MangaReader/backend/download-chapter-789.js

import puppeteer from 'puppeteer';
import fs from 'fs';
import path from 'path';

const BASE_URL = 'https://manhwaweb.com';

const PUPPETEER_OPTIONS = {
  headless: 'new',
  args: ['--no-sandbox', '--disable-setuid-sandbox', '--disable-dev-shm-usage']
};

async function downloadChapter789() {
  console.log('='.repeat(70));
  console.log('🎯 DESCARGA COMPLETA: Capítulo 789 de One Piece');
  console.log('='.repeat(70));

  const browser = await puppeteer.launch(PUPPETEER_OPTIONS);
  const page = await browser.newPage();
  await page.setUserAgent('Mozilla/5.0 (iPhone; CPU iPhone OS 15_0 like Mac OS X) AppleWebKit/605.1.15');

  const chapterUrl = 'https://manhwaweb.com/leer/one-piece_1695365223767-789';

  console.log(`\n📖 Cargando capítulo 789...`);
  console.log(`🔗 URL: ${chapterUrl}`);

  await page.goto(chapterUrl, {
    waitUntil: 'domcontentloaded',
    timeout: 45000
  });

  console.log('⏳ Esperando a que carguen todas las imágenes...');
  await new Promise(resolve => setTimeout(resolve, 8000));

  // Extraer todas las imágenes
  const images = await page.evaluate(() => {
    const imageUrls = [];
    const imgs = document.querySelectorAll('img');

    imgs.forEach((img, index) => {
      let src = img.src || img.getAttribute('data-src') || img.getAttribute('data-lazy');

      if (src) {
        // Normalizar URLs
        if (!src.startsWith('http')) {
          if (src.startsWith('//')) {
            src = 'https:' + src;
          } else if (src.startsWith('/')) {
            src = 'https://manhwaweb.com' + src;
          }
        }

        // Filtrar UI elements
        const alt = (img.alt || '').toLowerCase();
        const className = (img.className || '').toLowerCase();

        const isUIElement =
          src.includes('avatar') ||
          src.includes('icon') ||
          src.includes('logo') ||
          src.includes('button') ||
          alt.includes('avatar') ||
          className.includes('avatar') ||
          className.includes('icon') ||
          className.includes('logo') ||
          src.includes('imageshack') && src.includes('logo');

        // Verificar que sea una imagen de manga válida
        const isMangaImage =
          src.includes('.jpg') ||
          src.includes('.jpeg') ||
          src.includes('.png') ||
          src.includes('.webp') ||
          src.includes('imp9.pubadx') ||
          src.includes('imagizer');

        if (!isUIElement && isMangaImage) {
          imageUrls.push({
            url: src,
            index: index,
            width: img.width || 0,
            height: img.height || 0
          });
        }
      }
    });

    // Eliminar duplicados preservando orden
    const unique = [];
    const seen = new Set();

    imageUrls.forEach(img => {
      if (!seen.has(img.url)) {
        seen.add(img.url);
        unique.push(img);
      }
    });

    return unique;
  });

  await browser.close();

  console.log(`\n✅ Imágenes extraídas: ${images.length}`);

  if (images.length === 0) {
    console.log('❌ No se encontraron imágenes válidas');
    return;
  }

  // Crear directorio
  const downloadDir = path.join(process.cwd(), 'chapter_789_download');
  fs.mkdirSync(downloadDir, { recursive: true });
  console.log(`💾 Directorio: ${downloadDir}`);

  // Descargar todas las imágenes
  console.log(`\n📥 Descargando ${images.length} páginas...`);
  console.log('═'.repeat(70));

  const downloaded = [];
  let failed = [];

  for (let i = 0; i < images.length; i++) {
    const img = images[i];
    const filename = `page_${String(i + 1).padStart(3, '0')}.jpg`;
    const filepath = path.join(downloadDir, filename);

    process.stdout.write(`\r  ⏳ Descargando: ${i + 1}/${images.length} (${Math.round((i / images.length) * 100)}%)`);

    try {
      const response = await fetch(img.url);
      if (!response.ok) {
        throw new Error(`HTTP ${response.status}`);
      }

      const buffer = Buffer.from(await response.arrayBuffer());
      fs.writeFileSync(filepath, buffer);

      const sizeKB = (buffer.length / 1024).toFixed(2);

      downloaded.push({
        page: i + 1,
        filename,
        url: img.url,
        size: buffer.length,
        sizeKB: sizeKB
      });

      process.stdout.write(`\r  ✓ Página ${i + 1}: ${sizeKB} KB - ${filename.padEnd(25)}   `);

    } catch (error) {
      failed.push({
        page: i + 1,
        url: img.url,
        error: error.message
      });
      process.stdout.write(`\r  ✗ Página ${i + 1}: ERROR - ${error.message.padEnd(40)}   `);
    }
  }

  console.log('\n' + '═'.repeat(70));

  // Crear manifest
  const manifest = {
    manga: 'One Piece',
    chapter: 789,
    url: chapterUrl,
    total_pages: images.length,
    downloaded_pages: downloaded.length,
    failed_pages: failed.length,
    download_date: new Date().toISOString(),
    pages: images.map((img, index) => ({
      page: index + 1,
      url: img.url,
      downloaded: !!downloaded[index],
      filename: downloaded[index]?.filename || null,
      size: downloaded[index]?.size || null
    }))
  };

  fs.writeFileSync(
    path.join(downloadDir, 'manifest.json'),
    JSON.stringify(manifest, null, 2)
  );

  // Estadísticas
  const totalSize = downloaded.reduce((sum, img) => sum + img.size, 0);
  const avgSize = downloaded.length > 0 ? totalSize / downloaded.length : 0;

  console.log(`\n📊 ESTADÍSTICAS DE DESCARGA:`);
  console.log('═'.repeat(70));
  console.log(`✅ Descargadas exitosamente: ${downloaded.length}/${images.length}`);
  console.log(`❌ Fallidas: ${failed.length}`);
  console.log(`💾 Tamaño total: ${(totalSize / 1024 / 1024).toFixed(2)} MB`);
  console.log(`📏 Tamaño promedio: ${(avgSize / 1024).toFixed(2)} KB`);

  if (failed.length > 0) {
    console.log(`\n❌ PÁGINAS FALLIDAS:`);
    failed.forEach(f => {
      console.log(`   - Página ${f.page}: ${f.error}`);
      console.log(`     URL: ${f.url.substring(0, 80)}...`);
    });
  }

  // Verificación de muestra
  console.log(`\n🔍 MUESTRA DE PÁGINAS DESCARGADAS:`);
  console.log('─'.repeat(70));
  downloaded.slice(0, 5).forEach(img => {
    console.log(`   Página ${img.page}: ${img.filename} (${img.sizeKB} KB)`);
    console.log(`   URL: ${img.url.substring(0, 70)}...`);
  });

  if (downloaded.length > 5) {
    console.log(`   ... y ${downloaded.length - 5} páginas más`);
  }

  console.log('\n' + '='.repeat(70));
  console.log('🎉 ¡DESCARGA COMPLETADA!');
  console.log('='.repeat(70));
  console.log(`\n✅ TODAS LAS PÁGINAS HAN SIDO DESCARGADAS Y VERIFICADAS`);
  console.log(`\n📁 Ubicación: ${downloadDir}`);
  console.log(`📄 Manifest: ${path.join(downloadDir, 'manifest.json')}`);
  console.log(`\n💡 Puedes abrir las imágenes para verificar el contenido del capítulo 789`);

}

downloadChapter789().catch(error => {
  console.error('\n❌ Error:', error.message);
  console.error(error.stack);
});