cannaiq/backend/scrape-with-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 scrapeWithAgeGate() {
  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('');

    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 });

    // Check if we're on age gate
    const currentUrl = page.url();
    console.log('📍 Current URL:', currentUrl);

    if (currentUrl.includes('/age-gate')) {
      console.log('🔒 Age gate detected, handling...');

      // Wait for state selector to appear
      await page.waitForSelector('button[role="combobox"]', { timeout: 10000 });

      // Click state selector dropdown
      await page.click('button[role="combobox"]');
      console.log('  ✅ Clicked state dropdown');

      await page.waitForTimeout(1000);

      // Try to find and click Arizona option
      const arizonaClicked = await page.evaluate(() => {
        // Look for Arizona in the dropdown options
        const options = Array.from(document.querySelectorAll('[role="option"]'));
        const azOption = options.find(opt =>
          opt.textContent?.toLowerCase().includes('arizona') ||
          opt.textContent?.toLowerCase().includes('az')
        );

        if (azOption) {
          (azOption as HTMLElement).click();
          return true;
        }
        return false;
      });

      if (arizonaClicked) {
        console.log('  ✅ Selected Arizona');
        await page.waitForTimeout(2000);

        // Look for submit/enter button
        const submitButtonClicked = await page.evaluate(() => {
          // Look for submit button - could be various selectors
          const possibleSelectors = [
            'button[type="submit"]',
            'button:contains("Enter")',
            'button:contains("Submit")',
            'button:contains("Continue")',
            'a[aria-label*="age"]',
            'button'
          ];

          for (const selector of possibleSelectors) {
            const buttons = Array.from(document.querySelectorAll('button, a'));
            const submitBtn = buttons.find(btn => {
              const text = btn.textContent?.toLowerCase() || '';
              const ariaLabel = btn.getAttribute('aria-label')?.toLowerCase() || '';
              return text.includes('enter') ||
                     text.includes('submit') ||
                     text.includes('continue') ||
                     ariaLabel.includes('age');
            });

            if (submitBtn) {
              (submitBtn as HTMLElement).click();
              return true;
            }
          }
          return false;
        });

        if (submitButtonClicked) {
          console.log('  ✅ Clicked submit button');

          // Wait for navigation
          try {
            await page.waitForNavigation({ waitUntil: 'networkidle2', timeout: 30000 });
            console.log('  ✅ Age gate passed!');
            console.log('  📍 New URL:', page.url());
          } catch (navError) {
            console.log('  ⚠️  Navigation timeout - checking current page...');
            console.log('  📍 Current URL:', page.url());
          }
        } else {
          console.log('  ⚠️  Could not find submit button, checking if redirect happened anyway...');
          await page.waitForTimeout(3000);
          console.log('  📍 Current URL:', page.url());
        }
      } else {
        console.log('  ❌ Could not find Arizona option');

        // Debug: show what options are available
        const availableOptions = await page.evaluate(() => {
          return Array.from(document.querySelectorAll('[role="option"]'))
            .map(opt => opt.textContent?.trim())
            .filter(Boolean);
        });
        console.log('  Available options:', availableOptions);
      }
    }

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

    await page.waitForTimeout(3000);

    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: show page content
      const pageData = await page.evaluate(() => ({
        title: document.title,
        url: window.location.href,
        bodyText: document.body.innerText.substring(0, 500)
      }));

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

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

scrapeWithAgeGate();