# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/construction_app.py : connecteur Construction APP / Les
#   Habitations APP (constructionappinc.com) — constructeur de Portneuf qui
#   loue ses immeubles à Pont-Rouge, Donnacona, Saint-Marc-des-Carrières,
#   Portneuf, Saint-Basile, Val-Bélair et Sainte-Catherine-de-la-J.-Cartier.
#   WordPress/Divi rendu serveur : /vente-et-location/ empile des blocs
#   `.et_pb_row` — h1 « Ville (Québec) », libellé « LOGEMENT À LOUER »/
#   « MAISONS DE VILLE EN LOCATION », adresse, caractéristiques (chambres,
#   sdb, types n ½) et photos. Ni prix ni date (Tier B).
#   Granularité : immeuble (un bloc = une annonce).
# -----------------------------------------------------------------------------
from __future__ import annotations

import re

from bs4 import BeautifulSoup

from ..schema import Listing, strip_accents
from .base import BaseConnector

BASE = "https://www.constructionappinc.com"
LIST_URL = f"{BASE}/vente-et-location/"

LABEL_RE = re.compile(r"(LOGEMENTS?|MAISONS? DE VILLE)"
                      r".{0,40}(À LOUER|EN LOCATION)", re.I | re.S)
CITY_RE = re.compile(r"^(.+?)\s*\(Qu[ée]bec\)$", re.I)
ADDR_RE = re.compile(r"^\d[\d\s\wàâéèêëîïôùûç.'-]*,\s*.+", re.I)
ADDR2_RE = re.compile(r"^(rue|avenue|av\.|boul)", re.I)
BED_RE = re.compile(r"Chambre\(?s?\)?\s*:\s*(\d+)", re.I)
BATH_RE = re.compile(r"Salle\(?s?\)?\s*de\s*bain\s*:\s*(\d+)", re.I)
TYPE_RE = re.compile(r"\b(\d)\s*1/2\b")

CONTACT = {"email": "info@habitationsapp.com", "phone": "418-410-0715"}


def _slug(text: str) -> str:
    return re.sub(r"-+", "-",
                  re.sub(r"[^a-z0-9]+", "-", strip_accents(text.lower()))
                  ).strip("-")


class ConstructionAppConnector(BaseConnector):
    source_id = "construction_app"
    request_delay = 0.6

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

        seen: set[str] = set()
        for row in soup.select("div.et_pb_row"):
            try:
                lst = self._parse_row(row, seen)
            except Exception:
                continue
            if lst is not None:
                listings.append(lst)
        return listings

    def _parse_row(self, row, seen: set[str]) -> Listing | None:
        txt = row.get_text("\n", strip=True)
        if not LABEL_RE.search(txt):
            return None
        lines = [ln.strip() for ln in txt.split("\n") if ln.strip()]

        city = ""
        for ln in lines:
            m = CITY_RE.match(ln)
            if m:
                city = m.group(1).strip()
                break

        address = next(
            (ln for ln in lines
             if ADDR_RE.match(ln) and not ln.lower().startswith(("t 4", "418"))
             and "1/2" not in ln), "")
        if not address:
            address = next((ln for ln in lines
                            if ADDR2_RE.match(ln) and "," in ln), "")
        if not address:
            return None                # bloc générique sans adresse
        if not city:                   # ville = dernier segment de l'adresse
            tail = address.split(",")[-1].strip()
            if tail and not re.match(r"^suite", tail, re.I):
                city = tail

        ext_id = _slug(address)
        if not ext_id or ext_id in seen:
            return None
        seen.add(ext_id)

        blob = " | ".join(lines)
        bedrooms = bathrooms = None
        m = BED_RE.search(blob)
        if m:
            bedrooms = float(m.group(1))
        m = BATH_RE.search(blob)
        if m:
            bathrooms = float(m.group(1))
        types = list(dict.fromkeys(f"{n}½" for n in TYPE_RE.findall(blob)))
        unit_type = types[0] if len(types) == 1 else ""

        # commodités : lignes courtes ni adresse/ville/label/contacts
        amenities = [ln for ln in lines
                     if 3 < len(ln) <= 60
                     and ln not in (address, city)
                     and not CITY_RE.match(ln)
                     and not LABEL_RE.search(ln)
                     and not re.search(r"(chambre|salle|1/2|@|\d{3}[ -]\d{4}"
                                       r"|nouveau projet|contactez|photos?\b"
                                       r"|plans? des|en savoir)",
                                       ln, re.I)][:12]

        images = [img.get("src", "") for img in row.find_all("img")
                  if img.get("src", "").startswith("http")
                  and "icon" not in img.get("src", "").lower()
                  and "logo" not in img.get("src", "").lower()][:10]

        label = "MAISONS DE VILLE" if re.search(r"maisons? de ville",
                                                blob, re.I) else "Logement"
        return Listing(
            source=self.source_id,
            external_id=ext_id,
            url=LIST_URL,
            title=f"{address}",
            address=f"{address}" + (f", {city}" if city
                                    and city.lower() not in address.lower()
                                    else ""),
            city=city,
            unit_type=unit_type,
            bedrooms=bedrooms,
            bathrooms=bathrooms,
            price_label="Prix sur demande",
            availability="Disponible",
            description=f"{label} à louer au {address} ({city}) — "
                        "Les Habitations APP (Construction APP). "
                        + (f"Types offerts : {', '.join(types)}. "
                           if types else "") + " ".join(amenities[:6]),
            amenities=amenities,
            details={"contact": dict(CONTACT),
                     **({"types": types} if len(types) > 1 else {})},
            images=images,
        )