# -----------------------------------------------------------------------------
# Lou-Ka — Agrégateur de logements à louer (province de Québec)
# Auteur : Simon-Pierre Boucher — contact@spboucher.ai
# connectors/plantation.py : connecteur Domaine de la Plantation
#   (domainedelaplantation.ca) — Shawinigan, secteur Saint-Georges-de-Champlain
#   (entre Grand-Mère et le Lac-à-la-Tortue). Nouveau quartier de 4½ neufs de
#   Construction Michael Massicotte. Site vitrine Pixpa page unique rendu
#   serveur : section « Prix » avec paires <h3> typologie (« Demi sous-sol
#   4 ½ » / « Étages 4 ½ ») + <h4> prix (« 1295$ par mois »), section
#   « Inclusions » en items, photos Pixpa en data-src (lazy-load).
#   Granularité = typologie (pas d'inventaire par unité publié).
#   external_id = typologie en slug — stable.
# -----------------------------------------------------------------------------
from __future__ import annotations

import re

from bs4 import BeautifulSoup

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

BASE = "https://www.domainedelaplantation.ca"
PAGE_URL = f"{BASE}/"

ADDRESS = "Saint-Georges-de-Champlain, Shawinigan"

PRICE_RE = re.compile(r"([\d\s  ]{3,7})\$\s*par\s*mois", re.I)
UNIT_RE = re.compile(r"(\d)\s*(?:½|1/2|&frac12;)")
IMG_RE = re.compile(
    r"https://px-web-images[\w.-]*\.pixpa\.com/[^\"'\s)]+", re.I)


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


def _num(txt: str) -> float | None:
    n = re.sub(r"[\s  ]", "", txt or "")
    try:
        return float(n)
    except ValueError:
        return None


class PlantationConnector(BaseConnector):
    source_id = "plantation"
    request_delay = 0.8

    def fetch(self) -> list[Listing]:
        html = self.get(PAGE_URL).text
        soup = BeautifulSoup(html, "html.parser")

        # Photos (lazy-load Pixpa) — communes au projet
        images = [u for u in dict.fromkeys(
            m.group(0) for el in soup.select("[data-src]")
            for m in [IMG_RE.match(el["data-src"])] if m)][:12]

        # Inclusions : items de la section « Inclusions »
        amenities: list[str] = []
        incl = soup.find("h2", string=re.compile(r"Inclusions", re.I))
        if incl:
            sec = incl.parent
            for _ in range(4):
                if sec is None:
                    break
                txt = sec.get_text("\n", strip=True)
                if len(txt) > 60:
                    break
                sec = sec.parent
            if sec is not None:
                for line in sec.get_text("\n", strip=True).split("\n"):
                    line = re.sub(r"\s+", " ", line).strip()
                    if line and not re.match(r"Inclusions", line, re.I) \
                            and len(line) < 60 and line not in amenities:
                        amenities.append(line)

        # Accroche du projet (texte héro)
        blurb = ""
        h3 = soup.find("h3", string=re.compile(r"Grand-Mère|Lac-à-la-Tortue"))
        if h3:
            blurb = re.sub(r"\s+", " ", h3.get_text(" ", strip=True))[:600]

        # Section « Prix » : paires h3 (typologie) + h4 (prix)
        listings: list[Listing] = []
        prix_h2 = soup.find("h2", string=re.compile(r"^\s*Prix\s*$", re.I))
        if not prix_h2:
            return listings
        sec = prix_h2
        for _ in range(4):
            sec = sec.parent
            if sec is None:
                return listings
            if sec.find("h4"):
                break

        current_label = ""
        for el in sec.find_all(["h3", "h4"]):
            text = re.sub(r"\s+", " ", el.get_text(" ", strip=True))
            if el.name == "h3":
                current_label = text
                continue
            pm = PRICE_RE.search(text)
            if not pm or not current_label:
                continue
            unit_type = ""
            um = UNIT_RE.search(current_label)
            if um:
                unit_type = normalize_unit_type(f"{um.group(1)} ½")
            listings.append(Listing(
                source=self.source_id,
                external_id=_slug(current_label),
                url=PAGE_URL,
                title=f"{current_label} — Domaine de la Plantation",
                address=ADDRESS,
                sector="Saint-Georges-de-Champlain",
                city="Shawinigan",
                unit_type=unit_type,
                price=_num(pm.group(1)),
                price_label=text,
                description=blurb,
                amenities=list(amenities),
                images=list(images),
            ))
            current_label = ""
        return listings