const LIST_URL = "https://racingnews365.nl/f1-nieuws";
const UA =
  "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36";

const SKIP = new Set([
  "f1-nieuws",
  "liveblog-formule-1",
  "coureurs",
  "formule-1-op-tv",
  "f1-podcast",
  "uitslag-formule-1",
  "indycar",
  "abonnementen",
  "premium",
  "video",
  "interview",
  "podcast",
  "f1",
]);

export type F1NewsItem = {
  id: string;
  title: string;
  url: string;
  imageUrl: string | null;
  publishedAt: string | null;
  source: string;
};

function decodeHtml(s: string): string {
  return s
    .replace(/"/g, '"')
    .replace(/'/g, "'")
    .replace(/'/g, "'")
    .replace(/&/g, "&")
    .replace(/&lt;/g, "<")
    .replace(/&gt;/g, ">")
    .replace(/&#(\d+);/g, (_, n) => String.fromCharCode(Number(n)));
}

function slugOf(url: string): string {
  try {
    return new URL(url).pathname.replace(/^\/+|\/+$/g, "");
  } catch {
    return "";
  }
}

export function parseRacingNews365(html: string, limit = 20): F1NewsItem[] {
  const items: F1NewsItem[] = [];
  const seen = new Set<string>();
  const openRe = /<a\b([^>]+)>/gi;
  let m: RegExpExecArray | null;
  while ((m = openRe.exec(html)) && items.length < limit) {
    const attrs = m[1] || "";
    if (!/\bcard\b/i.test(attrs)) continue;
    const href = attrs.match(/href="(https:\/\/racingnews365\.nl\/[^"]+)"/i)?.[1];
    if (!href) continue;
    const slug = slugOf(href);
    if (!slug || SKIP.has(slug) || slug.includes("/") || seen.has(href) || slug.length < 12) continue;
    const chunk = html.slice(m.index, m.index + 4500);
    const titleMatch =
      chunk.match(/data-ab-title="\d+">([^<]+)/i) ||
      chunk.match(/class="card__title"[^>]*>\s*<span[^>]*>([^<]+)/i);
    if (!titleMatch) continue;
    const title = decodeHtml(titleMatch[1].trim());
    if (!title) continue;
    const timeMatch = chunk.match(/datetime="([^"]+)"/i);
    const imgMatch = chunk.match(/https:\/\/cdn\.racingnews365\.com\/[^"'>\s]+/);
    let imageUrl = imgMatch ? imgMatch[0].replace(/&amp;/g, "&") : null;
    if (imageUrl) {
      imageUrl = imageUrl.replace(/width=\d+/i, "width=720").replace(/height=\d+/i, "height=405");
    }
    const idMatch = attrs.match(/data-id="(\d+)"/) || chunk.match(/data-id="(\d+)"/);
    seen.add(href);
    items.push({
      id: idMatch?.[1] || slug,
      title,
      url: href,
      imageUrl,
      publishedAt: timeMatch?.[1] || null,
      source: "RacingNews365",
    });
  }
  return items;
}

let newsCache: { at: number; items: F1NewsItem[] } | null = null;

export async function getF1News(limit = 16): Promise<F1NewsItem[]> {
  if (newsCache && Date.now() - newsCache.at < 8 * 60_000) {
    return newsCache.items.slice(0, limit);
  }
  const res = await fetch(LIST_URL, {
    headers: {
      Accept: "text/html,application/xhtml+xml",
      "User-Agent": UA,
      "Accept-Language": "nl-NL,nl;q=0.9",
    },
    signal: AbortSignal.timeout(18_000),
  });
  if (!res.ok) throw new Error(`RacingNews365 HTTP ${res.status}`);
  const html = await res.text();
  const items = parseRacingNews365(html, 24);
  newsCache = { at: Date.now(), items };
  return items.slice(0, limit);
}