# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/axia.py : connecteur AXIA Appartements (axiaappartements.com)
#   Complexe locatif neuf à Lachine (Montréal), géré par Pur Immobilia.
#   Site vitrine WordPress/WPBakery une page : deux configurations
#   (2 chambres 4½ et 3 chambres 5½) avec superficie et prix
#   « à partir de » — 1 annonce par configuration.
# -----------------------------------------------------------------------------
from __future__ import annotations

import re

from bs4 import BeautifulSoup

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

BASE = "https://www.axiaappartements.com"

ADDRESS = "200, boul. Saint-Joseph, Lachine (Québec) H8S 2L3"

IMG_RE = re.compile(
    r"https://www\.axiaappartements\.com/wp-content/uploads/"
    r"[^\"\s\\]+?\.(?:jpg|jpeg|webp)", re.I)


class AxiaConnector(BaseConnector):
    source_id = "axia"
    request_delay = 0.6

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

        # Photos du complexe (icônes/logos exclus)
        images = [u for u in dict.fromkeys(IMG_RE.findall(html))
                  if not re.search(r"icone|logo|favicon|fleche|-\d+x\d+\.",
                                   u, re.I)][:25]

        # Description (meta + paragraphe « APPARTEMENTS » sur les unités)
        desc = ""
        og = soup.find("meta", attrs={"property": "og:description"}) or \
            soup.find("meta", attrs={"name": "description"})
        if og and og.get("content"):
            desc = og["content"].strip()
        for p in soup.find_all("p"):
            txt = re.sub(r"\s+", " ", p.get_text(" ", strip=True))
            if "électroménagers" in txt:      # descriptif des unités
                desc = (desc + " " + txt).strip()
                break
        desc = desc[:600]
        amenities = []
        for el in soup.select(".icon-label"):
            txt = re.sub(r"\s+", " ", el.get_text(" ", strip=True)).strip()
            if txt and txt not in amenities:
                amenities.append(txt)

        # Contact (liens tel:/mailto: structurés — location résidentielle)
        contact: dict = {}
        tel = soup.select_one('a[href^="tel:"]')
        if tel:
            digits = re.sub(r"\D", "", tel.get("href", ""))[-10:]
            if len(digits) == 10:
                contact["phone"] = "-".join(
                    (digits[:3], digits[3:6], digits[6:]))
        mail = soup.select_one('a[href^="mailto:"]')
        if mail:
            contact["email"] = mail.get("href", "")[7:].split("?")[0]

        # Disponibilité (texte du site) + promotion (dans la description,
        # pour ne pas fausser la date de disponibilité normalisée)
        page_text = re.sub(r"\s+", " ", soup.get_text(" ", strip=True))
        availability = "Disponible"
        if re.search(r"Maintenant disponible à la location", page_text, re.I):
            availability = "Maintenant disponible à la location"
        m = re.search(r"(\d\s*MOIS OFFERTS[^<*]{0,80})", page_text, re.I)
        if m:
            promo = "Promotion : " + re.sub(r"\s+", " ", m.group(1)).strip()
            desc = (desc + " — " + promo)[:600]

        # Cartes de prix : « 2 chambres (4 ½) », superficie, « À partir de … $ »
        for card in soup.select(".price-card"):
            try:
                lst = self._parse_card(card, images, desc, amenities,
                                       availability, contact)
            except Exception:
                continue
            if lst:
                listings.append(lst)
        return listings

    def _parse_card(self, card, images, desc, amenities,
                    availability, contact=None) -> Listing | None:
        head = card.get_text(" ", strip=True)          # « 2 chambres (4 ½) »
        m = re.search(r"(\d)\s*chambres?", head, re.I)
        mtype = re.search(r"\(\s*(\d)\s*(?:½|1/2)\s*\)", head)
        if not (m or mtype):
            return None
        if mtype:
            unit_type = f"{mtype.group(1)}½"
        else:
            n = int(m.group(1))
            unit_type = {0: "Studio", 1: "3½", 2: "4½", 3: "5½"}.get(
                n, f"{n} chambres")

        # superficie et prix dans les blocs suivants du même conteneur
        container = card.parent
        sqft = price_label = ""
        if container:
            sq = container.select_one(".superficie")
            if sq:
                sqft = sq.get_text(" ", strip=True)
            mt = container.select_one(".montant")
            if mt:
                price_label = mt.get_text(" ", strip=True)

        price = parse_price(price_label)
        n_ch = m.group(1) if m else {"4½": "2", "5½": "3"}.get(unit_type, "")

        return Listing(
            source=self.source_id,
            external_id=f"axia-lachine-{unit_type.replace('½', '.5')}",
            url=BASE + "/#appartements",
            title=f"AXIA Appartements — {n_ch} chambres ({unit_type})",
            address=ADDRESS,
            sector="Lachine",
            city="Montréal",
            unit_type=unit_type,
            price=price,
            price_label=re.sub(r"\s+", " ", price_label).strip(),
            availability=availability,
            description=(f"{sqft}. {desc}" if sqft else desc)[:600],
            amenities=amenities[:15],
            details={"contact": dict(contact)} if contact else {},
            images=images,
        )