Update app.py

app.py

@@ -2077,6 +2077,35 @@ def clean_and_format_markdown(raw_text):
     return formatted_text
 
 
+def deep_clean_text(text):
+    # Normalize Unicode to a standard form
+    text = unicodedata.normalize('NFKC', text)
+
+    # Remove zero-width characters and other invisible Unicode chars (ZWSP: \u200B, ZWNBSP: \uFEFF, etc.)
+    text = re.sub(r'[\u200B\uFEFF\u200C\u200D]', '', text)
+
+    # Replace escaped and actual newlines with a space
+    text = text.replace('\\n', ' ').replace('\n', ' ')
+
+    # Remove HTML tags if any
+    text = re.sub(r'<[^>]*>', '', text)
+
+    # Insert space after punctuation if missing
+    text = re.sub(r'([.,!?])(\S)', r'\1 \2', text)
+
+    # Fix 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)
+
+    # Replace multiple spaces with a single space
+    text = re.sub(r'\s+', ' ', text).strip()
+
+    # You can also replace unusual hyphens or dashes with a standard dash
+    text = text.replace('−', '-')
+
+    return text
+
+
 
 prompt = st.chat_input("")
 global combined_text
@@ -2229,7 +2258,10 @@ def handle_prompt(prompt):
                     })
                     full_response = output["output"]
 
-                    #cleaned_text = clean_text(full_response)
+                    cleaned_text = deep_clean_text(full_response)
+
+                    # Use Markdown with a styled div to ensure proper word wrapping
+                    
                     #formatted_text = clean_and_format_markdown(cleaned_text)
 
                     if "Trust Tip" not in full_response and "Suggestion" not in full_response:
@@ -2238,12 +2270,18 @@ def handle_prompt(prompt):
                         
                         # 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}"
+                        
                     else:
                         # If already present, just use the full response
                         combined_text = full_response
                     with response_placeholder:
                         with st.chat_message("assistant"):
-                            st.write(combined_text,unsafe_allow_html=True)
+                            html_content = f"""
+                            <div style="white-space: pre-wrap; word-wrap: break-word;">
+                                {combined_text}
+                            </div>
+                            """
+                            st.write(html_content,unsafe_allow_html=True)
                     st.session_state.chat_history.append({"role": "assistant", "content": combined_text})
                     copy_to_clipboard(combined_text)