# -----------------------------------------------------------------------------
# Lou-Ka — Location court terme
# connectors/chaletsbsl.py : Chalets BSL (chaletsbsl.com) — 4 chalets avec spa
#   sur un domaine privé de Saint-Simon-de-Rimouski (Bas-Saint-Laurent).
#
# Méthode : sitemap.xml (index) → sitemap_sections_*.xml → URLs /chalets/<slug>
#   + lastmod (clé du cache détail ; les /en/ sont ignorées). Pages statiques
#   (CMS maison Bootstrap) :
#     - <h3 class="mt-4 fw-600"> = titre (préfixe « Chalets BSL - » retiré) ;
#     - <h5 class="prix"> « À partir de 610$ pour 2 nuits » → price_night ;
#     - <p class="capacite"> « 2 pers. 1 chbre. 1 sdb. » ;
#     - description = <p> entre la capacité et le bloc country-info ;
#     - commodités = <h6> des blocs country-name ;
#     - galerie = var chaletImgs = {"ete": […], …} (JSON par saison) ;
#     - lat/lng = const myLatLng = { lat: …, lng: … } ; CITQ en pied de fiche.
# -----------------------------------------------------------------------------
from __future__ import annotations

import html as _html
import json
import re

from ..schema import StListing
from .base import StConnector

SITE = "https://chaletsbsl.com"
SITEMAP = SITE + "/sitemap.xml"

_TAG_RE = re.compile(r"<[^>]+>")


def _text(fragment: str) -> str:
    return _html.unescape(re.sub(r"\s+", " ", _TAG_RE.sub(" ", fragment))).strip()


class ChaletsBsl(StConnector):
    source_id = "chaletsbsl"
    request_delay = 1.0

    # -- page détail ----------------------------------------------------------
    def _detail(self, url: str) -> dict:
        h = self.get(url).text
        d: dict = {}

        m = re.search(r'(?s)<h3 class="mt-4 fw-600[^"]*"[^>]*>(.*?)</h3>', h)
        if m:
            d["title"] = re.sub(r"^Chalets BSL\s*-\s*", "", _text(m.group(1)))

        # « À partir de 610$ pour 2 nuits » → 305 $/nuit
        m = re.search(r"À partir de\s*([\d\s]+)\$\s*pour\s*(\d+)\s*nuits", h)
        if m:
            total = float(m.group(1).replace(" ", ""))
            nights = int(m.group(2))
            if nights and 20 <= total / nights <= 20000:
                d["price_night"] = round(total / nights)
                d["price_ref"] = f"{total:g} $ pour {nights} nuits"

        # « 2 pers. 1 chbre. 1 sdb. »
        m = re.search(r'(?s)<p class="capacite[^"]*"[^>]*>(.*?)</p>', h)
        if m:
            frag = _text(m.group(1))
            for pat, key in ((r"(\d+)\s*pers", "capacity"),
                             (r"(\d+)\s*chbre", "bedrooms"),
                             (r"(\d+)\s*sdb", "bathrooms")):
                mm = re.search(pat, frag)
                if mm:
                    d[key] = float(mm.group(1))

        # description : les <p> entre la capacité et le bloc country-info
        i = h.find('class="capacite')
        j = h.find("country-info")
        if 0 < i < j:
            paras = [_text(p) for p in
                     re.findall(r"(?s)<p[^>]*>(.*?)</p>", h[i:j])]
            texte = "\n".join(p for p in paras if len(p) > 40)
            if texte:
                d["description"] = texte[:5000]

        # commodités : les <h6> (tous portés par les blocs country-name)
        amen: list[str] = []
        for h6 in re.findall(r"(?s)<h6[^>]*>(.*?)</h6>", h):
            t = _text(h6)
            if t and t not in amen:
                amen.append(t)
        if amen:
            d["amenities"] = amen

        m = re.search(r"const myLatLng = \{\s*lat:\s*(-?\d+\.\d+),"
                      r"\s*lng:\s*(-?\d+\.\d+)", h)
        if m:
            d["lat"], d["lng"] = float(m.group(1)), float(m.group(2))
        m = re.search(r"CITQ\D{0,25}(\d{6})", h)
        if m:
            d["citq"] = m.group(1)

        # galerie : var chaletImgs = {"ete": […], "hiver": […]}
        m = re.search(r"var chaletImgs = (\{.*?\});", h, re.S)
        if m:
            try:
                seasons = json.loads(m.group(1))
                imgs: list[str] = []
                for key in ("ete", *sorted(k for k in seasons if k != "ete")):
                    for u in seasons.get(key) or []:
                        if isinstance(u, str) and u.startswith("https://") \
                                and u not in imgs and len(imgs) < 20:
                            imgs.append(u)
                if imgs:
                    d["images"] = imgs
            except ValueError:
                pass
        return d

    # -- contrat --------------------------------------------------------------
    def fetch(self) -> list[StListing]:
        index = self.get(SITEMAP).text
        entries: list[tuple[str, str]] = []
        for sub in re.findall(r"<loc>([^<]+)</loc>", index):
            if "sitemap_sections" not in sub:
                continue
            xml = self.get(sub).text
            entries += re.findall(r"(?s)<url>\s*<loc>([^<]+)</loc>"
                                  r"(?:\s*<lastmod>([^<]*)</lastmod>)?", xml)

        listings: list[StListing] = []
        vus: set[str] = set()
        for url, lastmod in entries:
            m = re.match(r"https://chaletsbsl\.com/chalets/([^/]+)/?$", url)
            if not m:
                continue
            slug = m.group(1)
            if slug in vus:
                continue
            vus.add(slug)

            det = self.detail(slug, lastmod or "v1",
                              lambda u=url: self._detail(u))
            title = det.get("title") or ""
            if not title:
                continue

            price = det.get("price_night")
            details = {k: v for k, v in {
                "price_ref": det.get("price_ref") or "",
                "domain": "Chalets BSL",
            }.items() if v}

            listings.append(StListing(
                source=self.source_id,
                external_id=slug,
                url=url,
                title=title,
                property_type="Chalet",
                city="Saint-Simon-de-Rimouski",
                region="Bas-Saint-Laurent",
                price_night=float(price) if price else None,
                price_label=f"à partir de {price:g} $ / nuit" if price else "",
                capacity=det.get("capacity"),
                bedrooms=det.get("bedrooms"),
                bathrooms=det.get("bathrooms"),
                citq=det.get("citq") or "",
                description=det.get("description") or "",
                amenities=det.get("amenities") or [],
                details=details,
                images=det.get("images") or [],
                lat=det.get("lat"),
                lng=det.get("lng"),
            ))
        return listings