Update app.py

app.py

@@ -577,35 +577,28 @@ def clean_response(response_text):
     response_text = re.sub(r'<\|assistant\|>', '', response_text, flags=re.DOTALL)
 
     # Extract the document name and page number
-    document_match = re.search(r"Document\(metadata=\{'source':'(.+?)','page':(\d+)\}", response_text)
+    document_match = re.search(r"Document\(metadata=\{'source':'(.+?)','page':(\d+)\},page_content='(.*?)'", response_text)
     if document_match:
         document_name = document_match.group(1).split('/')[-1]  # Get the document name
         page_number = document_match.group(2)  # Get the page number
+        page_content = document_match.group(3)  # Get the actual page content
     else:
         document_name = "Unknown"
         page_number = "Unknown"
-
-    # Remove the 'Document(metadata=...' part and keep only the page content
-    response_text = re.sub(r'Document\(metadata=\{.*?\},page_content=', '', response_text, flags=re.DOTALL)
-    
-    # Remove any mention of "Document:" in the response
-    response_text = re.sub(r'- Document:.*', '', response_text)
+        page_content = "No content available"
 
     # Remove any unwanted escape characters like \u and \u00
-    response_text = re.sub(r'\\u[0-9A-Fa-f]{4}', '', response_text)
+    page_content = re.sub(r'\\u[0-9A-Fa-f]{4}', '', page_content)
 
     # Ensure proper spacing between words and dates (e.g., "born04/04/1963" becomes "born 04/04/1963")
-    response_text = re.sub(r'([a-zA-Z])(\d)', r'\1 \2', response_text)  # Letter followed by a number
-    response_text = re.sub(r'(\d)([a-zA-Z])', r'\1 \2', response_text)  # Number followed by a letter
+    page_content = re.sub(r'([a-zA-Z])(\d)', r'\1 \2', page_content)  # Letter followed by a number
+    page_content = re.sub(r'(\d)([a-zA-Z])', r'\1 \2', page_content)  # Number followed by a letter
 
     # Ensure there are spaces between capital letters and lowercase (e.g., "CognomeBELLAVISTA" becomes "Cognome BELLAVISTA")
-    response_text = re.sub(r'([a-z])([A-Z])', r'\1 \2', response_text)
-
-    # Remove the phrase "Sure! The Responses are as follows:" from the actual content
-    response_text = re.sub(r'Sure! The Responses are as follows:', '', response_text).strip()
+    page_content = re.sub(r'([a-z])([A-Z])', r'\1 \2', page_content)
 
     # Clean up the text by removing extra whitespace
-    cleaned_response = re.sub(r'\s+', ' ', response_text).strip()
+    cleaned_response = re.sub(r'\s+', ' ', page_content).strip()
 
     # Format the final response with bullet points and re-add "Sure! The Responses are as follows:"
     final_response = f"""
@@ -623,6 +616,7 @@ Sure! The Responses are as follows:
 
 
 
+
 # Define a new template specifically for GPT-4o-mini in VDB Details mode
 gpt4o_mini_template_details = f"""
 As a highly specialized assistant, I provide precise, detailed, and informative responses. On this bright day of {current_date}, I'm equipped to assist with all your queries about Birmingham, Alabama, offering detailed insights tailored to your needs.