#!/usr/bin/env node
'use strict';

// 从官网 sitemap 生成 KB 类目地图 bridge/kb/11-site-map.md（旧 bridge site-map.md 的移植）。
// 只生成导航/类目页 Q/A，不包含商品详情页；生成后需重建 KB 索引。
const fs = require('fs');
const path = require('path');

const ROOT = path.resolve(__dirname, '..');
const OUT = path.join(ROOT, 'bridge', 'kb', '11-site-map.md');
const SITEMAP = process.env.SITE_MAP_URL || 'https://www.yehwang.com/sitemap.xml';
const BASE = process.env.SITE_BASE_URL || 'https://www.yehwang.com';
const WHITELIST = new Set(['jewelry', 'clothes', 'accessories', 'packing-and-display', 'bestsellers', 'new', 'top-deal', 'campaigns', 'brands', 'wholesale-service', 'factory-direct-wholesale-en', 'dropshipping', 'restock-request']);

function titleCase(segment) {
  return String(segment || '')
    .replace(/-/g, ' ')
    .replace(/\b\w/g, (c) => c.toUpperCase());
}

function buildEntries(urls) {
  const out = [];
  const seen = new Set();
  for (const url of urls) {
    if (!url.startsWith(BASE)) continue;
    const pathPart = url.slice(BASE.length).split('?')[0];
    if (!pathPart || pathPart === '/' || /\.html$/i.test(pathPart)) continue;
    const segments = pathPart.split('/').filter(Boolean);
    if (!segments.length || !WHITELIST.has(segments[0])) continue;
    if (segments.length > 4) continue;
    if (seen.has(pathPart)) continue;
    seen.add(pathPart);
    out.push({ path: pathPart, url: BASE + pathPart, name: titleCase(segments[segments.length - 1]) });
  }
  out.sort((a, b) => a.path.localeCompare(b.path));
  return out;
}

function render(entries) {
  const head = [
    '---',
    'title: Yehwang Website Category Map (Site Map)',
    'category: site-map',
    `source: ${SITEMAP}`,
    'generated: true',
    'source_url:',
    `  - ${SITEMAP}`,
    `last_updated: ${new Date().toISOString().slice(0, 10)}`,
    '---',
    '',
    'Every Yehwang category page has a direct URL. Use these links to send customers straight to the right page instead of a search query.',
    '',
  ];
  const body = entries.map((e) => `### Q: Where can I find ${e.name}?\nBrowse our ${e.name} category: ${e.url}\n`);
  return head.join('\n') + '\n' + body.join('\n');
}

(async () => {
  const res = await fetch(SITEMAP, {
    headers: { 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0 Safari/537.36' },
    signal: AbortSignal.timeout(90000),
  });
  if (!res.ok) throw new Error(`sitemap HTTP ${res.status}`);
  const xml = await res.text();
  const urls = [...xml.matchAll(/<loc>([^<]+)<\/loc>/g)].map((m) => m[1].trim());
  const entries = buildEntries(urls);
  if (!entries.length) throw new Error('generated 0 site-map entries');
  fs.writeFileSync(OUT, render(entries), 'utf8');
  console.log(JSON.stringify({ ts: new Date().toISOString(), site_map_kb: 'done', entries: entries.length, out: OUT }));
})().catch((err) => { console.error('[build-site-kb]', err.message); process.exit(1); });