fixed a few errors

awsfunctions.py

@@ -82,33 +82,48 @@ def download_files_from_s3(local_folder, file_path_list):
 def download_folder_from_s3(local_folder, aws_folder_prefix):
     s3 = boto3.client('s3')
     bucket_name = os.getenv("AWS_BUCKET_NAME")
-    folder_prefix = aws_folder_prefix
     
+    if not bucket_name:
+        raise ValueError("AWS_BUCKET_NAME environment variable is not set")
+
     try:
+        # Create the local folder if it doesn't exist
+        os.makedirs(local_folder, exist_ok=True)
+
         # List objects in the S3 bucket
         paginator = s3.get_paginator('list_objects_v2')
-        page_iterator = paginator.paginate(Bucket=bucket_name, Prefix=folder_prefix)
-        
-        # Download filtered files
+        page_iterator = paginator.paginate(Bucket=bucket_name, Prefix=aws_folder_prefix)
+
+        # Process objects
         for page in page_iterator:
             for obj in page.get('Contents', []):
                 key = obj['Key']
                                 
+                # Determine if it's a file or directory
+                if obj['Size'] == 0:
+                    continue
+                
                 # Construct local file path
-                local_path = os.path.join(local_folder, key)
+                local_path = os.path.join(local_folder, os.path.relpath(key, aws_folder_prefix))
                 os.makedirs(os.path.dirname(local_path), exist_ok=True)
+
                 try:
                     print(f"Downloading: {key} -> {local_path}")
                     s3.download_file(bucket_name, key, local_path)
                     print(f"Downloaded: {local_path}")
-                except Exception as e:
-                    print(f"Error downloading {key}: {e}")
-                    raise e
+                except ClientError as e:
+                    if e.response['Error']['Code'] == 'AccessDenied':
+                        print(f"Permission denied when trying to download {key}: {e}")
+                    elif e.response['Error']['Code'] == 'NoSuchKey':
+                        print(f"The object {key} does not exist in the bucket.")
+                    elif e.response['Error']['Code'] == "":
+                        pass
+                    else:
+                        print(f"An error occurred while downloading {key}: {e}")
+                        raise e
     
-    except NoCredentialsError:
-        print("No AWS credentials found.")
     except Exception as e:
-        print(f"An error occurred: {e}")
+        print(f"An unexpected error occurred : {e}")
 
 def delete_s3_folder(folder_path):
     bucket_name = os.getenv("AWS_BUCKET_NAME")

pdf_processing.py

@@ -113,7 +113,7 @@ def process_pdf(pdf_file):
     if check_file_exists_in_s3(os.path.join(aws_prefix_path, pdf_file.name)):
         temp_dir = tempfile.mkdtemp()
         download_folder_from_s3(local_folder=temp_dir, aws_folder_prefix=os.path.join(aws_prefix_path, "qdrant"))
-        client = qdrant_client.QdrantClient(path=os.path.join(temp_dir))
+        client = qdrant_client.QdrantClient(path=os.path.join(temp_dir, "qdrant"))
         image_store = QdrantVectorStore(client = client , collection_name=f"image_collection")
         text_store = QdrantVectorStore(client = client , collection_name=f"text_collection")
         index = MultiModalVectorStoreIndex.from_vector_store(vector_store=text_store, image_store=image_store)