refac

Co-Authored-By: Classic298 <27028174+Classic298@users.noreply.github.com>

backend/open_webui/utils/files.py

@@ -32,6 +32,7 @@ from open_webui.storage.provider import Storage
 
 BASE64_IMAGE_URL_PREFIX = re.compile(r'data:image/\w+;base64,', re.IGNORECASE)
 MARKDOWN_IMAGE_URL_PATTERN = re.compile(r'!\[(.*?)\]\((.+?)\)', re.IGNORECASE)
+FILE_CONTENT_URL_PATTERN = re.compile(r'^/api/v1/files/([^/?#]+)/content(?:[?#]|$)')
 
 # Extension-based MIME fallback, only used when ENABLE_IMAGE_CONTENT_TYPE_EXTENSION_FALLBACK is True.
 _IMAGE_MIME_FALLBACK = {
@@ -90,7 +91,8 @@ async def get_image_base64_from_url(url: str, user=None) -> Optional[str]:
         else:
             # Non-URL string — treat as file_id. Delegate to the canonical
             # file-ID resolver which enforces ownership/access checks.
-            return await get_image_base64_from_file_id(url, user=user)
+            file_id_match = FILE_CONTENT_URL_PATTERN.match(url)
+            return await get_image_base64_from_file_id(file_id_match.group(1) if file_id_match else url, user=user)
 
     except Exception:
         return None

backend/open_webui/utils/middleware.py

@@ -995,6 +995,29 @@ def extract_base64_images(value: Any, files: list) -> Any:
     return value
 
 
+async def store_tool_result_image(request, image_url, metadata, user):
+    """Keep saved tool images out of chat JSON, falling back to inline data if storage fails."""
+    metadata = metadata or {}
+    if (
+        not isinstance(image_url, str)
+        or not image_url.startswith('data:image/')
+        or not is_saved_chat_id(metadata.get('chat_id'))
+    ):
+        return image_url
+
+    try:
+        stored_url = await get_file_url_from_base64(
+            request,
+            image_url,
+            {key: metadata.get(key) for key in ('chat_id', 'message_id', 'session_id')},
+            user,
+        )
+        return stored_url or image_url
+    except Exception:
+        log.warning('Could not store tool image; retaining inline image')
+        return image_url
+
+
 async def process_tool_result(
     request,
     tool_function_name,
@@ -1425,6 +1448,12 @@ async def chat_completion_tools_handler(
                     )
 
                     if tool_result_files:
+                        for file_item in tool_result_files:
+                            if file_item.get('type') == 'image':
+                                file_item['url'] = await store_tool_result_image(
+                                    request, file_item.get('url'), metadata, user
+                                )
+
                         await event_emitter(
                             {
                                 'type': 'files',
@@ -2113,7 +2142,7 @@ async def convert_url_images_to_base64(form_data, user=None):
         new_content = []
 
         for item in content:
-            if not isinstance(item, dict) or item.get('type') != 'image_url':
+            if not isinstance(item, dict) or item.get('type') not in ('image_url', 'input_image'):
                 new_content.append(item)
                 continue
 
@@ -2130,13 +2159,15 @@ async def convert_url_images_to_base64(form_data, user=None):
 
             try:
                 base64_data = await get_image_base64_from_url(image_url, user=user)
-                if base64_data:
+                if base64_data and isinstance(image_url_data, str):
+                    new_content.append({**item, 'image_url': base64_data})
+                elif base64_data:
                     image_url_payload = {'url': base64_data}
                     if isinstance(image_url_data, dict) and image_url_data.get('detail'):
                         image_url_payload['detail'] = image_url_data['detail']
                     new_content.append(
                         {
-                            'type': 'image_url',
+                            'type': item['type'],
                             'image_url': image_url_payload,
                         }
                     )
@@ -3321,7 +3352,8 @@ async def drain_approved_tool_calls(request, form_data, user, model, metadata) -
         display_files = []
         for file_item in result.get('files', []):
             if file_item.get('type') == 'image' and file_item.get('url', '').startswith('data:'):
-                output_parts.append({'type': 'input_image', 'image_url': file_item['url']})
+                image_url = await store_tool_result_image(request, file_item['url'], metadata, user)
+                output_parts.append({'type': 'input_image', 'image_url': image_url})
             else:
                 display_files.append(file_item)
 
@@ -5907,7 +5939,8 @@ async def streaming_chat_response_handler(response, ctx):
                         for file_item in result.get('files', []):
                             if file_item.get('type') == 'image' and file_item.get('url', '').startswith('data:'):
                                 # LLM-only: add as input_image part, not frontend display output.
-                                output_parts.append({'type': 'input_image', 'image_url': file_item['url']})
+                                image_url = await store_tool_result_image(request, file_item['url'], metadata, user)
+                                output_parts.append({'type': 'input_image', 'image_url': image_url})
                             else:
                                 # Frontend display (MCP images, audio, etc.)
                                 display_files.append(file_item)
@@ -6071,6 +6104,8 @@ async def streaming_chat_response_handler(response, ctx):
                                     }
                                 )
 
+                        new_form_data = await convert_url_images_to_base64(new_form_data, user=user)
+
                         if filter_functions:
                             new_form_data, _ = await process_filter_functions(
                                 request=request,