const express = require('express');
const puppeteer = require('puppeteer');
const cheerio = require('cheerio');
const fs = require('fs');
const path = require('path');
const axios = require('axios');
const { URL } = require('url');

const app = express();
app.use(express.json());

class AdvancedWebCloner {
  constructor() {
    this.baseOutputDir = path.join(__dirname, '..', 'cloned-sites');
    this.ensureDirectoryExists(this.baseOutputDir);
    this.visitedPages = new Set();
    this.downloadedAssets = new Map();
    this.pageQueue = [];
  }

  ensureDirectoryExists(dirPath) {
    if (!fs.existsSync(dirPath)) {
      fs.mkdirSync(dirPath, { recursive: true });
    }
  }

  sanitizeFilename(filename) {
    return filename.replace(/[<>:"/\\|?*]/g, '_').replace(/\s+/g, '_').substring(0, 200);
  }

  getUniqueFilename(urlPath, existingFiles = new Set()) {
    const parts = urlPath.split('/').filter(p => p);
    const filename = parts.pop() || 'file';
    const prefix = parts.slice(-3).join('_');
    let baseName = this.sanitizeFilename(prefix ? `${prefix}_${filename}` : filename);
    
    // Garantir unicidade
    let finalName = baseName;
    let counter = 1;
    while (existingFiles.has(finalName)) {
      const ext = path.extname(baseName);
      const name = baseName.replace(ext, '');
      finalName = `${name}_${counter}${ext}`;
      counter++;
    }
    
    return finalName;
  }

  async downloadAsset(url, outputPath, retries = 3) {
    for (let attempt = 1; attempt <= retries; attempt++) {
      try {
        // Para Google Fonts CSS, precisamos de user-agent específico para obter woff2
        const isGoogleFonts = url.includes('fonts.googleapis.com');
        
        const response = await axios.get(url, {
          responseType: 'arraybuffer',
          timeout: 30000,
          maxContentLength: 50 * 1024 * 1024, // 50MB
          headers: {
            'User-Agent': isGoogleFonts 
              ? 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36'
              : 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
            'Accept': '*/*',
            'Accept-Encoding': 'gzip, deflate, br',
            'Connection': 'keep-alive',
            'Referer': isGoogleFonts ? 'https://fonts.googleapis.com/' : url
          }
        });
        
        this.ensureDirectoryExists(path.dirname(outputPath));
        fs.writeFileSync(outputPath, response.data);
        return true;
      } catch (error) {
        if (attempt === retries) {
          // Não logar erro para fontes - muito verboso
          if (!url.includes('.woff') && !url.includes('.ttf') && !url.includes('.eot')) {
            console.log(`    ⚠️ Falha após ${retries} tentativas: ${url.substring(0, 80)}`);
          }
          return false;
        }
        await new Promise(resolve => setTimeout(resolve, 1000 * attempt));
      }
    }
    return false;
  }

  getAssetInfo(assetUrl, baseUrl, siteDir, existingFiles) {
    try {
      const fullUrl = new URL(assetUrl, baseUrl);
      const pathname = fullUrl.pathname;
      
      // Preservar a estrutura de pastas original
      // Remove a barra inicial e mantém o caminho completo
      let relativePath = pathname.startsWith('/') ? pathname.substring(1) : pathname;
      
      // Se o caminho estiver vazio, gerar um nome
      if (!relativePath || relativePath === '') {
        relativePath = 'index_' + Date.now();
      }
      
      // Sanitizar apenas caracteres inválidos, mas manter a estrutura de pastas
      relativePath = relativePath.replace(/[<>:"|?*]/g, '_');
      
      const fullPath = path.join(siteDir, relativePath);
      
      // Criar diretórios necessários
      this.ensureDirectoryExists(path.dirname(fullPath));
      
      return {
        fullPath: fullPath,
        relativePath: relativePath,
        fullUrl: fullUrl.href
      };
    } catch (error) {
      return null;
    }
  }

  extractUrlsFromCSS(cssContent, cssUrl) {
    const urls = [];
    
    // Regex mais abrangente para capturar URLs em CSS
    const patterns = [
      /url\(['"]?([^'")\s]+)['"]?\)/g,  // url(...)
      /@import\s+['"]([^'"]+)['"]/g,     // @import "..."
      /@import\s+url\(['"]?([^'")\s]+)['"]?\)/g  // @import url(...)
    ];
    
    patterns.forEach(regex => {
      let match;
      while ((match = regex.exec(cssContent)) !== null) {
        const url = match[1];
        if (url && !url.startsWith('data:') && !url.startsWith('#')) {
          try {
            const fullUrl = new URL(url, cssUrl).href;
            urls.push({ original: url, full: fullUrl });
          } catch(e) {
            // Tentar como caminho relativo
            try {
              const baseUrl = new URL(cssUrl);
              const resolvedUrl = new URL(url, baseUrl.origin + baseUrl.pathname.substring(0, baseUrl.pathname.lastIndexOf('/') + 1)).href;
              urls.push({ original: url, full: resolvedUrl });
            } catch(e2) {}
          }
        }
      }
    });
    
    return urls;
  }

  async extractAllAssets(page, pageUrl, $) {
    const assets = new Map();
    
    // 1. CSS - todos os tipos
    $('link[rel="stylesheet"], link[rel="preload"][as="style"]').each((_, el) => {
      const href = $(el).attr('href');
      if (href && !href.startsWith('data:')) {
        try { assets.set(href, new URL(href, pageUrl).href); } catch(e) {}
      }
    });
    
    // 2. JavaScript - todos os tipos
    $('script[src]').each((_, el) => {
      const src = $(el).attr('src');
      if (src && !src.startsWith('data:')) {
        try { assets.set(src, new URL(src, pageUrl).href); } catch(e) {}
      }
    });
    
    // 3. Imagens - múltiplos atributos
    $('img, picture source, [data-src], [data-lazy-src], [data-original]').each((_, el) => {
      const attrs = ['src', 'data-src', 'data-lazy-src', 'data-original', 'data-srcset', 'srcset'];
      attrs.forEach(attr => {
        const value = $(el).attr(attr);
        if (value && !value.startsWith('data:')) {
          // Processar srcset
          if (attr.includes('srcset')) {
            value.split(',').forEach(item => {
              const url = item.trim().split(' ')[0];
              if (url && !url.startsWith('data:')) {
                try { assets.set(url, new URL(url, pageUrl).href); } catch(e) {}
              }
            });
          } else {
            try { assets.set(value, new URL(value, pageUrl).href); } catch(e) {}
          }
        }
      });
    });
    
    // 4. Background images - style attributes e data attributes
    $('[style*="url"], [data-background], [data-bg], [data-background-image]').each((_, el) => {
      const attrs = ['style', 'data-background', 'data-bg', 'data-background-image'];
      attrs.forEach(attr => {
        const value = $(el).attr(attr) || '';
        const matches = value.match(/url\(['"]?([^'")\s]+)['"]?\)/g);
        if (matches) {
          matches.forEach(match => {
            const url = match.replace(/url\(['"]?/, '').replace(/['"]?\)/, '');
            if (url && !url.startsWith('data:')) {
              try { assets.set(url, new URL(url, pageUrl).href); } catch(e) {}
            }
          });
        }
      });
    });
    
    // 5. Vídeos e áudio
    $('video, audio, source, video[poster]').each((_, el) => {
      const attrs = ['src', 'poster'];
      attrs.forEach(attr => {
        const value = $(el).attr(attr);
        if (value && !value.startsWith('data:')) {
          try { assets.set(value, new URL(value, pageUrl).href); } catch(e) {}
        }
      });
    });
    
    // 6. Favicon e icons
    $('link[rel*="icon"], link[rel="apple-touch-icon"]').each((_, el) => {
      const href = $(el).attr('href');
      if (href && !href.startsWith('data:')) {
        try { assets.set(href, new URL(href, pageUrl).href); } catch(e) {}
      }
    });
    
    // 7. Preload e prefetch
    $('link[rel="preload"], link[rel="prefetch"]').each((_, el) => {
      const href = $(el).attr('href');
      if (href && !href.startsWith('data:')) {
        try { assets.set(href, new URL(href, pageUrl).href); } catch(e) {}
      }
    });
    
    // 8. Fonts do Google Fonts - baixar o CSS e as fontes
    const googleFontsLinks = [];
    $('link[href*="fonts.googleapis.com"]').each((_, el) => {
      const href = $(el).attr('href');
      if (href) {
        googleFontsLinks.push(href);
        try { assets.set(href, new URL(href, pageUrl).href); } catch(e) {}
      }
    });
    
    // 9. Fonts locais e de outros CDNs
    $('link[href*="fonts.gstatic.com"], link[href*="font"]').each((_, el) => {
      const href = $(el).attr('href');
      if (href) {
        try { assets.set(href, new URL(href, pageUrl).href); } catch(e) {}
      }
    });
    
    // 9. Extrair assets de inline styles
    const inlineStyles = await page.evaluate(() => {
      const styles = [];
      document.querySelectorAll('style').forEach(style => {
        styles.push(style.textContent);
      });
      return styles;
    });
    
    inlineStyles.forEach(styleContent => {
      const matches = styleContent.match(/url\(['"]?([^'")\s]+)['"]?\)/g);
      if (matches) {
        matches.forEach(match => {
          const url = match.replace(/url\(['"]?/, '').replace(/['"]?\)/, '');
          if (url && !url.startsWith('data:') && !url.startsWith('#')) {
            try { assets.set(url, new URL(url, pageUrl).href); } catch(e) {}
          }
        });
      }
    });
    
    return assets;
  }

  async discoverInternalLinks(page, baseHostClean, startUrl) {
    return await page.evaluate((baseHostClean, startUrl) => {
      const links = new Set();
      const baseUrl = new URL(startUrl);
      
      document.querySelectorAll('a[href]').forEach(a => {
        try {
          const href = a.href;
          const url = new URL(href);
          
          // Verificar se é link interno
          if (url.hostname.includes(baseHostClean) || url.hostname === baseUrl.hostname) {
            // Filtrar extensões de arquivo
            if (!url.pathname.match(/\.(css|js|png|jpg|jpeg|gif|svg|webp|ico|pdf|zip|doc|docx|xls|xlsx|mp4|mp3|avi)$/i)) {
              // Remover query strings e fragments para normalizar
              const cleanUrl = `${url.origin}${url.pathname}`.replace(/\/$/, '');
              if (cleanUrl && 
                  !cleanUrl.includes('mailto:') && 
                  !cleanUrl.includes('tel:') &&
                  !cleanUrl.includes('javascript:')) {
                links.add(cleanUrl);
              }
            }
          }
        } catch(e) {}
      });
      
      return Array.from(links);
    }, baseHostClean, startUrl);
  }

  async cloneWebsite(startUrl) {
    const browser = await puppeteer.launch({
      headless: 'new',
      args: [
        '--no-sandbox',
        '--disable-setuid-sandbox',
        '--disable-dev-shm-usage',
        '--disable-accelerated-2d-canvas',
        '--disable-gpu'
      ]
    });
    
    try {
      const page = await browser.newPage();
      
      // Configurar viewport grande para capturar mais conteúdo
      await page.setViewport({ width: 1920, height: 1080 });
      
      // Configurar user agent realista
      await page.setUserAgent('Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36');
      
      console.log(`\n🚀 INICIANDO CLONAGEM PROFUNDA DE: ${startUrl}\n`);
      
      const parsedUrl = new URL(startUrl);
      const baseHost = parsedUrl.hostname;
      const baseHostClean = baseHost.replace('www.', '');
      
      // Criar diretório
      const siteName = this.sanitizeFilename(baseHost);
      const timestamp = new Date().toISOString().replace(/[:.]/g, '-');
      const siteDir = path.join(this.baseOutputDir, `${siteName}_${timestamp}`);
      this.ensureDirectoryExists(siteDir);
      
      console.log(`📁 Diretório: ${siteDir}\n`);
      
      // Fase 1: Descobrir todas as páginas
      console.log(`🔍 FASE 1: Descobrindo páginas do site...`);
      
      const response = await page.goto(startUrl, {
        waitUntil: 'networkidle0',
        timeout: 60000
      });
      
      await page.waitForTimeout(3000);
      
      // Descobrir links internos
      const internalLinks = await this.discoverInternalLinks(page, baseHostClean, startUrl);
      
      // Criar mapa de páginas
      const pageMap = new Map();
      const existingFiles = new Set();
      
      // Adicionar página inicial
      pageMap.set(startUrl, { file: 'index.html', title: 'Home' });
      existingFiles.add('index.html');
      
      // Adicionar páginas descobertas
      for (const link of internalLinks) {
        if (!pageMap.has(link)) {
          const urlObj = new URL(link);
          let pathname = urlObj.pathname.replace(/^\/|\/$/g, '');
          
          let filename;
          if (!pathname) {
            filename = 'home.html';
          } else {
            filename = pathname.replace(/\//g, '_') + '.html';
          }
          
          // Garantir unicidade
          let finalFilename = filename;
          let counter = 1;
          while (existingFiles.has(finalFilename)) {
            finalFilename = filename.replace('.html', `_${counter}.html`);
            counter++;
          }
          
          pageMap.set(link, { file: finalFilename, title: pathname || 'page' });
          existingFiles.add(finalFilename);
        }
      }
      
      console.log(`✅ Encontradas ${pageMap.size} páginas para clonar\n`);
      
      // Fase 2: Clonar cada página
      console.log(`📄 FASE 2: Clonando páginas e coletando assets...`);
      
      const globalAssets = new Map();
      const pageContents = new Map();
      const assetFiles = new Set();
      
      let pageIndex = 0;
      for (const [pageUrl, pageInfo] of pageMap) {
        pageIndex++;
        console.log(`\n  [${pageIndex}/${pageMap.size}] ${pageInfo.file}`);
        
        try {
          await page.goto(pageUrl, {
            waitUntil: 'networkidle0',
            timeout: 60000
          });
          
          // Esperar carregamento completo
          await page.waitForTimeout(3000);
          
          // Scroll completo para carregar lazy content
          console.log(`    📜 Fazendo scroll para carregar conteúdo lazy...`);
          await page.evaluate(async () => {
            await new Promise((resolve) => {
              let totalHeight = 0;
              const distance = 300;
              const timer = setInterval(() => {
                const scrollHeight = document.body.scrollHeight;
                window.scrollBy(0, distance);
                totalHeight += distance;
                
                if (totalHeight >= scrollHeight) {
                  clearInterval(timer);
                  window.scrollTo(0, 0);
                  setTimeout(resolve, 1000);
                }
              }, 100);
            });
          });
          
          await page.waitForTimeout(2000);
          
          // Capturar conteúdo
          const content = await page.content();
          pageContents.set(pageUrl, content);
          
          const $ = cheerio.load(content);
          
          // Extrair todos os assets
          console.log(`    🔍 Extraindo assets...`);
          const pageAssets = await this.extractAllAssets(page, pageUrl, $);
          
          // Adicionar ao conjunto global
          for (const [original, full] of pageAssets) {
            globalAssets.set(original, full);
          }
          
          console.log(`    ✓ ${pageAssets.size} assets encontrados`);
          
        } catch (err) {
          console.log(`    ❌ Erro: ${err.message}`);
        }
      }
      
      console.log(`\n✅ Total de ${globalAssets.size} assets únicos encontrados\n`);
      
      // Fase 3: Baixar todos os assets
      console.log(`📥 FASE 3: Baixando assets...`);
      
      const replacements = new Map();
      const batchSize = 5; // Reduzido para evitar sobrecarga
      const assetEntries = [...globalAssets.entries()];
      let successCount = 0;
      let failCount = 0;
      
      for (let i = 0; i < assetEntries.length; i += batchSize) {
        const batch = assetEntries.slice(i, i + batchSize);
        const promises = batch.map(async ([originalUrl, fullUrl]) => {
          const info = this.getAssetInfo(fullUrl, startUrl, siteDir, assetFiles);
          if (info) {
            const ok = await this.downloadAsset(info.fullUrl, info.fullPath);
            if (ok) {
              replacements.set(originalUrl, info.relativePath);
              replacements.set(info.fullUrl, info.relativePath);
              
              // Variações de protocolo
              if (info.fullUrl.startsWith('https://')) {
                replacements.set(info.fullUrl.replace('https://', 'http://'), info.relativePath);
              } else if (info.fullUrl.startsWith('http://')) {
                replacements.set(info.fullUrl.replace('http://', 'https://'), info.relativePath);
              }
              
              successCount++;
              return { ok: true, info, fullUrl };
            } else {
              failCount++;
            }
          }
          return { ok: false };
        });
        
        await Promise.all(promises);
        
        const progress = Math.round((i + batch.length) / assetEntries.length * 100);
        console.log(`  📊 Progresso: ${progress}% (${successCount} ok, ${failCount} falhas)`);
      }
      
      console.log(`\n✅ ${successCount} assets baixados com sucesso\n`);
      
      // Fase 4: Extrair assets de CSS
      console.log(`🎨 FASE 4: Extraindo assets de arquivos CSS...`);
      
      const cssAssets = new Map();
      let cssAssetCount = 0;
      
      for (const [original, local] of replacements) {
        if (local.endsWith('.css')) {
          try {
            const cssPath = path.join(siteDir, local);
            if (fs.existsSync(cssPath)) {
              const cssContent = fs.readFileSync(cssPath, 'utf8');
              
              // Determinar URL base do CSS
              let cssBaseUrl = original;
              for (const [origUrl, localPath] of replacements) {
                if (localPath === local) {
                  cssBaseUrl = origUrl;
                  break;
                }
              }
              
              const cssUrls = this.extractUrlsFromCSS(cssContent, cssBaseUrl);
              
              for (const cssUrl of cssUrls) {
                if (!replacements.has(cssUrl.full) && !cssAssets.has(cssUrl.full)) {
                  cssAssets.set(cssUrl.original, cssUrl.full);
                }
              }
            }
          } catch(e) {}
        }
      }
      
      if (cssAssets.size > 0) {
        console.log(`  📦 Encontrados ${cssAssets.size} assets em CSS (fontes, imagens, etc.)`);
        
        for (const [originalUrl, fullUrl] of cssAssets) {
          const info = this.getAssetInfo(fullUrl, startUrl, siteDir, assetFiles);
          if (info) {
            const ok = await this.downloadAsset(info.fullUrl, info.fullPath);
            if (ok) {
              replacements.set(originalUrl, info.relativePath);
              replacements.set(info.fullUrl, info.relativePath);
              cssAssetCount++;
              
              // Atualizar CSS files para substituir URLs
              for (const [origCss, localCss] of replacements) {
                if (localCss.endsWith('.css')) {
                  try {
                    const cssPath = path.join(siteDir, localCss);
                    if (fs.existsSync(cssPath)) {
                      let cssContent = fs.readFileSync(cssPath, 'utf8');
                      const escaped = originalUrl.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
                      cssContent = cssContent.replace(new RegExp(escaped, 'g'), info.relativePath);
                      fs.writeFileSync(cssPath, cssContent, 'utf8');
                    }
                  } catch(e) {}
                }
              }
            }
          }
        }
        
        console.log(`  ✅ ${cssAssetCount} assets de CSS baixados`);
      } else {
        console.log(`  ℹ️ Nenhum asset adicional encontrado em CSS`);
      }
      
      // Fase 5: Processar e salvar páginas
      console.log(`\n💾 FASE 5: Processando e salvando páginas HTML...\n`);
      
      const sortedReplacements = [...replacements.entries()]
        .sort((a, b) => b[0].length - a[0].length);
      
      const elementsToRemove = [
        '.cookie-banner', '.cookie-notice', '.cookie-consent', '.cookies-banner',
        '.lgpd-banner', '.lgpd-notice', '.lgpd-popup', '.lgpd',
        '#cookie-banner', '#cookie-notice', '#cookie-consent', '#cookies',
        '#lgpd-banner', '#lgpd', '#lgpd-notice',
        '[class*="cookie"]', '[class*="lgpd"]', '[class*="gdpr"]',
        '[id*="cookie"]', '[id*="lgpd"]', '[id*="gdpr"]',
        '.popup-overlay', '.modal-overlay', '.consent-banner',
        '[class*="consent"]', '[class*="privacy-banner"]',
        '.joinchat', '#joinchat', '.joinchat__button',
      ];
      
      let pagesSaved = 0;
      for (const [pageUrl, pageInfo] of pageMap) {
        let html = pageContents.get(pageUrl);
        if (!html) continue;
        
        console.log(`  💾 ${pageInfo.file}`);
        
        // Remover elementos indesejados
        const $ = cheerio.load(html);
        
        for (const selector of elementsToRemove) {
          try { $(selector).remove(); } catch(e) {}
        }
        
        // Remover scripts problemáticos e externos que não funcionam offline
        $('script').each((_, el) => {
          const src = $(el).attr('src') || '';
          const content = $(el).html() || '';
          
          // Lista de scripts a remover
          const removePatterns = [
            'cookie', 'lgpd', 'consent', 'gdpr',
            'recaptcha', 'google-analytics', 'googletagmanager', 'gtag', 'gtm',
            'facebook.net', 'fbevents', 'fbq',
            'doubleclick.net', 'googleadservices',
            'analytics.js', 'ga.js'
          ];
          
          const shouldRemove = removePatterns.some(pattern => 
            src.includes(pattern) || content.includes(pattern)
          );
          
          if (shouldRemove && !src.includes('jquery') && !src.includes('elementor')) {
            $(el).remove();
          }
        });
        
        // Remover iframes externos (Google Tag Manager, etc)
        $('iframe[src*="googletagmanager"], iframe[src*="facebook"], iframe[src*="doubleclick"]').remove();
        
        // Remover noscript tags de tracking
        $('noscript').each((_, el) => {
          const content = $(el).html() || '';
          if (content.includes('googletagmanager') || content.includes('facebook')) {
            $(el).remove();
          }
        });
        
        html = $.html();
        
        // Substituir assets - usar substituição mais precisa
        for (const [original, local] of sortedReplacements) {
          // Escapar caracteres especiais
          const escaped = original.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
          
          // Substituir em diferentes contextos
          // 1. Em atributos HTML (src, href, data-src, etc)
          html = html.replace(new RegExp(`(src|href|data-src|data-lazy-src|data-background|poster)=["']${escaped}["']`, 'gi'), `$1="${local}"`);
          
          // 2. Em CSS inline (background-image, etc)
          html = html.replace(new RegExp(`url\\(['"]?${escaped}['"]?\\)`, 'gi'), `url("${local}")`);
          
          // 3. Substituição direta
          html = html.replace(new RegExp(escaped, 'g'), local);
        }
        
        // Substituir links internos
        for (const [url, info] of pageMap) {
          const escaped = url.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
          html = html.replace(new RegExp(`href=["']${escaped}["']`, 'gi'), `href="${info.file}"`);
          
          const urlObj = new URL(url);
          if (urlObj.pathname && urlObj.pathname !== '/') {
            const pathname = urlObj.pathname.replace(/\/$/, '');
            html = html.replace(
              new RegExp(`href=["']${pathname.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}["']`, 'gi'),
              `href="${info.file}"`
            );
          }
        }
        
        // Links restantes do domínio -> index.html
        html = html.replace(
          new RegExp(`href=["']https?://(www\\.)?${baseHostClean.replace(/\./g, '\\.')}[^"']*["']`, 'gi'),
          'href="index.html"'
        );
        
        // Adicionar script para suprimir erros e melhorar compatibilidade offline
        const offlineScript = `
<script>
// Suprimir erros de scripts externos
window.addEventListener('error', function(e) {
  // Ignorar erros de recursos externos
  if (e.filename && (
    e.filename.includes('recaptcha') ||
    e.filename.includes('google') ||
    e.filename.includes('facebook') ||
    e.filename.includes('doubleclick')
  )) {
    e.preventDefault();
    return true;
  }
}, true);

// Criar stubs para funções que podem estar faltando
window.fbq = window.fbq || function() { console.log('fbq stub called'); };
window.gtag = window.gtag || function() { console.log('gtag stub called'); };
window.dataLayer = window.dataLayer || [];
window.ga = window.ga || function() { console.log('ga stub called'); };

// Desabilitar reCAPTCHA se presente
if (typeof grecaptcha !== 'undefined') {
  grecaptcha.ready = function(cb) { console.log('reCAPTCHA disabled offline'); };
}

console.log('🌐 Site clonado carregado em modo offline');
</script>
`;
        
        // Inserir antes do </body>
        html = html.replace('</body>', offlineScript + '</body>');
        
        fs.writeFileSync(path.join(siteDir, pageInfo.file), html, 'utf8');
        pagesSaved++;
      }
      
      // Salvar informações
      const info = {
        originalUrl: startUrl,
        clonedAt: new Date().toISOString(),
        pages: [...pageMap.entries()].map(([url, info]) => ({ url, file: info.file })),
        assetsDownloaded: successCount + cssAssetCount,
        assetsFailed: failCount,
        totalAssets: globalAssets.size + cssAssets.size,
        directory: siteDir
      };
      
      fs.writeFileSync(path.join(siteDir, 'clone-info.json'), JSON.stringify(info, null, 2));
      fs.writeFileSync(path.join(siteDir, 'abrir-site.bat'), `@echo off\nstart index.html`);
      
      console.log(`\n${'='.repeat(60)}`);
      console.log(`🎉 CLONAGEM CONCLUÍDA COM SUCESSO!`);
      console.log(`${'='.repeat(60)}`);
      console.log(`📄 Páginas clonadas: ${pagesSaved}`);
      console.log(`📦 Assets baixados: ${successCount + cssAssetCount}`);
      console.log(`⚠️  Assets com falha: ${failCount}`);
      console.log(`📁 Localização: ${siteDir}`);
      console.log(`${'='.repeat(60)}\n`);
      
      return {
        url: startUrl,
        title: 'Site Clonado',
        statusCode: response.status(),
        links: [],
        images: [],
        contentLength: 0,
        timestamp: new Date().toISOString(),
        cloneInfo: {
          directory: siteDir,
          pagesCloned: pagesSaved,
          assetsDownloaded: successCount + cssAssetCount,
          assetsFailed: failCount,
          totalAssets: globalAssets.size + cssAssets.size,
          success: true
        }
      };
      
    } finally {
      await browser.close();
    }
  }
}

// API
app.post('/crawl', async (req, res) => {
  try {
    const { url } = req.body;
    if (!url) return res.status(400).json({ error: 'URL obrigatória' });
    
    const cloner = new AdvancedWebCloner();
    const result = await cloner.cloneWebsite(url);
    res.json({ success: true, data: result });
  } catch (error) {
    console.error('❌ Erro:', error.message);
    res.status(500).json({ success: false, error: error.message });
  }
});

app.get('/health', (req, res) => {
  res.json({ status: 'ok', service: 'advanced-web-cloner' });
});

app.listen(3001, () => {
  console.log('🌐 Advanced Web Cloner rodando em http://localhost:3001');
  console.log('🚀 Pronto para clonagem profunda de sites!\n');
});