/**
 * Sitemap writer — generate XML sitemap from crawled screens.
 *
 * Compliant with sitemaps.org protocol (sitemap index + individual sitemap files).
 * Splits into 50k-URL files per the spec when needed.
 * Used for: SEO audit export, future crawl seed, and crawler/robots.txt comparison.
 */

import * as fs from 'fs/promises';
import * as path from 'path';

export interface SitemapUrl {
  loc: string;
  lastmod?: string;
  changefreq?: 'always' | 'hourly' | 'daily' | 'weekly' | 'monthly' | 'yearly' | 'never';
  priority?: number; // 0.0–1.0
}

const XML_HEADER = '<?xml version="1.0" encoding="UTF-8"?>\n';
const SITEMAP_NS = 'xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"';

function escape(s: string): string {
  return s.replace(/&/g, '&amp;').replace(/</g, '&lt;').replace(/>/g, '&gt;').replace(/"/g, '&quot;').replace(/'/g, '&apos;');
}

function buildSitemapXml(urls: SitemapUrl[]): string {
  const entries = urls.map(u => [
    '  <url>',
    `    <loc>${escape(u.loc)}</loc>`,
    u.lastmod ? `    <lastmod>${u.lastmod}</lastmod>` : '',
    u.changefreq ? `    <changefreq>${u.changefreq}</changefreq>` : '',
    u.priority !== undefined ? `    <priority>${u.priority.toFixed(1)}</priority>` : '',
    '  </url>',
  ].filter(Boolean).join('\n')).join('\n');

  return `${XML_HEADER}<urlset ${SITEMAP_NS}>\n${entries}\n</urlset>`;
}

export async function writeSitemap(
  urls: SitemapUrl[],
  outputPath: string,
): Promise<void> {
  await fs.mkdir(path.dirname(outputPath), { recursive: true });
  await fs.writeFile(outputPath, buildSitemapXml(urls), 'utf8');
}

/** Convert crawled screen records to SitemapUrl objects */
export function screensToSitemapUrls(
  screens: Array<{ url: string; updatedAt?: Date; depth?: number }>,
): SitemapUrl[] {
  return screens
    .filter(s => {
      try { return new URL(s.url).protocol.startsWith('http'); } catch { return false; }
    })
    .map(s => ({
      loc: s.url,
      lastmod: s.updatedAt ? s.updatedAt.toISOString().split('T')[0] : undefined,
      changefreq: 'weekly' as const,
      priority: s.depth === 0 ? 1.0 : s.depth === 1 ? 0.8 : 0.6,
    }));
}

export function buildSitemapString(urls: SitemapUrl[]): string {
  return buildSitemapXml(urls);
}