Update app.py

app.py

@@ -157,9 +157,11 @@ def query_vectara(text):
     else:
         return f"Error: {response.status_code}"
 
-def clean_text(text):
-    # Function to clean text using regex
-    cleaned_text = re.sub(r'[^\w\s]', '', text)  # Remove special characters except spaces
+def remove_references(text):
+    # Regex pattern to find references like [1], [1][2], etc.
+    pattern = r'\[\d+\]+'
+    # Replace found patterns with an empty string
+    cleaned_text = re.sub(pattern, '', text)
     return cleaned_text
 
 def evaluate_content(user_input):
@@ -169,8 +171,11 @@ def evaluate_content(user_input):
     summary = vectara_response_json.get("summary", "")
     sources = vectara_response_json.get("sources", [])
     
-    # Clean summary text
-    summary_clean = clean_text(summary)
+    # Remove references from the summary text
+    summary_no_refs = remove_references(summary)
+    
+    # Clean summary text to remove special characters
+    summary_clean = clean_text(summary_no_refs)
     
     # Process sources to extract and clean necessary information
     sources_info = ""
@@ -185,7 +190,7 @@ def evaluate_content(user_input):
         
         sources_info += f"Title: {title_clean}, Author: {author_clean}, Page: {page_number}\n"
     
-    # Generate text based on the cleaned summary
+    # Generate text based on the cleaned and reference-removed summary
     olmo_output = generate_text(summary_clean)
     olmo_output_clean = clean_text(olmo_output)