couple_matters/tools/get_address.js

const puppeteer = require('puppeteer');

const url = 'https://minsu.xiaozhu.com/detail?luId=354701406371854&startDate=2025-12-21&endDate=2025-12-22';

console.log('📍 Extracting address from listing...\n');

(async () => {
  const browser = await puppeteer.launch({
    headless: "new",
    args: ['--no-sandbox', '--disable-setuid-sandbox']
  });

  const page = await browser.newPage();
  await page.setUserAgent('Mozilla/5.0 (iPhone; CPU iPhone OS 14_7_1 like Mac OS X) AppleWebKit/605.1.15');

  try {
    console.log('🌐 Loading page...');
    await page.goto(url, { waitUntil: 'networkidle2', timeout: 30000 });
    await new Promise(resolve => setTimeout(resolve, 5000));

    console.log('📸 Taking screenshot...');
    await page.screenshot({ path: './xiaozhu_detail_page.png', fullPage: true });

    console.log('🔍 Extracting information...\n');

    const info = await page.evaluate(() => {
      const result = {
        title: '',
        address: '',
        location: '',
        nearbyLandmarks: [],
        metro: [],
        price: '',
        amenities: [],
        fullText: ''
      };

      // Title
      const titleEl = document.querySelector('h1, .title, [class*="title"]');
      if (titleEl) result.title = titleEl.textContent.trim();

      // Look for address keywords
      const bodyText = document.body.textContent;
      result.fullText = bodyText.substring(0, 2000);

      // Common address patterns
      const addressPatterns = [
        /地址[：:]\s*(.{5,50})/,
        /位于[：:]?\s*(.{5,50})/,
        /详细地址[：:]\s*(.{5,50})/,
        /([^，。]{2,}路\d+号[^，。]{0,20})/,
        /([^，。]{2,}街\d+号[^，。]{0,20})/
      ];

      for (const pattern of addressPatterns) {
        const match = bodyText.match(pattern);
        if (match && match[1]) {
          result.address = match[1].trim();
          break;
        }
      }

      // Look for specific elements
      const allElements = document.querySelectorAll('div, span, p');
      allElements.forEach(el => {
        const text = el.textContent.trim();

        // Address
        if (text.includes('地址') || text.includes('位于')) {
          if (text.length < 100 && text.length > 5) {
            result.location = text;
          }
        }

        // Metro
        if (text.includes('地铁') || text.includes('号线')) {
          if (text.length < 50) {
            result.metro.push(text);
          }
        }

        // Landmarks
        if (text.includes('医院') || text.includes('公园') || text.includes('商场')) {
          if (text.length < 30) {
            result.nearbyLandmarks.push(text);
          }
        }

        // Amenities
        if (text.includes('厨房') || text.includes('冰箱') || text.includes('洗衣机')) {
          if (text.length < 20) {
            result.amenities.push(text);
          }
        }
      });

      // Price
      const priceEl = document.querySelector('[class*="price"]');
      if (priceEl) result.price = priceEl.textContent.trim();

      return result;
    });

    console.log('📋 LISTING INFORMATION:\n');
    console.log(`Title: ${info.title || 'Not found'}`);
    console.log(`\nPrice: ${info.price || 'Not found'}`);
    console.log(`\nAddress: ${info.address || 'Not found in structured format'}`);
    console.log(`\nLocation Info: ${info.location || 'Not found'}`);

    if (info.metro.length > 0) {
      console.log(`\nMetro: ${info.metro.slice(0, 3).join(', ')}`);
    }

    if (info.nearbyLandmarks.length > 0) {
      console.log(`\nNearby: ${info.nearbyLandmarks.slice(0, 5).join(', ')}`);
    }

    if (info.amenities.length > 0) {
      console.log(`\nAmenities: ${info.amenities.slice(0, 5).join(', ')}`);
    }

    console.log('\n\n📄 PAGE TEXT PREVIEW (first 500 chars):\n');
    console.log(info.fullText.substring(0, 500));
    console.log('\n\n💡 Check xiaozhu_detail_page.png for full page screenshot');

  } catch (err) {
    console.error('❌ Error:', err.message);
  } finally {
    await browser.close();
  }
})();