# -----------------------------------------------------------------------------
# Immo-Ka — Agrégateur de maisons à vendre (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/duquesimms.py : Duque Simms (duquesimms.com) — agence indépendante
#   de Montréal (NDG, Côte-des-Neiges, Ouest-de-l'Île), ~25 inscriptions.
#
#   Site WordPress/Elementor (plateforme proprius/e-mmobilier, photos sur
#   cdn.proprius.e-mmobilier.ca). Découverte par property-sitemap.xml : URLs
#   fr /property/{n° Centris}/ (les /en/ sont des doublons de langue).
#
#   TOUT vient de la fiche détail (du.enrich) : JSON-LD RealEstateListing
#   (description, contentLocation adresse complète + GPS, offers[].price avec
#   businessFunction Sell, keywords = type, seller = courtier nom/téléphone)
#   + blocs HTML property-characteristic (libellé/valeurs), property-expense
#   (taxes, énergie), property-evaluation, compteurs chambres/salles de bain
#   (icônes bedroom.svg/bathroom.svg suivies du nombre), galerie swiper
#   cdn …-pi.jpg (les -c.jpg sont des portraits). Les locations (prix de
#   loyer) sont écartées par le seuil de prix vente.
#
#   Agence première main (pas d'infixe _ag_) : ses fiches battent les copies
#   portails au même n° Centris dans la dédup.
# -----------------------------------------------------------------------------
from __future__ import annotations

import html as _html
import os
import re

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

SITE = "https://duquesimms.com"
SITEMAP = SITE + "/property-sitemap.xml"
AGENCY = "Duque Simms"
DETAIL_LIMIT = int(os.environ.get("IMMOKA_DUQUESIMMS_DETAIL_LIMIT", "80"))
MIN_SALE_PRICE = 20000     # sous ce prix : loyer mensuel (location) → écarté

_LOC_RE = re.compile(r"<loc>\s*(https://duquesimms\.com/property/(\d{6,10})/?)\s*</loc>")
_ADDR_RE = re.compile(r"^(.*?),\s*(.+?)\s*([A-Z]\d[A-Z]\s?\d[A-Z]\d)?$")
_CHAR_RE = re.compile(r'property-characteristic-type">(.*?)</span>(.*?)</div>', re.S)
_SUBVAL_RE = re.compile(r'property-subcharacteristic-value">(.*?)</span>', re.S)
_EXP_RE = re.compile(r'property-expense-label">(.*?)</span>.*?property-money-amount">'
                     r'([\d\s ,]+)</span>', re.S)
_EVAL_RE = re.compile(r'property-evaluation-label">(.*?)</span>.*?property-money-amount">'
                      r'([\d\s ,]+)</span>', re.S)
_BEDS_RE = re.compile(r'bedroom\.svg.{0,900}?elementor-widget-container">\s*(\d+)\s*<', re.S)
_BATHS_RE = re.compile(r'bathroom\.svg.{0,900}?elementor-widget-container">\s*(\d+)\s*<', re.S)
_IMG_RE = re.compile(r'https://cdn\.proprius\.e-mmobilier\.ca/([A-Z0-9]+-pi)\.jpe?g', re.I)


class DuqueSimmsConnector(BaseConnector):
    source_id = "duquesimms"
    request_delay = 0.4

    def fetch(self) -> list[PropertyListing]:
        # SiteGround sert un challenge sgcaptcha en HTTP 202 (sans exception) :
        # get_resilient escalade (Oxylabs → Scrapfly ASP → Bright Data)
        xml = self.get_resilient(SITEMAP).text
        by_id: dict[str, PropertyListing] = {}
        for url, mls in _LOC_RE.findall(xml):
            by_id.setdefault(mls, PropertyListing(
                source=self.source_id, external_id=mls,
                url=url if url.endswith("/") else url + "/",
                title="", mls=mls, agency=AGENCY, broker_name=AGENCY,
            ))
        listings = list(by_id.values())
        du.enrich(self, listings, DETAIL_LIMIT, _parse_detail, key="v1",
                  fetch_html=lambda u: self.get_resilient(u).text)
        out = []
        for lst in listings:
            if not lst.price or lst.price < MIN_SALE_PRICE:
                continue               # location (loyer) ou fiche vide
            if not lst.title:
                lst.title = f"{lst.property_type or 'Propriété'} à vendre — " \
                            f"{lst.address or lst.city}".strip(" —")
            if lst.price and not lst.price_label:
                lst.price_label = f"{int(lst.price):,} $".replace(",", " ")
            out.append(lst)
        return out


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


def _parse_detail(html: str) -> dict:
    node = None
    for n in du.ld_nodes(html):
        if n.get("@type") == "RealEstateListing":
            node = n
            break
    if node is None:
        return {}
    out: dict = {}
    details: dict = {}
    features: list[str] = []

    desc = _html.unescape(str(node.get("description") or "")).strip()
    if desc:
        out["description"] = desc

    kw = node.get("keywords") or []
    if isinstance(kw, str):
        kw = [kw]
    if kw:
        out["property_type"] = str(kw[0]).strip()
        details["Type de propriété"] = out["property_type"]
        if len(kw) > 1:
            details["Catégorie"] = str(kw[1]).strip()

    loc = node.get("contentLocation") or {}
    if isinstance(loc, dict):
        full = _html.unescape(str(loc.get("address") or "")).strip()
        m = _ADDR_RE.match(full)
        if m:
            out["address"] = m.group(1).strip()
            out["city"] = m.group(2).strip()
            if m.group(3):
                details["Code postal"] = m.group(3)
        elif full:
            out["address"] = full
        try:
            out["lat"], out["lng"] = float(loc["latitude"]), float(loc["longitude"])
        except (KeyError, TypeError, ValueError):
            pass

    offers = node.get("offers") or []
    if isinstance(offers, dict):
        offers = [offers]
    sellers: list[str] = []
    for off in offers:
        if not isinstance(off, dict):
            continue
        if "Sell" not in str(off.get("businessFunction") or "Sell"):
            continue
        if out.get("price") is None and off.get("price"):
            try:
                out["price"] = float(str(off["price"]).replace(" ", ""))
            except ValueError:
                pass
        seller = off.get("seller") or {}
        if isinstance(seller, dict) and seller.get("name"):
            name = str(seller["name"]).strip()
            if name not in sellers:
                sellers.append(name)
            if "broker_phone" not in out:
                tel = str(seller.get("telephone") or "").strip()
                if re.fullmatch(r"\d{10}", tel):
                    out["broker_phone"] = f"{tel[:3]} {tel[3:6]}-{tel[6:]}"
    if sellers:
        out["broker_name"] = " et ".join(sellers[:2])

    bm, tm = _BEDS_RE.search(html), _BATHS_RE.search(html)
    if bm:
        out["bedrooms"] = int(bm.group(1))
    if tm:
        out["bathrooms"] = int(tm.group(1))

    for label, blob in _CHAR_RE.findall(html):
        label = _txt(label)
        vals = [_txt(v) for v in _SUBVAL_RE.findall(blob)]
        vals = [v for v in vals if v]
        if label and vals:
            details[label] = ", ".join(vals)
            features.append(f"{label} : {details[label]}"
                            if len(details[label]) < 40 else label)

    for rx in (_EXP_RE, _EVAL_RE):
        for label, amount in rx.findall(html):
            label, amount = _txt(label), _txt(amount).replace(" ", " ")
            if label and amount:
                details[label] = f"{amount} $"

    imgs, seen = [], set()
    for base in _IMG_RE.findall(html):
        if base not in seen:
            seen.add(base)
            imgs.append(f"https://cdn.proprius.e-mmobilier.ca/{base}.jpg")
    if imgs:
        out["images"] = imgs

    if details:
        out["details"] = details
    if features:
        out["features"] = features
    return out