URL Canonicalizer For Cache Keys

const DEFAULT_PORTS = new Map([
  ["http:", "80"],
  ["https:", "443"],
]);

export function canonicalizeUrl(input, options = {}) {
  if (typeof input !== "string" || input.trim() === "") {
    throw new TypeError("input must be a non-empty URL string");
  }
  const {
    stripFragment = true,
    sortQuery = true,
    dropQuery = [],
    keepTrailingSlash = true,
  } = options;
  if (!Array.isArray(dropQuery)) throw new TypeError("dropQuery must be an array");
  const dropped = new Set(dropQuery.map((key) => String(key)));

  let url;
  try {
    url = new URL(input.trim());
  } catch {
    throw new TypeError("invalid URL: " + input);
  }
  if (url.protocol !== "http:" && url.protocol !== "https:") {
    throw new TypeError("only http and https URLs are supported");
  }

  if (stripFragment) url.hash = "";
  if (DEFAULT_PORTS.get(url.protocol) === url.port) url.port = "";
  if (url.pathname === "") url.pathname = "/";
  if (!keepTrailingSlash && url.pathname.length > 1 && url.pathname.endsWith("/")) {
    url.pathname = url.pathname.slice(0, -1);
  }

  if (sortQuery) {
    const pairs = [...url.searchParams.entries()].filter(([key]) => !dropped.has(key));
    pairs.sort((left, right) => {
      if (left[0] !== right[0]) return left[0] < right[0] ? -1 : 1;
      if (left[1] === right[1]) return 0;
      return left[1] < right[1] ? -1 : 1;
    });
    url.search = "";
    for (const [key, value] of pairs) url.searchParams.append(key, value);
  }
  return url.toString();
}

const urlKeyCache = new Map();

// Memoize canonical results so one logical URL maps to one cache key.
export function urlCacheKey(input, options = {}) {
  const fingerprint = [input, options.sortQuery ?? true, options.stripFragment ?? true, options.keepTrailingSlash ?? true, (options.dropQuery ?? []).join(',')].join('|');
  let key = urlKeyCache.get(fingerprint);
  if (key === undefined) {
    key = canonicalizeUrl(input, options);
    urlKeyCache.set(fingerprint, key);
  }
  return key;
}