mb-grid-solutions.com/scripts/check-html.ts

import axios from "axios";
import * as cheerio from "cheerio";
import * as fs from "fs";
import * as path from "path";
import { execSync } from "child_process";

const targetUrl =
  process.argv[2] ||
  process.env.NEXT_PUBLIC_BASE_URL ||
  "http://localhost:3000";
const gatekeeperPassword = process.env.GATEKEEPER_PASSWORD || "mbgrid";

async function main() {
  console.log(`\n🚀 Starting HTML Validation for: ${targetUrl}`);
  console.log(`📊 Limit: None (Full Sitemap)\n`);

  try {
    const sitemapUrl = `${targetUrl.replace(/\/$/, "")}/sitemap.xml`;
    console.log(`📥 Fetching sitemap from ${sitemapUrl}...`);

    const response = await axios.get(sitemapUrl, {
      headers: { Cookie: `mintel_gatekeeper_session=${gatekeeperPassword}` },
      validateStatus: (status) => status < 400,
    });

    const $ = cheerio.load(response.data, { xmlMode: true });
    let urls = $("url loc")
      .map((i, el) => $(el).text())
      .get();

    const urlPattern = /https?:\/\/[^\/]+/;
    urls = [...new Set(urls)]
      .filter((u) => u.startsWith("http"))
      .map((u) => u.replace(urlPattern, targetUrl.replace(/\/$/, "")))
      .sort();

    console.log(`✅ Found ${urls.length} URLs in sitemap.`);

    if (urls.length === 0) {
      console.error("❌ No URLs found in sitemap. Is the site up?");
      process.exit(1);
    }

    const outputDir = path.join(process.cwd(), ".htmlvalidate-tmp");
    if (fs.existsSync(outputDir))
      fs.rmSync(outputDir, { recursive: true, force: true });
    fs.mkdirSync(outputDir, { recursive: true });

    console.log(`📥 Fetching HTML for ${urls.length} pages...`);
    for (let i = 0; i < urls.length; i++) {
      const u = urls[i];
      try {
        const res = await axios.get(u, {
          headers: {
            Cookie: `mintel_gatekeeper_session=${gatekeeperPassword}`,
          },
          validateStatus: (status) => status < 400,
        });

        // Generate a safe filename that retains URL information
        const urlStr = new URL(u);
        const safePath = (urlStr.pathname + urlStr.search).replace(
          /[^a-zA-Z0-9]/g,
          "_",
        );
        const filename = `${safePath || "index"}.html`;

        fs.writeFileSync(path.join(outputDir, filename), res.data);
      } catch (err: unknown) {
        const errorBody = err instanceof Error ? err.message : String(err);
        console.error(`❌ HTTP Error fetching ${u}: ${errorBody}`);
        throw new Error(`Failed to fetch page: ${u} - ${errorBody}`);
      }
    }

    console.log(`\n💻 Executing html-validate...`);
    try {
      execSync(`npx html-validate .htmlvalidate-tmp/*.html`, {
        stdio: "inherit",
      });
      console.log(`✅ HTML Validation passed perfectly!`);
    } catch {
      console.error(`❌ HTML Validation found issues.`);
      process.exit(1);
    }
  } catch (error: unknown) {
    const errorBody = error instanceof Error ? error.message : String(error);
    console.error(`\n❌ Error during HTML Validation:`, errorBody);
    process.exit(1);
  } finally {
    const outputDir = path.join(process.cwd(), ".htmlvalidate-tmp");
    if (fs.existsSync(outputDir))
      fs.rmSync(outputDir, { recursive: true, force: true });
  }
}

main();