"""Parser for Northwoods Storage Systems Llc."""

from __future__ import annotations

import re

from bs4 import BeautifulSoup

from src.parsers.base import BaseParser, ParseResult, UnitResult


class Facility003695Parser(BaseParser):
    """Extract storage units from Northwoods Storage Systems Llc."""

    platform = "custom_facility_003695"

    _UNIT_RE = re.compile(
        r"(\d+\s*[\'\'\u2032]?\s*[xX\u00d7]\s*\d+\s*[\'\'\u2032]?)"
        r"[^\$]{0,120}"
        r"\$(\d[\d,.]*)",
        re.DOTALL,
    )

    _PRICE_SIZE_RE = re.compile(
        r"\$(\d[\d,.]*)"
        r".{0,120}"
        r"(\d+\s*[\'\'\u2032]?\s*[xX\u00d7]\s*\d+\s*[\'\'\u2032]?)",
        re.DOTALL,
    )

    _SIZE_ONLY_RE = re.compile(
        r"(\d+\s*[\'\'\u2032]?\s*[xX\u00d7]\s*\d+\s*[\'\'\u2032]?)"
    )

    def parse(self, html: str, url: str = "") -> ParseResult:
        soup = BeautifulSoup(html, "lxml")
        result = ParseResult(platform=self.platform, parser_name=self.__class__.__name__)

        for tag in soup.find_all(["script", "style"]):
            tag.decompose()

        body_text = soup.get_text(separator="\n")

        seen: set[tuple[str, str]] = set()

        # Try size-then-price pattern
        for m in self._UNIT_RE.finditer(body_text):
            size_text = m.group(1).strip()
            price_text = m.group(2).strip()
            key = (size_text, price_text)
            if key in seen:
                continue
            seen.add(key)

            unit = UnitResult()
            unit.size = size_text
            w, ln, sq = self.normalize_size(size_text)
            if w is not None:
                unit.metadata = {"width": w, "length": ln, "sqft": sq}
            unit.price = self.normalize_price(price_text)
            unit.description = m.group(0).strip()[:200]
            if unit.size or unit.price:
                result.units.append(unit)

        # Try price-then-size pattern if no results
        if not result.units:
            for m in self._PRICE_SIZE_RE.finditer(body_text):
                price_text = m.group(1).strip()
                size_text = m.group(2).strip()
                key = (size_text, price_text)
                if key in seen:
                    continue
                seen.add(key)

                unit = UnitResult()
                unit.size = size_text
                w, ln, sq = self.normalize_size(size_text)
                if w is not None:
                    unit.metadata = {"width": w, "length": ln, "sqft": sq}
                unit.price = self.normalize_price(price_text)
                unit.description = m.group(0).strip()[:200]
                if unit.size or unit.price:
                    result.units.append(unit)

        # Fallback: extract sizes without prices
        if not result.units:
            seen_sizes: set[str] = set()
            for m in self._SIZE_ONLY_RE.finditer(body_text):
                size_text = m.group(1).strip()
                if size_text in seen_sizes:
                    continue
                w, ln, sq = self.normalize_size(size_text)
                if w is None or w < 3 or ln < 3:
                    continue
                seen_sizes.add(size_text)
                unit = UnitResult()
                unit.size = size_text
                unit.metadata = {"width": w, "length": ln, "sqft": sq}
                result.units.append(unit)

        if not result.units:
            result.warnings.append("No units found via regex")

        return result