klz-cables.com/scripts/generate-pdf-datasheets.ts

#!/usr/bin/env ts-node
/**
 * PDF Datasheet Generator (React-PDF)
 *
 * Renders PDFs via `@react-pdf/renderer`.
 *
 * Source of truth:
 * - All technical data + cross-section tables: Excel files in `data/excel/`
 * - Product description text: MDX files in `data/products/{en,de}/*.mdx`
 */

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

import * as XLSX from 'xlsx';
const matter = require('gray-matter') as (src: string) => { data: unknown; content: string };

import type { ProductData } from './pdf/model/types';
import { generateDatasheetPdfBuffer } from './pdf/react-pdf/generate-datasheet-pdf';
import { generateFileName, normalizeValue, stripHtml } from './pdf/model/utils';

const CONFIG = {
  outputDir: path.join(process.cwd(), 'public/datasheets'),
  chunkSize: 10,
} as const;

const EXCEL_FILES = [
  path.join(process.cwd(), 'data/excel/high-voltage.xlsx'),
  path.join(process.cwd(), 'data/excel/medium-voltage-KM.xlsx'),
  path.join(process.cwd(), 'data/excel/medium-voltage-KM 170126.xlsx'),
  path.join(process.cwd(), 'data/excel/low-voltage-KM.xlsx'),
  path.join(process.cwd(), 'data/excel/solar-cables.xlsx'),
] as const;

type MdxProduct = {
  slug: string;
  title: string;
  sku: string;
  categories: string[];
  images: string[];
  descriptionHtml: string;
};

type MdxIndex = Map<string, MdxProduct>; // key: normalized designation/title

function ensureOutputDir(): void {
  if (!fs.existsSync(CONFIG.outputDir)) {
    fs.mkdirSync(CONFIG.outputDir, { recursive: true });
  }
}

function normalizeExcelKey(value: string): string {
  return String(value || '')
    .toUpperCase()
    .replace(/-\d+$/g, '')
    .replace(/[^A-Z0-9]+/g, '');
}

function extractDescriptionHtmlFromMdxBody(body: string): string {
  const content = String(body || '').trim();
  if (!content) return '';

  // MDX product files are wrapped like:
  // <ProductTabs technicalData={...}>
  //   <section>...</section>
  // </ProductTabs>
  // For PDF, we only want the inner description content.
  const withoutOpen = content.replace(/^\s*<ProductTabs[\s\S]*?>\s*/i, '');
  const withoutClose = withoutOpen.replace(/\s*<\/ProductTabs>\s*$/i, '');
  return withoutClose.trim();
}

function buildMdxIndex(locale: 'en' | 'de'): MdxIndex {
  const dir = path.join(process.cwd(), 'data/products', locale);
  const idx: MdxIndex = new Map();
  if (!fs.existsSync(dir)) return idx;

  const files = fs
    .readdirSync(dir)
    .filter(f => f.endsWith('.mdx'))
    .sort();

  for (const file of files) {
    const filePath = path.join(dir, file);
    const raw = fs.readFileSync(filePath, 'utf8');
    const parsed = matter(raw);
    const data = (parsed.data || {}) as any;

    const title = normalizeValue(String(data.title || ''));
    if (!title) continue;

    const sku = normalizeValue(String(data.sku || ''));
    const categories = Array.isArray(data.categories) ? data.categories.map((c: any) => normalizeValue(String(c))).filter(Boolean) : [];
    const images = Array.isArray(data.images) ? data.images.map((i: any) => normalizeValue(String(i))).filter(Boolean) : [];

    const descriptionHtml = extractDescriptionHtmlFromMdxBody(parsed.content);

    const slug = path.basename(file, '.mdx');
    idx.set(normalizeExcelKey(title), { slug, title, sku, categories, images, descriptionHtml });
  }

  return idx;
}

function findKeyByHeaderValue(headerRow: Record<string, unknown>, pattern: RegExp): string | null {
  for (const [k, v] of Object.entries(headerRow || {})) {
    const text = normalizeValue(String(v ?? ''));
    if (!text) continue;
    if (pattern.test(text)) return k;
  }
  return null;
}

function readExcelRows(filePath: string): Array<Record<string, unknown>> {
  if (!fs.existsSync(filePath)) return [];
  const workbook = XLSX.readFile(filePath, { cellDates: false, cellNF: false, cellText: false });
  const sheetName = workbook.SheetNames[0];
  if (!sheetName) return [];
  const sheet = workbook.Sheets[sheetName];
  if (!sheet) return [];

  return XLSX.utils.sheet_to_json(sheet, {
    defval: '',
    raw: false,
    blankrows: false,
  }) as Array<Record<string, unknown>>;
}

function readDesignationsFromExcelFile(filePath: string): Map<string, string> {
  const rows = readExcelRows(filePath);
  if (!rows.length) return new Map();

  // Legacy sheets use "Part Number" as a column key.
  // The new MV sheet uses __EMPTY* keys and stores the human headers in row 0 values.
  const headerRow = rows[0] || {};
  const partNumberKey =
    (Object.prototype.hasOwnProperty.call(headerRow, 'Part Number') ? 'Part Number' : null) ||
    findKeyByHeaderValue(headerRow, /^part\s*number$/i) ||
    '__EMPTY';

  const out = new Map<string, string>();
  for (const r of rows) {
    const pn = normalizeValue(String(r?.[partNumberKey] ?? ''));
    if (!pn || pn === 'Units' || pn === 'Part Number') continue;

    const key = normalizeExcelKey(pn);
    if (!key) continue;

    // Keep first-seen designation string (stable filenames from MDX slug).
    if (!out.has(key)) out.set(key, pn);
  }

  return out;
}

function loadAllExcelDesignations(): Map<string, string> {
  const out = new Map<string, string>();
  for (const filePath of EXCEL_FILES) {
    const m = readDesignationsFromExcelFile(filePath);
    Array.from(m.entries()).forEach(([k, v]) => {
      if (!out.has(k)) out.set(k, v);
    });
  }
  return out;
}

async function loadProductsFromExcelAndMdx(locale: 'en' | 'de'): Promise<ProductData[]> {
  const mdxIndex = buildMdxIndex(locale);
  const excelDesignations = loadAllExcelDesignations();

  const products: ProductData[] = [];
  let id = 1;

  Array.from(excelDesignations.entries()).forEach(([key, designation]) => {
    const mdx = mdxIndex.get(key) || null;

    const title = mdx?.title || designation;
    const slug =
      mdx?.slug ||
      title
        .toLowerCase()
        .replace(/[^a-z0-9]+/g, '-')
        .replace(/-+/g, '-')
        .replace(/^-|-$/g, '');

    // Only the product description comes from MDX. Everything else is Excel-driven
    // during model building (technicalItems + voltage tables).
    const descriptionHtml = mdx?.descriptionHtml || '';

    products.push({
      id: id++,
      name: title,
      shortDescriptionHtml: '',
      descriptionHtml,
      images: mdx?.images || [],
      featuredImage: (mdx?.images && mdx.images[0]) || null,
      sku: mdx?.sku || title,
      slug,
      translationKey: slug,
      locale,
      categories: (mdx?.categories || []).map(name => ({ name })),
      attributes: [],
    });
  });

  // Deterministic order: by slug, then name.
  products.sort((a, b) => (a.slug || '').localeCompare(b.slug || '') || a.name.localeCompare(b.name));

  // Drop products that have no readable name.
  return products.filter(p => stripHtml(p.name));
}

async function processChunk(products: ProductData[], chunkIndex: number, totalChunks: number): Promise<void> {
  console.log(`\nProcessing chunk ${chunkIndex + 1}/${totalChunks} (${products.length} products)...`);

  for (const product of products) {
    try {
      const locale = (product.locale || 'en') as 'en' | 'de';
      const buffer = await generateDatasheetPdfBuffer({ product, locale });
      const fileName = generateFileName(product, locale);
      fs.writeFileSync(path.join(CONFIG.outputDir, fileName), buffer);
      console.log(`✓ ${locale.toUpperCase()}: ${fileName}`);
      await new Promise(resolve => setTimeout(resolve, 25));
    } catch (error) {
      console.error(`✗ Failed to process product ${product.id}:`, error);
    }
  }
}

async function processProductsInChunks(): Promise<void> {
  console.log('Starting PDF generation (React-PDF)');
  ensureOutputDir();

  const onlyLocale = normalizeValue(String(process.env.PDF_LOCALE || '')).toLowerCase();
  const locales: Array<'en' | 'de'> = onlyLocale === 'de' || onlyLocale === 'en' ? [onlyLocale] : ['en', 'de'];

  const allProducts: ProductData[] = [];
  for (const locale of locales) {
    const products = await loadProductsFromExcelAndMdx(locale);
    allProducts.push(...products);
  }

  if (allProducts.length === 0) {
    console.log('No products found');
    return;
  }

  // Dev convenience: generate only one product subset.
  // IMPORTANT: apply filters BEFORE PDF_LIMIT so the limit works within the filtered set.
  let products = allProducts;

  const match = normalizeValue(String(process.env.PDF_MATCH || '')).toLowerCase();
  if (match) {
    products = products.filter(p => {
      const hay = [p.slug, p.translationKey, p.sku, p.name]
        .filter(Boolean)
        .join(' ')
        .toLowerCase();
      return hay.includes(match);
    });
  }

  const limit = Number(process.env.PDF_LIMIT || '0');
  products = Number.isFinite(limit) && limit > 0 ? products.slice(0, limit) : products;

  const enProducts = products.filter(p => (p.locale || 'en') === 'en');
  const deProducts = products.filter(p => (p.locale || 'en') === 'de');
  console.log(`Found ${enProducts.length} EN + ${deProducts.length} DE products`);

  const totalChunks = Math.ceil(products.length / CONFIG.chunkSize);
  for (let i = 0; i < totalChunks; i++) {
    const chunk = products.slice(i * CONFIG.chunkSize, (i + 1) * CONFIG.chunkSize);
    await processChunk(chunk, i, totalChunks);
  }

  console.log('\n✅ PDF generation completed!');
  console.log(`Generated ${enProducts.length} EN + ${deProducts.length} DE PDFs`);
  console.log(`Output: ${CONFIG.outputDir}`);
}

async function main(): Promise<void> {
  const start = Date.now();
  try {
    await processProductsInChunks();
    console.log(`\nTime: ${((Date.now() - start) / 1000).toFixed(2)}s`);
  } catch (error) {
    console.error('Fatal error:', error);
    process.exit(1);
  }
}

main().catch(console.error);

export { main as generatePDFDatasheets };