Upload 6 files

api_clients.py

@@ -10,11 +10,13 @@ from config import (
     GENERIC_SYSTEM_PROMPT_WITH_SEARCH, FollowUpSystemPrompt
 )
 from chat_processing import (
-    history_to_messages, messages_to_history, create_multimodal_message,
+    history_to_messages, messages_to_history,
     remove_code_block, apply_search_replace_changes, send_to_sandbox,
     history_to_chatbot_messages, get_gradio_language
 )
-from file_processing import extract_text_from_file
+from file_processing import (
+    extract_text_from_file, create_multimodal_message
+)
 from web_extraction import extract_website_content
 
 # HF Inference Client

app.py

@@ -9,8 +9,10 @@ from config import (
 from api_clients import generation_code, tavily_client
 from chat_processing import (
     clear_history, history_to_chatbot_messages, update_image_input_visibility,
-    get_gradio_language, send_to_sandbox
+    get_gradio_language, send_to_sandbox,
 )
+from file_processing import create_multimodal_message
+from web_extraction import enhance_query_with_search
 
 
 def demo_card_click(e: gr.EventData):

chat_processing.py

@@ -88,6 +88,26 @@ def update_image_input_visibility(model):
     is_glm_vl = model.get("id") == "THUDM/GLM-4.1V-9B-Thinking"
     return gr.update(visible=is_ernie_vl or is_glm_vl)
 
+def create_multimodal_message(text, image=None):
+    """Create a multimodal message with text and optional image"""
+    if image is None:
+        return {"role": "user", "content": text}
+    
+    from file_processing import process_image_for_model
+    content = [
+        {
+            "type": "text",
+            "text": text
+        },
+        {
+            "type": "image_url",
+            "image_url": {
+                "url": process_image_for_model(image)
+            }
+        }
+    ]
+    
+    return {"role": "user", "content": content}
 def apply_search_replace_changes(original_html: str, changes_text: str) -> str:
     """Apply search/replace changes to HTML content"""
     if not changes_text.strip():

file_processing.py

@@ -67,4 +67,24 @@ def extract_text_from_file(file_path):
         else:
             return ""
     except Exception as e:
-        return f"Error extracting text: {e}"
+        return f"Error extracting text: {e}"
+
+def create_multimodal_message(text, image=None):
+    """Create a multimodal message with text and optional image"""
+    if image is None:
+        return {"role": "user", "content": text}
+    
+    content = [
+        {
+            "type": "text",
+            "text": text
+        },
+        {
+            "type": "image_url",
+            "image_url": {
+                "url": process_image_for_model(image)
+            }
+        }
+    ]
+    
+    return {"role": "user", "content": content}