# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/per.py : connecteur Les Immeubles Paul-E. Richard
#   (immeublesper.com — 18 immeubles à Limoilou, Charlesbourg et Beauport).
#   Les unités en vedette sont listées sur /logements/ ; chaque fiche
#   /logement/<slug>/ fournit secteur, adresse, format, prix, disponibilité,
#   caractéristiques et galerie de photos.
# -----------------------------------------------------------------------------
from __future__ import annotations

import re

from bs4 import BeautifulSoup

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

BASE = "https://immeublesper.com"
LIST_URL = f"{BASE}/logements/"

_SIZE_SUFFIX = re.compile(r"-\d{2,4}x\d{2,4}(?=\.(?:jpg|jpeg|png|webp)$)", re.I)
_IMG_BLACKLIST = re.compile(r"ico-|logo|slide-\d|favicon", re.I)


class PERConnector(BaseConnector):
    source_id = "per"
    request_delay = 0.6
    max_details = 60        # garde-fou de fetch des fiches

    def fetch(self) -> list[Listing]:
        html = self.get(LIST_URL).text
        slugs: list[str] = []
        for m in re.finditer(r'href="https://immeublesper\.com/logement/([^/"]+)/?"', html):
            if m.group(1) not in slugs:
                slugs.append(m.group(1))

        listings: list[Listing] = []
        for slug in slugs[:self.max_details]:
            try:
                lst = self._parse_detail(slug)
                if lst:
                    listings.append(lst)
            except Exception:
                continue
        return listings

    # -- fiche /logement/<slug>/ -------------------------------------------------
    def _parse_detail(self, slug: str) -> Listing | None:
        url = f"{BASE}/logement/{slug}/"
        html = self.get(url).text
        soup = BeautifulSoup(html, "html.parser")

        # Informations générales : paires "ls-label"/"ls-data"
        info: dict[str, str] = {}
        for div in soup.select("div.ls-info"):
            label = div.select_one("span.ls-label")
            data = div.select_one("span.ls-data")
            if label and data:
                key = label.get_text(strip=True).rstrip(":").lower()
                info[key] = data.get_text(" ", strip=True)

        unit_raw = info.get("format", "")
        # exclusions : stationnement / commercial / rangement
        blob = " ".join([slug, unit_raw] + list(info.values()))
        if re.search(r"stationnement|commercial|rangement|garage|entrep[oô]t",
                     unit_raw + " " + slug, re.I):
            return None

        sector = info.get("secteur", "")
        address = info.get("adresse", "")
        price_label = info.get("prix", "")
        availability = info.get("disponibilité", info.get("disponibilite", ""))

        # numéro d'unité : texte "#3" entre les ls-info
        unit_no = ""
        details_div = soup.select_one("div.ls-single-details")
        if details_div:
            m = re.search(r"#\s*([\w\-]+)", details_div.get_text(" ", strip=True))
            if m:
                unit_no = f"#{m.group(1)}"

        # caractéristiques
        amenities: list[str] = []
        for h2 in soup.find_all("h2"):
            if "caract" in h2.get_text(strip=True).lower():
                ul = h2.find_next("ul")
                if ul:
                    for li in ul.find_all("li"):
                        t = re.sub(r"\s+", " ", li.get_text(" ", strip=True))
                        if t and not t.startswith("N.B.") and t not in amenities:
                            amenities.append(t)
                break
        for key, lbl in (("nombre de chambres", "chambre(s)"),
                         ("étage", "étage"), ("etage", "étage")):
            if info.get(key):
                amenities.append(f"{info[key]} {lbl}")

        # Animaux : champ structuré "Animaux permis: Oui/Non" de la fiche.
        # Valeur explicite (prioritaire sur la dérivation) — le texte brut
        # "Animaux permis : Non" serait sinon mal lu par la normalisation.
        pets = None
        pets_raw = (info.get("animaux permis") or "").strip()
        if pets_raw:
            amenities.append(f"Animaux permis : {pets_raw}")
            low = pets_raw.lower()
            if low.startswith("non"):
                pets = "non"
            elif low.startswith("oui"):
                pets = "oui"
            else:
                pets = "conditions"

        # images : galerie fancybox + image principale (pleine taille, dédupliquées)
        images: list[str] = []
        for a in soup.select("div.gallerie a[href], a.fancybox-thumb[href]"):
            u = a.get("href", "")
            if re.search(r"\.(?:jpg|jpeg|png|webp)$", u, re.I):
                u = _SIZE_SUFFIX.sub("", u)
                if u.startswith("/"):
                    u = BASE + u
                if u.startswith("http") and not _IMG_BLACKLIST.search(u) and u not in images:
                    images.append(u)
        main = soup.select_one("div.ls-single-image img[src]")
        if main:
            u = _SIZE_SUFFIX.sub("", main["src"])
            if u.startswith("http") and not _IMG_BLACKLIST.search(u) and u not in images:
                images.insert(0, u)
        if not images:   # repli : toutes les images d'uploads de la page
            for u in re.findall(r'https://immeublesper\.com/wp-content/uploads/'
                                r'[^"\'\s]+\.(?:jpg|jpeg|png|webp)', html, re.I):
                u = _SIZE_SUFFIX.sub("", u)
                if not _IMG_BLACKLIST.search(u) and u not in images:
                    images.append(u)

        title = ", ".join(x for x in (address or slug.replace("-", " "),
                                      unit_no, unit_raw) if x)
        return Listing(
            source=self.source_id,
            external_id=slug,
            url=url,
            title=title,
            address=address,
            sector=sector,
            city=infer_city(sector, default="Québec"),
            unit_type=normalize_unit_type(unit_raw),
            price=parse_price(price_label),
            price_label=price_label,
            availability=availability,
            pets=pets,
            amenities=amenities,
            images=images[:25],
        )