From 440894f8d3ead0ef438e89cbb5a1b46ff3cd58af Mon Sep 17 00:00:00 2001
From: Gabriel Ecegi <gabriel.ecegi@blogic.cz>
Date: Sat, 14 Dec 2024 10:45:27 +0100
Subject: [PATCH] Fix process/files/batch

---
 backend/open_webui/apps/retrieval/main.py     | 21 +++++++------------
 .../apps/webui/routers/knowledge.py           | 15 +++++++++----
 2 files changed, 18 insertions(+), 18 deletions(-)

diff --git a/backend/open_webui/apps/retrieval/main.py b/backend/open_webui/apps/retrieval/main.py
index 86ea6bf41..4da322e70 100644
--- a/backend/open_webui/apps/retrieval/main.py
+++ b/backend/open_webui/apps/retrieval/main.py
@@ -1062,30 +1062,24 @@ def process_files_batch(
     Process a batch of files and save them to the vector database.
     """
     results: List[BatchProcessFilesResult] = []
-    errors: List[BatchProcessFilesResult]   = []
+    errors: List[BatchProcessFilesResult] = []
     collection_name = form_data.collection_name
-    
 
     # Prepare all documents first
     all_docs: List[Document] = []
-    for file_request in form_data.files:
+    for file in form_data.files:
         try:
-            file = Files.get_file_by_id(file_request.file_id)
-            if not file:
-                log.error(f"process_files_batch: File {file_request.file_id} not found")
-                raise ValueError(f"File {file_request.file_id} not found")
-
-            text_content = file_request.content
+            text_content = file.data.get("content", "")
             
             docs: List[Document] = [
                 Document(
                     page_content=text_content.replace("<br/>", "\n"),
                     metadata={
                         **file.meta,
-                        "name": file_request.filename,
+                        "name": file.filename,
                         "created_by": file.user_id,
                         "file_id": file.id,
-                        "source": file_request.filename,
+                        "source": file.filename,
                     },
                 )
             ]
@@ -1101,9 +1095,9 @@ def process_files_batch(
             ))
 
         except Exception as e:
-            log.error(f"process_files_batch: Error processing file {file_request.file_id}: {str(e)}")
+            log.error(f"process_files_batch: Error processing file {file.id}: {str(e)}")
             errors.append(BatchProcessFilesResult(
-                file_id=file_request.file_id,
+                file_id=file.id,
                 status="failed",
                 error=str(e)
             ))
@@ -1139,7 +1133,6 @@ def process_files_batch(
         errors=errors
     )
 
-
 class ProcessTextForm(BaseModel):
     name: str
     content: str
diff --git a/backend/open_webui/apps/webui/routers/knowledge.py b/backend/open_webui/apps/webui/routers/knowledge.py
index ccc2251d1..21361c7f7 100644
--- a/backend/open_webui/apps/webui/routers/knowledge.py
+++ b/backend/open_webui/apps/webui/routers/knowledge.py
@@ -548,10 +548,17 @@ def add_files_to_knowledge_batch(
         files.append(file)
 
     # Process files
-    result = process_files_batch(BatchProcessFilesForm(
-        files=files,
-        collection_name=id
-    ))
+    try:
+        result = process_files_batch(BatchProcessFilesForm(
+            files=files,
+            collection_name=id
+        ))
+    except Exception as e:
+        log.error(f"add_files_to_knowledge_batch: Exception occurred: {e}", exc_info=True)
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=str(e)
+        )
     
     # Add successful files to knowledge base
     data = knowledge.data or {}