"""Yandex Realty ЖК landing page parser.

URL pattern: /{city}/kupit/novostrojka/<slug>-<id>/
Reference target: ЖК Татлин (id=1592987, slug=tatlin) — comfort+, June 2023,
PRINZIP, rating 4.3, 1505 ratings, 353 text reviews, coords (56.855312, 60.576668).
"""

from __future__ import annotations

import logging
import re
from typing import Any

from pydantic import BaseModel, Field
from selectolax.parser import HTMLParser, Node

from app.services.scraper_settings import get_scraper_delay
from app.services.scrapers.base import BaseScraper
from app.services.scrapers.yandex_helpers import (
    RE_JK_ID,
    parse_house_class,
    parse_house_type,
)

logger = logging.getLogger(__name__)


# ── Models ───────────────────────────────────────────────────────────────────


class JKMetroStation(BaseModel):
    name: str
    walk_min: int | None = None


class YandexNewbuildingInfo(BaseModel):
    """ЖК landing scrape payload."""

    ext_id: str  # '1592987'
    ext_slug: str  # 'tatlin'
    source_url: str

    # Identification
    name: str | None = None  # 'ЖК «Татлин»'
    address: str | None = None  # 'Екатеринбург, ул. Черепанова / ул. Готвальда'

    # Coords (inline in HTML — Yandex unique vs Avito/Cian)
    lat: float | None = None
    lon: float | None = None

    # Classification + commission
    house_class: str | None = None  # 'comfort_plus' etc.
    commission_year: int | None = None
    commission_month: str | None = None  # 'июнь' (RU)

    # Footprint
    total_floors: int | None = None  # 35
    corpus_count: int | None = None  # 3
    total_area_ha: float | None = None  # 1.5

    # Developer
    developer_name: str | None = None  # 'PRINZIP'
    developer_url: str | None = None
    developer_other_jk: list[str] = Field(default_factory=list)

    # Reviews
    rating: float | None = None  # 4.3
    ratings_count: int | None = None  # 1505 — оценок
    text_reviews_count: int | None = None  # 353 — текстовых отзывов

    # NLP / description
    description: str | None = None

    # Metro
    metro_stations: list[JKMetroStation] = Field(default_factory=list)

    # House type (best-effort NLP from description)
    house_type: str | None = None

    raw_payload: dict[str, Any] | None = None


# ── Regex constants ───────────────────────────────────────────────────────────

RE_FLOORS_TOWERS = re.compile(r"(\d+)-этажн\w+\s+башн", re.IGNORECASE)
RE_AREA_HA = re.compile(r"участке\s+([\d,]+)\s*га", re.IGNORECASE)
RE_COMMISSION = re.compile(
    r"(?:введ[её]н|сдан)\s+в\s+эксплуатацию\s+в\s+(\w+)\s+(\d{4})", re.IGNORECASE
)
RE_RATING = re.compile(r"(\d[.,]\d)\s+из\s+5")
RE_RATINGS_COUNT = re.compile(r"(\d+)\s+оценок", re.IGNORECASE)
RE_TEXT_REVIEWS = re.compile(r"Смотреть\s+все\s+(\d+)\s+отзыв", re.IGNORECASE)
RE_CORPUS_COUNT_WORD = re.compile(
    r"(одн[ау]|две|три|четыре|пять|шесть|семь|восемь|\d+)\s+(?:[\d\s-]*)\s*-?\s*этажн",
    re.IGNORECASE,
)
RE_METRO_INLINE = re.compile(
    r"([А-ЯЁ][А-Яа-яё\s-]{2,30}?)\s+(\d+)\s*мин",
)

# Ekb-specific coord ranges (extend per-city later)
LAT_RANGE = (55.5, 57.5)
LON_RANGE = (59.5, 61.5)

# Word → number map
_WORD_NUM: dict[str, int] = {
    "одна": 1,
    "одну": 1,
    "одной": 1,
    "две": 2,
    "три": 3,
    "четыре": 4,
    "пять": 5,
    "шесть": 6,
    "семь": 7,
    "восемь": 8,
}


# ── Scraper class ─────────────────────────────────────────────────────────────


class YandexNewbuildingScraper(BaseScraper):
    name = "yandex_realty_nb"
    base_url = "https://realty.yandex.ru"
    request_delay_sec = 5.0  # class default; instance value loaded from scraper_settings

    def __init__(self) -> None:
        super().__init__()
        self.request_delay_sec = get_scraper_delay(self.name)

    async def fetch_around(
        self, lat: float, lon: float, radius_m: int = 1000
    ) -> list:  # type: ignore[override]
        raise NotImplementedError(
            "YandexNewbuildingScraper is JK-slug-based; use fetch_jk(slug, id) instead."
        )

    async def fetch_jk(
        self, jk_slug: str, jk_id: str, city: str = "ekaterinburg"
    ) -> YandexNewbuildingInfo | None:
        url = f"{self.base_url}/{city}/kupit/novostrojka/{jk_slug}-{jk_id}/"
        try:
            response = await self._http_get(url)
        except Exception:
            logger.exception("yandex nb fetch failed: %s", url)
            return None
        if response.status_code != 200:
            logger.warning("yandex nb returned %d for %s", response.status_code, url)
            return None
        result = self.parse(response.text, jk_slug=jk_slug, jk_id=jk_id, source_url=url)
        await self.sleep_between_requests()
        return result

    def parse(
        self, html: str, jk_slug: str, jk_id: str, source_url: str
    ) -> YandexNewbuildingInfo:
        tree = HTMLParser(html)
        body = tree.body
        body_text = body.text(strip=True) if body else ""

        # Header
        h1 = tree.css_first("h1")
        name = h1.text(strip=True) if h1 else None

        # Sections (heading-based extraction)
        location_text = _find_section_text(tree, "Расположение") or ""
        description_section = (
            _find_section_text(tree, "О комплексе")
            or _find_section_text(tree, "Расположение, транспортная доступность")
            or location_text
        )

        # Coords — inline scan within plausible range
        lat, lon = _extract_coords(html)

        # Class
        house_class = parse_house_class(description_section) or parse_house_class(body_text)

        # Commission
        commission_year: int | None = None
        commission_month: str | None = None
        c_m = RE_COMMISSION.search(description_section) or RE_COMMISSION.search(body_text)
        if c_m:
            commission_month = c_m.group(1).lower()
            try:
                commission_year = int(c_m.group(2))
            except ValueError:
                pass

        # Floors + corpus + area
        floors_m = RE_FLOORS_TOWERS.search(description_section) or RE_FLOORS_TOWERS.search(
            body_text
        )
        total_floors = int(floors_m.group(1)) if floors_m else None

        corpus_count = _parse_corpus_count(description_section or body_text)

        area_m = RE_AREA_HA.search(description_section) or RE_AREA_HA.search(body_text)
        total_area_ha = float(area_m.group(1).replace(",", ".")) if area_m else None

        # Developer
        dev_link = tree.css_first('[data-test="CARD_DEV_BADGE_DEVELOPER_LINK"]')
        developer_name: str | None = None
        developer_url: str | None = None
        if dev_link is not None:
            developer_name = dev_link.text(strip=True) or None
            href = dev_link.attributes.get("href", "")
            if href:
                developer_url = (
                    href
                    if href.startswith("http")
                    else f"https://realty.yandex.ru{href}"
                )

        # Developer's other JKs
        other_jk_block = tree.css_first('[data-test="CardDevSites"]')
        developer_other_jk: list[str] = []
        if other_jk_block is not None:
            for jk_a in other_jk_block.css("a"):
                t = jk_a.text(strip=True)
                if t and t not in developer_other_jk:
                    developer_other_jk.append(t)

        # Reviews
        rating_m = RE_RATING.search(body_text)
        rating = float(rating_m.group(1).replace(",", ".")) if rating_m else None
        rcount_m = RE_RATINGS_COUNT.search(body_text)
        ratings_count = int(rcount_m.group(1)) if rcount_m else None
        treviews_m = RE_TEXT_REVIEWS.search(body_text)
        text_reviews_count = int(treviews_m.group(1)) if treviews_m else None

        # Address (header area before lat/lon block)
        address = _extract_jk_address(body_text, name)

        # Metro stations
        metro_stations = _parse_metro(body_text)

        # NLP house_type fallback
        house_type = parse_house_type(description_section) or parse_house_type(body_text)

        return YandexNewbuildingInfo(
            ext_id=jk_id,
            ext_slug=jk_slug,
            source_url=source_url,
            name=name,
            address=address,
            lat=lat,
            lon=lon,
            house_class=house_class,
            commission_year=commission_year,
            commission_month=commission_month,
            total_floors=total_floors,
            corpus_count=corpus_count,
            total_area_ha=total_area_ha,
            developer_name=developer_name,
            developer_url=developer_url,
            developer_other_jk=developer_other_jk[:10],
            rating=rating,
            ratings_count=ratings_count,
            text_reviews_count=text_reviews_count,
            description=description_section or None,
            metro_stations=metro_stations,
            house_type=house_type,
            raw_payload={
                "description_len": len(description_section or ""),
                "body_len": len(body_text),
            },
        )


# ── helpers ───────────────────────────────────────────────────────────────────


def _extract_coords(html: str) -> tuple[float | None, float | None]:
    """Scan HTML for any `<lat>.\\d+` and `<lon>.\\d+` matching plausible city range."""
    lats = [
        float(m)
        for m in re.findall(r"\b(\d{2}\.\d{4,8})\b", html)
        if LAT_RANGE[0] <= float(m) <= LAT_RANGE[1]
    ]
    lons = [
        float(m)
        for m in re.findall(r"\b(\d{2}\.\d{4,8})\b", html)
        if LON_RANGE[0] <= float(m) <= LON_RANGE[1]
    ]
    lat = lats[0] if lats else None
    lon = lons[0] if lons else None
    return lat, lon


def _find_section_text(tree: HTMLParser, heading: str) -> str | None:
    for h in tree.css("h2, h3"):
        if heading.lower() in (h.text(strip=True) or "").lower():
            parts: list[str] = []
            node: Node | None = h.next
            while node is not None:
                tag = (node.tag or "").lower()
                if tag in {"h2", "h3"}:
                    break
                txt = node.text(strip=True) if hasattr(node, "text") else ""
                if txt:
                    parts.append(txt)
                node = node.next
            return " ".join(parts).strip() or None
    return None


def _parse_corpus_count(text: str) -> int | None:
    m = RE_CORPUS_COUNT_WORD.search(text)
    if not m:
        return None
    token = m.group(1).lower()
    if token.isdigit():
        return int(token)
    return _WORD_NUM.get(token)


def _parse_metro(text: str) -> list[JKMetroStation]:
    stations: list[JKMetroStation] = []
    for m in RE_METRO_INLINE.finditer(text):
        name = m.group(1).strip()
        # filter common noise
        if not 2 <= len(name) <= 40 or "ходьб" in name.lower() or "минут" in name.lower():
            continue
        try:
            stations.append(JKMetroStation(name=name, walk_min=int(m.group(2))))
        except ValueError:
            continue
        if len(stations) >= 5:
            break
    return stations


def _extract_jk_address(body_text: str, name: str | None) -> str | None:
    """Find address line — typically right after JK name, before metro entries."""
    if name and name in body_text:
        after_name = body_text.split(name, 1)[1]
        # Match 'Город, ул. X / ул. Y' until first metro mention
        m = re.match(
            r"\s*([А-ЯЁ][^•]{5,150}?)(?:\s+[А-ЯЁ][а-яё]+\s+\d+\s*мин|\d+\s+оценок|$)",
            after_name,
        )
        if m:
            addr = m.group(1).strip().rstrip(",").strip()
            if 10 < len(addr) < 200:
                return addr
    return None


# Expose RE_JK_ID at module level for external use (e.g. SERP link → newbuilding detail)
__all__ = [
    "RE_JK_ID",
    "JKMetroStation",
    "YandexNewbuildingInfo",
    "YandexNewbuildingScraper",
]