cannaiq/backend/scrape-bypass-age-gate.ts

import puppeteer from 'puppeteer-extra';
import StealthPlugin from 'puppeteer-extra-plugin-stealth';
import { Pool } from 'pg';

puppeteer.use(StealthPlugin());

const pool = new Pool({
  connectionString: 'postgresql://sail:password@localhost:5432/dutchie_menus'
});

async function scrapeBypassAgeGate() {
  let browser;

  try {
    // Get random proxy
    const proxyResult = await pool.query(`
      SELECT host, port, protocol FROM proxies
      ORDER BY RANDOM() LIMIT 1
    `);

    const proxy = proxyResult.rows[0];
    const proxyUrl = `${proxy.protocol}://${proxy.host}:${proxy.port}`;

    console.log('🔌 Proxy:', `${proxy.host}:${proxy.port}`);

    // Launch browser with proxy
    browser = await puppeteer.launch({
      headless: true,
      args: [
        '--no-sandbox',
        '--disable-setuid-sandbox',
        `--proxy-server=${proxyUrl}`
      ]
    });

    const page = await browser.newPage();

    // Mobile Chrome UA
    const mobileUA = 'Mozilla/5.0 (Linux; Android 10; SM-G973F) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Mobile Safari/537.36';
    await page.setUserAgent(mobileUA);

    console.log('📱 User-Agent:', mobileUA);
    console.log('');

    // Set age gate bypass cookie/localStorage
    // First, go to the main domain to set cookies
    console.log('🍪 Setting age verification cookies...');
    await page.goto('https://curaleaf.com', { waitUntil: 'domcontentloaded' });

    // Set cookies for age verification
    await page.evaluateOnNewDocument(() => {
      // Try various localStorage/cookie approaches
      try {
        // Set localStorage for age verification
        localStorage.setItem('age-verified', 'true');
        localStorage.setItem('curaleaf-age-verified', 'true');
        localStorage.setItem('state', 'arizona');
        localStorage.setItem('selectedState', 'arizona');
      } catch (e) {}
    });

    // Set cookies manually
    await page.setCookie(
      {
        name: 'age-verified',
        value: 'true',
        domain: '.curaleaf.com',
        path: '/',
      },
      {
        name: 'curaleaf-age-gate',
        value: 'passed',
        domain: '.curaleaf.com',
        path: '/',
      },
      {
        name: 'state',
        value: 'arizona',
        domain: '.curaleaf.com',
        path: '/',
      }
    );

    console.log('  ✅ Cookies set');
    console.log('');

    const url = 'https://curaleaf.com/stores/curaleaf-dispensary-phoenix-airport/brands';
    console.log('🌐 Going to:', url);
    console.log('');

    await page.goto(url, { waitUntil: 'networkidle2', timeout: 60000 });
    await page.waitForTimeout(5000);

    // Check current URL
    const currentUrl = page.url();
    console.log('📍 Current URL:', currentUrl);

    if (currentUrl.includes('/age-gate')) {
      console.log('⚠️  Still on age gate - cookies didn\'t work');
      console.log('');

      // Debug: check what cookies/localStorage we have
      const storageData = await page.evaluate(() => {
        const cookies = document.cookie;
        const localStorageItems: any = {};
        for (let i = 0; i < localStorage.length; i++) {
          const key = localStorage.key(i);
          if (key) {
            localStorageItems[key] = localStorage.getItem(key);
          }
        }
        return { cookies, localStorage: localStorageItems };
      });

      console.log('Current cookies:', storageData.cookies);
      console.log('Current localStorage:', storageData.localStorage);
    } else {
      console.log('✅ Age gate bypassed!');
    }

    // Try to scrape brands
    console.log('');
    console.log('📦 Scraping brands...');

    const brands = await page.evaluate(() => {
      const selectors = [
        '[data-testid*="brand"]',
        '[class*="Brand"]',
        '[class*="brand"]',
        'a[href*="/brand/"]',
        '.brand-card',
        '.brand-item'
      ];

      const found = new Set<string>();

      selectors.forEach(selector => {
        document.querySelectorAll(selector).forEach(el => {
          const text = el.textContent?.trim();
          if (text && text.length > 0 && text.length < 50) {
            found.add(text);
          }
        });
      });

      return Array.from(found);
    });

    console.log(`Found ${brands.length} brands`);
    if (brands.length > 0) {
      console.log('─'.repeat(60));
      brands.forEach((b, i) => console.log(`  ${i + 1}. ${b}`));
      console.log('─'.repeat(60));
    } else {
      // Debug
      const pageData = await page.evaluate(() => ({
        title: document.title,
        url: window.location.href,
        bodyText: document.body.innerText.substring(0, 500),
        hasNextRoot: document.getElementById('__next') !== null
      }));

      console.log('');
      console.log('📄 PAGE DEBUG:');
      console.log('Title:', pageData.title);
      console.log('URL:', pageData.url);
      console.log('Has __next:', pageData.hasNextRoot);
      console.log('Text preview:', pageData.bodyText);
    }

  } catch (error: any) {
    console.error('❌ Error:', error.message);
  } finally {
    if (browser) await browser.close();
    await pool.end();
  }
}

scrapeBypassAgeGate();