"""Pure parser for the news-summarizer reduce JSON / geo / importance contract."""

from __future__ import annotations

import json
import re

_EMPTY = {"summary_en": "", "ticker": [], "map_items": []}
_KEEP = frozenset({"critical", "high"})
_RANK = {"critical": 0, "high": 1, "medium": 2, "low": 3}
_THINK_RE = re.compile(r"<think>.*?</think>", re.DOTALL)
_FENCE_RE = re.compile(r"```(?:json)?", re.IGNORECASE)

TICKER_HEADLINE_MAX = 140
MAP_HEADLINE_MAX = 160
TICKER_CAP = 12
MAP_CAP = 20


def parse_reduce_json(raw: str) -> dict:
    try:
        text = _THINK_RE.sub("", raw or "")
        text = _FENCE_RE.sub("", text)
        start = text.find("{")
        end = text.rfind("}")
        if start == -1 or end == -1 or end < start:
            return dict(_EMPTY)
        data = json.loads(text[start : end + 1])
        if not isinstance(data, dict):
            return dict(_EMPTY)
        summary = data.get("summary_en", "")
        ticker = data.get("ticker", [])
        map_items = data.get("map_items", [])
        return {
            "summary_en": summary if isinstance(summary, str) else "",
            "ticker": ticker if isinstance(ticker, list) else [],
            "map_items": map_items if isinstance(map_items, list) else [],
        }
    except Exception:
        return dict(_EMPTY)


def clamp_coords(lat, lon) -> tuple[float, float] | None:
    try:
        lat_f = float(lat)
        lon_f = float(lon)
    except (TypeError, ValueError):
        return None
    if not (-90 <= lat_f <= 90 and -180 <= lon_f <= 180):
        return None
    return (lat_f, lon_f)


def _trimmed_headline(row: dict, limit: int) -> str:
    headline = row.get("headline") or ""
    if not isinstance(headline, str):
        headline = str(headline)
    return headline.strip()[:limit]


def select_ticker(rows: list) -> list:
    flagged = []
    medium = []
    low = []
    for row in rows:
        imp = row.get("importance")
        if imp not in _RANK:
            continue
        headline = _trimmed_headline(row, TICKER_HEADLINE_MAX)
        if not headline:
            continue
        item = dict(row)
        item["headline"] = headline
        if imp in _KEEP:
            flagged.append(item)
        elif imp == "medium":
            medium.append(item)
        else:
            low.append(item)
        if len(flagged) >= TICKER_CAP:
            break
    if flagged:
        return flagged[:TICKER_CAP]
    return (medium + low)[:TICKER_CAP]


def select_map(items: list) -> list:
    out = []
    for row in items:
        if row.get("importance") not in _KEEP:
            continue
        headline = _trimmed_headline(row, MAP_HEADLINE_MAX)
        if not headline:
            continue
        coords = clamp_coords(row.get("lat"), row.get("lon"))
        if coords is None:
            continue
        item = dict(row)
        item["headline"] = headline
        item["lat"], item["lon"] = coords
        out.append(item)
        if len(out) >= MAP_CAP:
            break
    return out