Update app.py

app.py

@@ -2078,31 +2078,37 @@ def clean_and_format_markdown(raw_text):
     return formatted_text
 
 
-def deep_clean_text(text):
+def clean_and_format_markdown(text: str) -> str:
     # Normalize Unicode
     text = unicodedata.normalize('NFKC', text)
-    # Remove zero-width and invisible chars
+
+    # Remove zero-width and other invisible chars
     text = re.sub(r'[\u200B\uFEFF\u200C\u200D]', '', text)
-    # Remove control chars
+
+    # Remove control characters
     text = re.sub(r'[\x00-\x1F\x7F-\x9F]', '', text)
-    # Replace newlines
+
+    # Replace escaped and actual newlines with a single space
     text = text.replace('\\n', ' ').replace('\n', ' ')
-    # Remove HTML tags
+
+    # Remove HTML tags if any
     text = re.sub(r'<[^>]*>', '', text)
-    # Space after punctuation
+
+    # Ensure space after punctuation if missing
     text = re.sub(r'([.,!?])(\S)', r'\1 \2', text)
-    # Space between numbers and letters
+
+    # Ensure spacing between numbers and letters
     text = re.sub(r'(\d)([A-Za-z])', r'\1 \2', text)
     text = re.sub(r'([A-Za-z])(\d)', r'\1 \2', text)
-    # Normalize spaces
+
+    # Normalize multiple spaces
     text = re.sub(r'\s+', ' ', text).strip()
-    # Normalize dashes
-    text = text.replace('−', '-')
 
     return text
 
 
 
+
 prompt = st.chat_input("")
 global combined_text
 def handle_prompt(prompt):
@@ -2254,7 +2260,7 @@ def handle_prompt(prompt):
                     })
                     full_response = output["output"]
 
-                    cleaned_text = deep_clean_text(full_response)
+                    cleaned_text = clean_and_format_markdown(full_response)
 
                     # Use Markdown with a styled div to ensure proper word wrapping
                     
@@ -2265,14 +2271,14 @@ def handle_prompt(prompt):
                         trust_tip, suggestion = get_trust_tip_and_suggestion()
                         
                         # Combine the response text with Trust Tip and Suggestion
-                        combined_text = f"{full_response}\n\n---\n\n**Trust Tip**: {trust_tip}\n\n**Suggestion**: {suggestion}"
+                        combined_text = f"{cleaned_text}\n\n---\n\n**Trust Tip**: {trust_tip}\n\n**Suggestion**: {suggestion}"
                         
                     else:
                         # If already present, just use the full response
-                        combined_text = full_response
+                        combined_text = cleaned_text
                     with response_placeholder:
                         with st.chat_message("assistant"):
-                            st.markdown(f"<pre style='white-space:pre-wrap;word-wrap:break-word;'>{combined_text}</pre>", unsafe_allow_html=True)
+                            st.markdown(combined_text, unsafe_allow_html=True)
 
                             #st.write(combined_text,unsafe_allow_html=True)
                     st.session_state.chat_history.append({"role": "assistant", "content": combined_text})