Spaces:
Running
Running
Create server.js
Browse files
server.js
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
const express = require('express');
|
| 2 |
+
const axios = require('axios');
|
| 3 |
+
const cheerio = require('cheerio');
|
| 4 |
+
const cors = require('cors');
|
| 5 |
+
const bodyParser = require('body-parser');
|
| 6 |
+
|
| 7 |
+
const app = express();
|
| 8 |
+
app.use(cors());
|
| 9 |
+
app.use(bodyParser.json());
|
| 10 |
+
|
| 11 |
+
// Scraping endpoint
|
| 12 |
+
app.post('/api/scrape', async (req, res) => {
|
| 13 |
+
try {
|
| 14 |
+
const { url } = req.body;
|
| 15 |
+
|
| 16 |
+
const response = await axios.get(url, {
|
| 17 |
+
headers: {
|
| 18 |
+
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
|
| 19 |
+
}
|
| 20 |
+
});
|
| 21 |
+
|
| 22 |
+
const $ = cheerio.load(response.data);
|
| 23 |
+
|
| 24 |
+
const data = {
|
| 25 |
+
url,
|
| 26 |
+
content: $('body').text().substring(0, 5000), // Limit content
|
| 27 |
+
pageCount: 1,
|
| 28 |
+
imageCount: $('img').length,
|
| 29 |
+
linkCount: $('a').length,
|
| 30 |
+
timestamp: new Date().toISOString()
|
| 31 |
+
};
|
| 32 |
+
|
| 33 |
+
res.json(data);
|
| 34 |
+
} catch (error) {
|
| 35 |
+
res.status(500).json({
|
| 36 |
+
error: 'Scraping failed',
|
| 37 |
+
details: error.message
|
| 38 |
+
});
|
| 39 |
+
}
|
| 40 |
+
});
|
| 41 |
+
|
| 42 |
+
const PORT = process.env.PORT || 3001;
|
| 43 |
+
app.listen(PORT, () => {
|
| 44 |
+
console.log(`Server running on port ${PORT}`);
|
| 45 |
+
});
|