# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/louis_alexandre.py : connecteur Le Louis-Alexandre
#   (condoslouisalexandre.ca — Longueuil, 2100, rue René, près du
#   Vieux-Longueuil). WordPress + Elementor rendu serveur : la page
#   « Appartements disponibles » liste chaque unité en blocs Elementor
#   « Condo #NNN | Superficie de X pi² | n chambres, n sdb | PRIX $ / mois »
#   avec image du plan et fiche PDF par unité. Le numéro d'unité (#NNN)
#   sert d'identifiant stable.
# -----------------------------------------------------------------------------
from __future__ import annotations

import re

from bs4 import BeautifulSoup

from ..schema import Listing
from .base import BaseConnector

BASE = "https://condoslouisalexandre.ca"
LIST_URL = f"{BASE}/appartements-disponibles-longueuil/"

ADDRESS = "2100, rue René, Longueuil"
CITY = "Longueuil"

UNIT_RE = re.compile(r"^Condo\s*#(\d+)\s*$")
SQFT_RE = re.compile(r"Superficie\s+de\s+([\d\s]+)\s*pi", re.I)
ROOMS_RE = re.compile(r"(Studio|\d\s*chambres?)\s*,\s*(\d)\s*sdb", re.I)
PRICE_RE = re.compile(r"(\d[\d\s]{2,6})\s*\$\s*/\s*mois", re.I)
IMG_RE = re.compile(
    r"https://condoslouisalexandre\.ca/wp-content/uploads/"
    r"[^\"\s\\']+\.(?:jpe?g|png|webp)", re.I)
SKIP_IMG_RE = re.compile(r"logo|icon|favicon|-\d{2,3}x\d{2,3}\.", re.I)


class LouisAlexandreConnector(BaseConnector):
    source_id = "louis_alexandre"
    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")

        # contact de l'immeuble (mailto:/tel: structurés dans la page)
        contact: dict = {}
        m = re.search(r'href="mailto:([^"?]+)"', html)
        if m:
            contact["email"] = m.group(1).strip().lower()
        m = re.search(r'href="tel:\+?1?[\s.\-]?(\d{3})[\s.\-]?(\d{3})'
                      r'[\s.\-]?(\d{4})"', html)
        if m:
            contact["phone"] = f"{m.group(1)}-{m.group(2)}-{m.group(3)}"

        seen: set[str] = set()
        for node in soup.find_all(string=UNIT_RE):
            try:
                unit_no = UNIT_RE.match(node.strip()).group(1)
                if unit_no in seen:
                    continue

                # bloc de l'unité : on remonte jusqu'au conteneur Elementor
                # qui porte le prix (texte « NNN $ / mois »)
                block = node.parent
                while block is not None and "$" not in block.get_text():
                    block = block.parent
                if block is None:
                    continue
                text = block.get_text(" | ", strip=True)
                if len(re.findall(r"Condo\s*#", text)) != 1:
                    continue    # conteneur trop large (plusieurs unités)
                seen.add(unit_no)

                price = None
                price_label = ""
                m = PRICE_RE.search(text)
                if m:
                    price_label = m.group(0)
                    try:
                        val = float(m.group(1).replace(" ", "")
                                    .replace(" ", "").replace("\xa0", ""))
                        if 300 <= val <= 20000:
                            price = val
                    except ValueError:
                        pass

                area = None
                m = SQFT_RE.search(text)
                if m:
                    try:
                        area = float(m.group(1).replace(" ", "")
                                     .replace("\xa0", ""))
                    except ValueError:
                        area = None

                unit_type = ""
                bedrooms = None
                bathrooms = None
                m = ROOMS_RE.search(text)
                if m:
                    rooms = m.group(1)
                    bathrooms = float(m.group(2))
                    if re.match(r"studio", rooms, re.I):
                        unit_type = "Studio"
                        bedrooms = 0.0
                    else:
                        n = int(re.match(r"\d+", rooms).group(0))
                        bedrooms = float(n)
                        unit_type = f"{n + 2}½"

                # image du plan + fiche PDF : conteneur parent (colonne voisine)
                images: list[str] = []
                pdf = ""
                wrap = block.parent
                while wrap is not None and not wrap.find("img"):
                    wrap = wrap.parent
                if wrap is not None and \
                        len(re.findall(r"Condo\s*#", wrap.get_text())) == 1:
                    for u in dict.fromkeys(IMG_RE.findall(str(wrap))):
                        if not SKIP_IMG_RE.search(u):
                            images.append(u)
                    a = wrap.find("a", href=re.compile(r"\.pdf$", re.I))
                    if a:
                        pdf = a.get("href", "")

                details: dict = {}
                if contact:
                    details["contact"] = dict(contact)
                if pdf:
                    details["plan_pdf"] = pdf

                amenities = []
                if area:
                    amenities.append(f"Superficie de {int(area)} pi²")
                if bathrooms:
                    amenities.append(f"{int(bathrooms)} salle(s) de bain")

                listings.append(Listing(
                    source=self.source_id,
                    external_id=f"condo-{unit_no}",
                    url=LIST_URL,
                    title=f"Condo #{unit_no} — Le Louis-Alexandre",
                    address=ADDRESS,
                    city=CITY,
                    unit_type=unit_type,
                    bedrooms=bedrooms,
                    bathrooms=bathrooms,
                    price=price,
                    price_label=price_label,
                    area_sqft=area,
                    amenities=amenities,
                    details=details,
                    images=images[:10],
                ))
            except Exception:
                continue
        return listings