chore: move cli scraper

This commit is contained in:
karishmas6
2024-04-18 22:19:24 +05:30
parent 2803d302e3
commit 32ec5b6f45

View File

@@ -20,37 +20,4 @@ fastify.get('/', async (request, reply) => {
await fastify.listen(3000, (err, address) => {
if (err) throw err;
console.log(`Server listening on ${fastify.server.address().port}`)
});
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);
}
console.log('Scraped data:', scrapedData); // Replace with desired saving method
},
});
await crawler.run([{ url }]);
}
const url = process.argv[2];
const selectors = process.argv.slice(3); // Selectors are passed as subsequent arguments
const waitForSeconds = parseInt(process.argv[selectors.length + 2] || 2); // Optional wait time
if (url && selectors.length > 0) {
await scrapeData(url, selectors, waitForSeconds);
} else {
console.error('Please provide URL and selectors as arguments.');
}
});