Files
parcer/scraper/src/scraper.js

24 lines
796 B
JavaScript
Raw Normal View History

2024-04-18 22:20:06 +05:30
import { PlaywrightCrawler } from 'crawlee';
2024-04-18 22:19:40 +05:30
async function scrapeData(url, selectors, waitForSeconds = 2) {
const crawler = new PlaywrightCrawler({
requestHandler: async ({ page }) => {
await page.goto(url);
// Wait for specific time (optional)
await page.waitForTimeout(waitForSeconds * 1000);
console.log('Received selectors:', selectors);
const scrapedData = [];
for (const selector of selectors) {
const elementData = await page.$$eval(selector, elements => elements.map(el => el.textContent.trim()));
scrapedData.push(...elementData);
}
2024-04-19 01:01:16 +05:30
console.log('Scraped data:', scrapedData);
2024-04-18 22:19:40 +05:30
},
});
await crawler.run([{ url }]);
}