Update index.js

index.js

@@ -338,6 +338,74 @@ app.post('/process', upload.single('htmlFile'), (req, res) => {
   }
 });
 
+// Add this function to handle script execution
+function executeCheerioScript(html, scriptContent) {
+  try {
+    const $ = cheerio.load(html, {
+      decodeEntities: false,
+      xmlMode: false,
+      lowerCaseTags: true
+    });
+
+    // Create a safe function from the script content
+    const extractorFunction = new Function('$', `
+      ${scriptContent}
+    `);
+
+    // Execute the script with Cheerio instance
+    const result = extractorFunction($);
+    return {
+      success: true,
+      data: result,
+      error: null
+    };
+  } catch (err) {
+    return {
+      success: false,
+      data: null,
+      error: err.message
+    };
+  }
+}
+
+// cherio extract endpoint
+app.post('/extract', upload.single('htmlFile'), (req, res) => {
+  try {
+    const startTime = Date.now();
+    let htmlContent = req.file 
+      ? req.file.buffer.toString('utf8')
+      : req.body.html || '';
+
+    const extractorScript = req.body.script;
+
+    if (!htmlContent.trim()) {
+      return res.status(400).json({ error: 'No HTML content provided.' });
+    }
+
+    if (!extractorScript) {
+      return res.status(400).json({ error: 'No extractor script provided.' });
+    }
+
+    const result = executeCheerioScript(htmlContent, extractorScript);
+
+    return res.json({
+      success: result.success,
+      data: result.data,
+      error: result.error,
+      processingTime: `${Date.now() - startTime}ms`
+    });
+
+  } catch (err) {
+    console.error('Extraction failed:', err);
+    return res.status(500).json({ 
+      success: false,
+      error: 'Internal server error.',
+      details: err.message.substring(0, 100)
+    });
+  }
+});
+
+
 const PORT = process.env.PORT || 3000;
 app.listen(PORT, () => {
   console.log(`Server running on http://localhost:${PORT}`);