Update app.py

app.py

@@ -526,50 +526,6 @@ import re
 
 #     return final_response
 
-# def clean_response(response_text):
-#     # Remove system and user tags
-#     response_text = re.sub(r'<\|system\|>.*?<\|end\|>', '', response_text, flags=re.DOTALL)
-#     response_text = re.sub(r'<\|user\|>.*?<\|end\|>', '', response_text, flags=re.DOTALL)
-#     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)
-#     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
-#     else:
-#         document_name = "Unknown"
-#         page_number = "Unknown"
-
-#     # Remove the entire 'Document(metadata=...' and any mention of it from the response
-#     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)
-
-#     # Remove any unwanted escape characters like \u and \u00
-#     response_text = re.sub(r'\\u[0-9A-Fa-f]{4}', '', response_text)
-
-#     # Ensure proper spacing between words and dates
-#     response_text = re.sub(r'([a-zA-Z])(\d)', r'\1 \2', response_text)
-#     response_text = re.sub(r'(\d)([a-zA-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()
-
-#     # Clean up the text by removing extra whitespace
-#     cleaned_response = re.sub(r'\s+', ' ', response_text).strip()
-
-#     # Format the final response with bullet points
-#     final_response = f"""
-# Sure! Here is the response for your Query:
-# • Document name - {document_name}
-# • Page No - {page_number}
-# • Response - {cleaned_response}
-# """
-
-#     return final_response
-
 def clean_response(response_text):
     # Remove system and user tags
     response_text = re.sub(r'<\|system\|>.*?<\|end\|>', '', response_text, flags=re.DOTALL)
@@ -585,40 +541,32 @@ def clean_response(response_text):
         document_name = "Unknown"
         page_number = "Unknown"
 
-    # Remove document metadata and other unnecessary parts
-    response_text = re.sub(r"Document\(metadata=\{.*?\}\)", '', response_text, flags=re.DOTALL)
+    # Remove the entire 'Document(metadata=...' and any mention of it from the response
+    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)
 
-    # Remove unwanted parts like "Nome del documento" and "Numero di pagina"
-    response_text = re.sub(r"Nome del documento:.*\n?", '', response_text)
-    response_text = re.sub(r"Numero di pagina:.*\n?", '', response_text)
+    # Remove any unwanted escape characters like \u and \u00
+    response_text = re.sub(r'\\u[0-9A-Fa-f]{4}', '', response_text)
 
-    # Clean up and extract the relevant response content
-    response_text = re.sub(r'\\u[0-9A-Fa-f]{4}', '', response_text)  # Remove any unwanted escape characters
+    # Ensure proper spacing between words and dates
+    response_text = re.sub(r'([a-zA-Z])(\d)', r'\1 \2', response_text)
+    response_text = re.sub(r'(\d)([a-zA-Z])', r'\1 \2', response_text)
 
-    # Ensure proper spacing and remove extra text around the "Contenuto effettivo" and "Risposta" parts
-    response_text = re.sub(r'Contenuto effettivo: ', '', response_text)
-    response_text = re.sub(r'Risposta: ', '', response_text)
-    cleaned_response = response_text.strip()
+    # 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()
 
-    # Handle blank or empty responses
-    if not cleaned_response:
-        cleaned_response = "No relevant content found in the document."
+    # Clean up the text by removing extra whitespace
+    cleaned_response = re.sub(r'\s+', ' ', response_text).strip()
 
-    # Create a table format for the document name, page number, and response
+    # Format the final response with bullet points
     final_response = f"""
-    <br>Here is the response of your Query <br>
-    <table style="width:100%;border: 1px solid black;border-collapse: collapse;">
-        <tr>
-            <th style="border: 1px solid black;padding: 8px;text-align: left;">Document Name</th>
-            <th style="border: 1px solid black;padding: 8px;text-align: left;">Page No</th>
-        </tr>
-        <tr>
-            <td style="border: 1px solid black;padding: 8px;">{document_name}</td>
-            <td style="border: 1px solid black;padding: 8px;">{page_number}</td>
-        </tr>
-    </table>
-    <p><strong>Response:</strong> {cleaned_response}</p>
-    """
+Sure! Here is the response for your Query:
+• Document name - {document_name}
+• Page No - {page_number}
+• Response - {cleaned_response}
+"""
 
     return final_response
 
@@ -633,6 +581,8 @@ def clean_response(response_text):
 
 
 
+
+
 # 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.