cannaiq/backend/scrape-real-clicks.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 scrapeRealClicks() {
  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
    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);

    await page.goto(url, { waitUntil: 'networkidle2', timeout: 60000 });
    await page.waitForTimeout(3000);

    console.log('📍 Current URL:', page.url());

    if (page.url().includes('/age-gate')) {
      console.log('🔒 Handling age gate...');

      // Wait for dropdown button
      await page.waitForSelector('button[role="combobox"]', { timeout: 10000 });
      console.log('  ✅ Found state dropdown');

      // Click the dropdown with real browser click
      await page.click('button[role="combobox"]');
      console.log('  ✅ Clicked dropdown');

      await page.waitForTimeout(2000);

      // Wait for options to appear and click Arizona
      const arizonaClicked = await page.evaluate(() => {
        const options = Array.from(document.querySelectorAll('[role="option"]'));
        const azOption = options.find(opt =>
          opt.textContent?.toLowerCase().includes('arizona')
        ) as HTMLElement;

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

      if (arizonaClicked) {
        console.log('  ✅ Selected Arizona');

        // Wait a bit for React to process
        await page.waitForTimeout(3000);

        console.log('  📍 URL after selection:', page.url());

        // Check if there's a continue/enter button that appeared
        const buttons = await page.evaluate(() => {
          const allButtons = Array.from(document.querySelectorAll('button, a'));
          return allButtons.map(btn => ({
            text: btn.textContent?.trim() || '',
            visible: (btn as HTMLElement).offsetParent !== null,
            ariaLabel: btn.getAttribute('aria-label') || ''
          })).filter(b => b.visible);
        });

        console.log('  📋 Visible buttons:', buttons);

        // If still on age gate, wait for possible redirect
        if (page.url().includes('/age-gate')) {
          console.log('  ⏳ Still on age gate, waiting for redirect...');

          try {
            await page.waitForNavigation({ timeout: 10000 });
            console.log('  ✅ Redirected to:', page.url());
          } catch (e) {
            console.log('  ⚠️  No redirect happened');
          }
        }
      } else {
        console.log('  ❌ Could not find Arizona option');
      }
    }

    // Try to scrape
    console.log('');
    console.log('📦 Attempting to scrape brands...');
    console.log('📍 Final URL:', page.url());

    await page.waitForTimeout(3000);

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

      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,
        bodyText: document.body.innerText.substring(0, 800)
      }));

      console.log('');
      console.log('📄 PAGE CONTENT:');
      console.log('Title:', pageData.title);
      console.log('Text:', pageData.bodyText);
    }

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

scrapeRealClicks();