如何爬取 Tata 1mg | 1mg.com 药品数据爬虫指南

import asyncio
from playwright.async_api import async_playwright

async def scrape_1mg_playwright():
    async with async_playwright() as p:
        # 使用隐身设置启动浏览器
        browser = await p.chromium.launch(headless=True)
        page = await browser.new_page()
        
        # 前往分类页面
        await page.goto('https://www.1mg.com/categories/fitness-supplements-63', wait_until='networkidle')
        
        # 等待产品卡片加载
        await page.wait_for_selector('.style__product-card___1Y_A-')
        
        # 提取数据
        products = await page.query_selector_all('.style__product-card___1Y_A-')
        for item in products:
            name = await (await item.query_selector('.style__pro-title___3G3mI')).inner_text()
            price = await (await item.query_selector('.style__price-tag___3yJdp')).inner_text()
            print(f'Product: {name} | Price: {price}')
            
        await browser.close()

asyncio.run(scrape_1mg_playwright())

import scrapy

class OneMgSpider(scrapy.Spider):
    name = 'one_mg'
    allowed_domains = ['1mg.com']
    start_urls = ['https://www.1mg.com/categories/all-medicines-1']

    def parse(self, response):
        # 建议在该站点使用 Scrapy-Playwright 中间件
        for product in response.css('.style__product-card___1Y_A-'):
            yield {
                'name': product.css('.style__pro-title___3G3mI::text').get(),
                'price': product.css('.style__price-tag___3yJdp::text').get(),
                'link': response.urljoin(product.css('a::attr(href)').get())
            }

        # 处理简单的分页
        next_page = response.css('ul.pagination li.next a::attr(href)').get()
        if next_page:
            yield response.follow(next_page, self.parse)

const puppeteer = require('puppeteer');

(async () => {
  const browser = await puppeteer.launch({ headless: true });
  const page = await browser.newPage();
  
  // 伪装指纹
  await page.setUserAgent('Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36');

  try {
    await page.goto('https://www.1mg.com/categories/homeopathy-57', { waitUntil: 'networkidle2' });
    
    const data = await page.evaluate(() => {
      const items = Array.from(document.querySelectorAll('.style__product-card___1Y_A-'));
      return items.map(i => ({
        name: i.querySelector('.style__pro-title___3G3mI')?.innerText,
        price: i.querySelector('.style__price-tag___3yJdp')?.innerText
      }));
    });

    console.log(data);
  } catch (e) {
    console.error('Scraping failed:', e);
  } finally {
    await browser.close();
  }
})();