// author: simon-pierre boucher <contact@spboucher.ai>
export interface SitemapUrl {
  readonly loc: string;
  readonly lastModified?: string;
}

export interface ParsedSitemap {
  readonly urls: SitemapUrl[];
  readonly childSitemaps: string[];
}

function decodeXml(s: string): string {
  return s
    .replace(/<!\[CDATA\[([\s\S]*?)\]\]>/g, "$1")
    .replace(/&lt;/g, "<")
    .replace(/&gt;/g, ">")
    .replace(/&quot;/g, '"')
    .replace(/&#39;/g, "'")
    .replace(/&apos;/g, "'")
    .replace(/&amp;/g, "&")
    .trim();
}

function tag(block: string, name: string): string | undefined {
  const m = new RegExp(`<${name}[^>]*>([\\s\\S]*?)</${name}>`, "i").exec(block);
  return m?.[1] !== undefined ? decodeXml(m[1]) : undefined;
}

function blocks(xml: string, name: string): string[] {
  const out: string[] = [];
  const re = new RegExp(`<${name}[^>]*>([\\s\\S]*?)</${name}>`, "gi");
  let m: RegExpExecArray | null;
  while ((m = re.exec(xml)) !== null) {
    if (m[1] !== undefined) out.push(m[1]);
  }
  return out;
}

/**
 * Parse a sitemap or sitemap index (§14.3). Returns child sitemap URLs (for
 * recursion through indexes) and page URLs with optional lastmod.
 */
export function parseSitemap(xml: string): ParsedSitemap {
  const childSitemaps: string[] = [];
  for (const block of blocks(xml, "sitemap")) {
    const loc = tag(block, "loc");
    if (loc !== undefined && loc !== "") childSitemaps.push(loc);
  }

  const urls: SitemapUrl[] = [];
  for (const block of blocks(xml, "url")) {
    const loc = tag(block, "loc");
    if (loc === undefined || loc === "") continue;
    const lastmod = tag(block, "lastmod");
    urls.push(lastmod !== undefined ? { loc, lastModified: lastmod } : { loc });
  }

  if (urls.length === 0 && childSitemaps.length === 0) {
    const re = /<loc[^>]*>([\s\S]*?)<\/loc>/gi;
    let m: RegExpExecArray | null;
    while ((m = re.exec(xml)) !== null) {
      const loc = m[1] !== undefined ? decodeXml(m[1]) : "";
      if (loc !== "") urls.push({ loc });
    }
  }

  return { urls, childSitemaps };
}