@peopl-health/nexus 2.4.9-fix-pdf-processing → 2.4.9-fix-mime-type

package/lib/helpers/filesHelper.js

@@ -1,6 +1,7 @@
 const { PDFDocument } = require('pdf-lib');
 const { execFile } = require('child_process');
 const fs = require('fs').promises;
+const fsSync = require('fs');
 const path = require('path');
 const sharp = require('sharp');
 
@@ -22,28 +23,52 @@ async function convertPdfToImages(pdfName, existingPdfPath = null) {
     const args = ['-jpeg', pdfPath, outputPattern];
     logger.info('[convertPdfToImages] Running: pdftoppm', args.join(' '));
     
-    execFile('pdftoppm', args, (error, stdout, stderr) => {
+    const timeout = 30000;
+    let timedOut = false;
+    
+    const child = execFile('pdftoppm', args, { timeout, maxBuffer: 10 * 1024 * 1024 }, (error, stdout, stderr) => {
+      if (timedOut) {
+        return;
+      }
+      
       if (error) {
         logger.error('[convertPdfToImages] Error details:', { 
           error: error.message, 
           stderr, 
           pdfPath,
-          pdfExists: require('fs').existsSync(pdfPath)
+          pdfExists: fsSync.existsSync(pdfPath),
+          killed: error.killed,
+          signal: error.signal
         });
         return reject(new Error(`Error splitting PDF: ${stderr || error.message}`));
       }
 
-      fs.readdir(outputDir, (err, files) => {
-        if (err) {
-          return reject(new Error(`Error reading output directory: ${err.message}`));
-        }
+      logger.info('[convertPdfToImages] pdftoppm completed successfully');
 
-        const jpgFiles = files
-          .filter(file => file.startsWith(sanitizedName) && file.endsWith('.jpg'))
-          .map(file => path.join(outputDir, file));
+      fs.readdir(outputDir)
+        .then(files => {
+          const jpgFiles = files
+            .filter(file => file.startsWith(sanitizedName) && file.endsWith('.jpg'))
+            .map(file => path.join(outputDir, file));
 
-        resolve(jpgFiles);
-      });
+          logger.info(`[convertPdfToImages] Found ${jpgFiles.length} image files`);
+          resolve(jpgFiles);
+        })
+        .catch(err => {
+          logger.error('[convertPdfToImages] Error reading output directory:', { error: err.message });
+          reject(new Error(`Error reading output directory: ${err.message}`));
+        });
+    });
+
+    const timeoutId = setTimeout(() => {
+      timedOut = true;
+      child.kill('SIGTERM');
+      logger.error('[convertPdfToImages] Process timed out after 30 seconds', { pdfPath });
+      reject(new Error('PDF conversion timed out after 30 seconds'));
+    }, timeout);
+
+    child.on('exit', () => {
+      clearTimeout(timeoutId);
     });
   });
 }
@@ -136,46 +161,69 @@ const cleanupFiles = async (files) => {
 };
 
 async function downloadMediaAndCreateFile(code, reply) {
-  const resultMedia = await Message.findOne({
-    message_id: reply.message_id,
-    timestamp: reply.timestamp,
-    media: { $ne: null }
-  });
+  try {
+    const resultMedia = await Message.findOne({
+      message_id: reply.message_id,
+      timestamp: reply.timestamp,
+      media: { $ne: null }
+    });
 
-  if (!resultMedia) return [];
+    if (!resultMedia) return [];
 
-  if (!resultMedia.media || !resultMedia.media.key) {
-    logger.info('[downloadMediaAndCreateFile] No valid media found for message:', reply.message_id);
-    return [];
-  }
+    if (!resultMedia.media || !resultMedia.media.key) {
+      logger.info('[downloadMediaAndCreateFile] No valid media found for message:', reply.message_id);
+      return [];
+    }
 
-  const { bucketName, key } = resultMedia.media;
-  if (!bucketName || !key) return [];
-  
-  const [subType, fileName] = key.split('/');
-  
-  const sanitizedCode = sanitizeFilename(code, 20);
-  const sanitizedSubType = sanitizeFilename(subType, 10);
-  const sanitizedFileName = sanitizeFilename(fileName, 50);
-  
-  const sourceFile = `${sanitizedCode}-${sanitizedSubType}-${sanitizedFileName}`;
-  const downloadPath = path.join(__dirname, 'assets', 'tmp', sourceFile);
-  
-  logger.info('[downloadMediaAndCreateFile] Downloading file', { sourceFile, downloadPath, bucketName, key });
-  
-  await fs.mkdir(path.dirname(downloadPath), { recursive: true });
-  await downloadFileFromS3(bucketName, key, downloadPath);
+    const { bucketName, key } = resultMedia.media;
+    if (!bucketName || !key) return [];
+    
+    const [subType, fileName] = key.split('/');
+    
+    const sanitizedCode = sanitizeFilename(code, 20);
+    const sanitizedSubType = sanitizeFilename(subType, 10);
+    const sanitizedFileName = sanitizeFilename(fileName, 50);
+    
+    const sourceFile = `${sanitizedCode}-${sanitizedSubType}-${sanitizedFileName}`;
+    const downloadPath = path.join(__dirname, 'assets', 'tmp', sourceFile);
+    
+    logger.info('[downloadMediaAndCreateFile] Downloading file', { sourceFile, downloadPath, bucketName, key });
+    
+    await fs.mkdir(path.dirname(downloadPath), { recursive: true });
+    await downloadFileFromS3(bucketName, key, downloadPath);
 
-  const { name: baseName } = path.parse(sourceFile);
-  const fileNames = (subType === 'document' ||  subType === 'application')
-    ? await convertPdfToImages(baseName, downloadPath)
-    : [downloadPath];
+    const { name: baseName } = path.parse(sourceFile);
+    let fileNames = [];
+    
+    if (subType === 'document' || subType === 'application') {
+      try {
+        fileNames = await convertPdfToImages(baseName, downloadPath);
+        logger.info('[downloadMediaAndCreateFile] PDF converted successfully', { imageCount: fileNames.length });
+      } catch (conversionError) {
+        logger.error('[downloadMediaAndCreateFile] PDF conversion failed:', { 
+          error: conversionError.message,
+          sourceFile 
+        });
+        fileNames = [];
+      } finally {
+        try {
+          await fs.unlink(downloadPath);
+        } catch (unlinkError) {
+          logger.warn('[downloadMediaAndCreateFile] Failed to delete PDF:', { error: unlinkError.message });
+        }
+      }
+    } else {
+      fileNames = [downloadPath];
+    }
 
-  if (subType === 'document' ||  subType === 'application') {
-    await fs.unlink(downloadPath);
+    return fileNames;
+  } catch (error) {
+    logger.error('[downloadMediaAndCreateFile] Error processing media:', { 
+      error: error.message,
+      message_id: reply.message_id 
+    });
+    return [];
   }
-
-  return fileNames;
 }
 
 module.exports = {

package/lib/helpers/llmsHelper.js

@@ -1,7 +1,7 @@
 const llmConfig = require('../config/llmConfig.js');
 const { logger } = require('../utils/logger');
 const fs = require('fs');
-const mime = require('mime-types');
+const path = require('path');
 
 
 async function analyzeImage(imagePath, isSticker = false, contentType = null) {
@@ -30,14 +30,30 @@ async function analyzeImage(imagePath, isSticker = false, contentType = null) {
       };
     }
     
+    // Determine mime type from file extension
     let mimeType = contentType;
     if (!mimeType) {
-      if (imagePath.toLowerCase().endsWith('.webp')) {
-        mimeType = 'image/webp';
-      } else {
-        mimeType = mime.lookup(imagePath) || 'image/jpeg';
-      }
+      const ext = path.extname(imagePath).toLowerCase();
+      const mimeMap = {
+        '.jpg': 'image/jpeg',
+        '.jpeg': 'image/jpeg',
+        '.png': 'image/png',
+        '.gif': 'image/gif',
+        '.webp': 'image/webp'
+      };
+      mimeType = mimeMap[ext] || 'image/jpeg'; // Default to jpeg for pdftoppm output
+    }
+    
+    // Validate that mime type is supported by Claude
+    const supportedMimeTypes = ['image/jpeg', 'image/png', 'image/gif', 'image/webp'];
+    if (!supportedMimeTypes.includes(mimeType)) {
+      logger.warn('[analyzeImage] Unsupported mime type, defaulting to image/jpeg:', { 
+        originalMimeType: mimeType, 
+        imagePath 
+      });
+      mimeType = 'image/jpeg';
     }
+    
     if (mimeType === 'image/vnd.wap.wbmp') {
       logger.info('Skipping image with MIME type:', mimeType);
       return {
@@ -49,6 +65,7 @@ async function analyzeImage(imagePath, isSticker = false, contentType = null) {
       };
     }
     // Read the image file and convert to base64
+    logger.info('[analyzeImage] Reading image file:', { imagePath: imagePath.split('/').pop() });
     const imageBuffer = await fs.promises.readFile(imagePath);
     const base64Image = imageBuffer.toString('base64');
 
@@ -77,6 +94,7 @@ async function analyzeImage(imagePath, isSticker = false, contentType = null) {
         },
       ],
     });
+    logger.info('[analyzeImage] Description received');
     const description = messageDescription.content[0].text;
 
     // For stickers, skip medical analysis and table extraction

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@peopl-health/nexus",
-  "version": "2.4.9-fix-pdf-processing",
+  "version": "2.4.9-fix-mime-type",
   "description": "Core messaging and assistant library for WhatsApp communication platforms",
   "keywords": [
     "whatsapp",