WFDASH-WEBAPP/scripts/nss-scrape-bundle.js

// Run in the browser console on nssfasteners.co.za after login.
// Fetches all Grade 8 (101 items) + A2 (21 items) hex nut DIN934 product pages,
// parses the hidden unit prices out of the HTML, builds PosDescription strings,
// and triggers a CSV download.

(async () => {
  const BASE = 'https://nssfasteners.co.za/nuts/hex-nuts/hex-nut-din934.html';
  const FILTERS = [
    { grade: '8',  gradeId: '771', expectedCount: 101 },
    { grade: 'A2', gradeId: '773', expectedCount: 21  },
  ];

  // ---- helpers ----
  const fetchHtml = async (url) => {
    const res = await fetch(url, { credentials: 'include' });
    if (!res.ok) throw new Error(`HTTP ${res.status} for ${url}`);
    return await res.text();
  };

  const parsePage = (html) => {
    const doc = new DOMParser().parseFromString(html, 'text/html');
    const items = Array.from(doc.querySelectorAll('.product-item'));
    return items.map(item => {
      const titleEl = item.querySelector('.items-center.justify-center.text-primary');
      const title = titleEl ? titleEl.textContent.trim() : '';
      const sku = title.includes(' - ') ? title.split(' - ').pop().trim() : '';
      const linkEl = item.querySelector('a.product-item-photo, a.product.photo');
      const url = linkEl ? linkEl.getAttribute('href') : '';

      // pull the description block (Material / Diameter / Thread / Grade / Finish / Pack)
      const block = (item.textContent || '').replace(/\s+/g, ' ');
      const get = (label) => {
        const m = block.match(new RegExp(label + '\\s*:?\\s*([^\\s].*?)(?=\\s+(?:Diameter|Thread|Grade|Finish|Sold|Metric|Hi-Tensile|Stainless|Mild|Brass|$))', 'i'));
        return m ? m[1].trim() : '';
      };

      const material = /Hi-Tensile/i.test(block) ? 'Hi-Tensile'
                    : /Stainless Steel/i.test(block) ? 'Stainless Steel'
                    : /Mild Steel/i.test(block) ? 'Mild Steel'
                    : /Brass/i.test(block) ? 'Brass'
                    : '';
      const diameter = get('Diameter');
      const thread   = get('Thread');
      const grade    = get('Grade');
      const finish   = get('Finish');
      const pack     = /Sold as a Single Piece/i.test(block) ? 'Single Piece'
                     : /Sold as a Box/i.test(block)         ? 'Box'
                     : /Sold as a Pack/i.test(block)        ? 'Pack'
                     : '';

      // Unit price is in the HTML but hidden with display:none
      const unitPriceEl = item.querySelector('.unit-price');
      const priceMatch = unitPriceEl ? unitPriceEl.textContent.match(/R\s*([\d.]+)/) : null;
      const unitPrice = priceMatch ? priceMatch[1] : '';

      // Description is the parent product type (e.g. "Hex Nut DIN934")
      const desc = title.includes(' - ') ? title.split(' - ')[0].trim() : '';

      // Build PosDescription like: "M39 NUT 8 PLN"
      // diameter -> "M{value}", normalised
      const sizeMatch = diameter.match(/(\d+(?:\.\d+)?)\s*mm/i);
      const size = sizeMatch ? 'M' + sizeMatch[1].replace(/\.0$/, '') : diameter;
      const gradeShort = grade === 'Stainless' ? '' : grade;       // "8" or "A2"
      const posDescription = [size, 'NUT', gradeShort, finish.toUpperCase() === 'PLAIN' ? 'PLN' : finish.toUpperCase()]
        .filter(Boolean).join(' ');

      return {
        sku,
        posDescription,
        description: desc,
        material,
        grade,
        diameter,
        thread,
        finish,
        pack,
        unitPrice: unitPrice ? 'R ' + unitPrice : '',
        url,
      };
    });
  };

  const allRows = [];
  for (const f of FILTERS) {
    // First page
    const firstUrl = `${BASE}?gradematerial=${f.gradeId}`;
    const firstHtml = await fetchHtml(firstUrl);
    const firstRows = parsePage(firstHtml);
    allRows.push(...firstRows);

    // Discover total page count from toolbar
    const doc = new DOMParser().parseFromString(firstHtml, 'text/html');
    const m = (doc.querySelector('.toolbar-amount')?.textContent || '').match(/of\s+(\d+)/);
    const total = m ? parseInt(m[1], 10) : f.expectedCount;
    const pages = Math.ceil(total / 10);
    console.log(`Grade ${f.grade}: ${firstRows.length} items on page 1, ${pages} pages total, expected ${total}`);

    // Remaining pages
    for (let p = 2; p <= pages; p++) {
      const html = await fetchHtml(`${BASE}?gradematerial=${f.gradeId}&p=${p}`);
      const rows = parsePage(html);
      allRows.push(...rows);
      console.log(`Grade ${f.grade}: page ${p}/${pages} -> ${rows.length} items`);
    }
  }

  console.log(`TOTAL ROWS: ${allRows.length}`);

  // Build CSV
  const headers = ['SKU','PosDescription','Description','Material','Grade','Diameter','Thread','Finish','Pack','UnitPrice','SourceURL'];
  const esc = (v) => {
    const s = (v ?? '').toString().replace(/"/g, '""');
    return /[",\n]/.test(s) ? `"${s}"` : s;
  };
  const csv = [headers.join(',')]
    .concat(allRows.map(r => headers.map(h => esc(r[h.toLowerCase()])).join(',')))
    .join('\n');

  // Trigger download
  const blob = new Blob([csv], { type: 'text/csv;charset=utf-8' });
  const url = URL.createObjectURL(blob);
  const a = document.createElement('a');
  a.href = url;
  a.download = 'nss-hex-nut-din934-grade8-a2.csv';
  document.body.appendChild(a);
  a.click();
  document.body.removeChild(a);
  URL.revokeObjectURL(url);

  return { count: allRows.length, sample: allRows.slice(0, 3) };
})();