BASE}/inscriptions/inscription/{pid}",
            title=addr,
            address=addr,
            city=city,
            region=region,
            price=parse_price(price_label),
            price_label=price_label,
            bedrooms=int(beds.group(1)) if beds else None,
            bathrooms=int(baths.group(1)) if baths else None,
            images=[IMG_TMPL.format(id=pid, n=1)],
            broker_name="Keller Williams Urbain",
        )


def parse_kw_detail(html: str) -> dict:
    """Galerie complète + description (balise meta) de la fiche KW."""
    out: dict = {}
    seen, imgs = set(), []
    for u in _KW_IMG_RE.findall(html):
        if u not in seen:
            seen.add(u)
            imgs.append(u)
    if imgs:
        out["images"] = imgs
    m = _META_DESC_RE.search(html)
    if m:
        d = _html.unescape(m.group(1)).strip()
        if d and not d.lower().startswith(("keller williams", "trouvez")):
            out["description"] = d
    _det = du.centris_details(du.flatten(html))
    if _det:
        out.setdefault("details", {}).update(_det)
    return out