SixFlagsSuperCalendar/lib/scrapers/sixflags.ts

/**
 * Six Flags scraper — calls the internal CloudFront operating-hours API directly.
 *
 * API: https://d18car1k0ff81h.cloudfront.net/operating-hours/park/{apiId}?date=YYYYMM
 * Returns full month data in one request — no browser needed.
 *
 * Each park has a numeric API ID that must be discovered first (see scripts/discover.ts).
 * Once stored in the DB, this scraper never touches a browser again.
 *
 * Rate limiting: on 429/503, exponential backoff (30s → 60s → 120s), MAX_RETRIES attempts.
 */

const API_BASE = "https://d18car1k0ff81h.cloudfront.net/operating-hours/park";
const MAX_RETRIES = 3;
const BASE_BACKOFF_MS = 30_000;

export class RateLimitError extends Error {
  constructor(public readonly waitedMs: number) {
    super(`Rate limited — exhausted ${MAX_RETRIES} retries after ${waitedMs / 1000}s total wait`);
    this.name = "RateLimitError";
  }
}

const HEADERS = {
  "User-Agent":
    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 " +
    "(KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
  Accept: "application/json",
  "Accept-Language": "en-US,en;q=0.9",
  Referer: "https://www.sixflags.com/",
};

export interface DayResult {
  date: string; // YYYY-MM-DD
  isOpen: boolean;
  hoursLabel?: string;
}

function sleep(ms: number) {
  return new Promise<void>((r) => setTimeout(r, ms));
}

/** "04/05/2026" → "2026-04-05" */
function parseApiDate(d: string): string {
  const [m, day, y] = d.split("/");
  return `${y}-${m}-${day}`;
}

interface ApiOperatingItem {
  timeFrom: string; // "10:30" 24h
  timeTo: string;   // "20:00" 24h
}

interface ApiOperating {
  operatingTypeName: string; // "Park", "Special Event", etc.
  items: ApiOperatingItem[];
}

interface ApiDay {
  date: string;
  isParkClosed: boolean;
  operatings?: ApiOperating[];
}

/** "10:30" → "10:30am", "20:00" → "8pm", "12:00" → "12pm" */
function fmt24(time: string): string {
  const [h, m] = time.split(":").map(Number);
  const period = h >= 12 ? "pm" : "am";
  const h12 = h % 12 || 12;
  return m === 0 ? `${h12}${period}` : `${h12}:${String(m).padStart(2, "0")}${period}`;
}

interface ApiResponse {
  parkId: number;
  parkAbbreviation: string;
  parkName: string;
  dates: ApiDay[];
}

async function fetchApi(url: string, attempt = 0, totalWaitedMs = 0): Promise<ApiResponse> {
  const res = await fetch(url, { headers: HEADERS });

  if (res.status === 429 || res.status === 503) {
    const retryAfter = res.headers.get("Retry-After");
    const waitMs = retryAfter
      ? parseInt(retryAfter) * 1000
      : BASE_BACKOFF_MS * Math.pow(2, attempt);
    console.log(
      `  [rate-limited] HTTP ${res.status} — waiting ${waitMs / 1000}s (attempt ${attempt + 1}/${MAX_RETRIES})`
    );
    await sleep(waitMs);
    if (attempt < MAX_RETRIES) return fetchApi(url, attempt + 1, totalWaitedMs + waitMs);
    throw new RateLimitError(totalWaitedMs + waitMs);
  }

  if (!res.ok) throw new Error(`HTTP ${res.status} for ${url}`);
  return res.json() as Promise<ApiResponse>;
}

/**
 * Fetch operating hours for an entire month in a single API call.
 * apiId must be pre-discovered via scripts/discover.ts.
 */
export async function scrapeMonth(
  apiId: number,
  year: number,
  month: number
): Promise<DayResult[]> {
  const dateParam = `${year}${String(month).padStart(2, "0")}`;
  const url = `${API_BASE}/${apiId}?date=${dateParam}`;

  const data = await fetchApi(url);

  return data.dates.map((d): DayResult => {
    const date = parseApiDate(d.date);
    // Prefer the "Park" operating entry; fall back to first entry
    const operating =
      d.operatings?.find((o) => o.operatingTypeName === "Park") ??
      d.operatings?.[0];
    const item = operating?.items?.[0];
    const hoursLabel =
      item?.timeFrom && item?.timeTo
        ? `${fmt24(item.timeFrom)} – ${fmt24(item.timeTo)}`
        : undefined;
    // If the API says open but no hours are available, treat as closed
    const isOpen = !d.isParkClosed && hoursLabel !== undefined;
    return { date, isOpen, hoursLabel };
  });
}

/**
 * Fetch park info for a given API ID (used during discovery to identify park type).
 * Uses the current month so there's always some data.
 */
export async function fetchParkInfo(
  apiId: number
): Promise<Pick<ApiResponse, "parkId" | "parkAbbreviation" | "parkName"> | null> {
  const now = new Date();
  const dateParam = `${now.getFullYear()}${String(now.getMonth() + 1).padStart(2, "0")}`;
  const url = `${API_BASE}/${apiId}?date=${dateParam}`;
  try {
    const data = await fetchApi(url);
    return {
      parkId: data.parkId,
      parkAbbreviation: data.parkAbbreviation,
      parkName: data.parkName,
    };
  } catch {
    return null;
  }
}

/** Returns true if the API park name looks like a main theme park (not a water park or safari). */
export function isMainThemePark(parkName: string): boolean {
  const lower = parkName.toLowerCase();
  const waterParkKeywords = [
    "hurricane harbor",
    "safari",
    "water park",
    "waterpark",
    "schlitterbahn",
    "wave pool",
    "splash",
    "aquatic",
  ];
  return !waterParkKeywords.some((kw) => lower.includes(kw));
}