Update app.py

app.py

@@ -145,55 +145,23 @@ def send_feedback_via_email(name, email, feedback):
         st.error(f"Error sending email: {e}")
 
 def clean_text(text):
-    # Replace escaped newlines with actual newlines
-    text = text.replace('\\n', '\n')
-
+    soup = BeautifulSoup(text, 'html.parser')
+    
     # Convert <a> tags to Markdown links
-    def convert_links(match):
-        url = match.group(1)
-        link_text = match.group(2)
-        return f"[{link_text}]({url})"
-
-    # Handle <a> tags to preserve clickable URLs in Markdown format
-    text = re.sub(r'<a [^>]*href="([^"]+)"[^>]*>(.*?)</a>', convert_links, text)
-
-    # Remove <span>, <i>, <b>, and other unwanted HTML tags
-    text = re.sub(r'<span[^>]*>|</span>|<i[^>]*>|</i>|<b[^>]*>|</b>', '', text)
-
-    # Remove any remaining HTML tags
-    text = re.sub(r'<[^>]+>', '', text)
-
-    # Split the text into paragraphs
-    paragraphs = text.split('\n\n')
-
-    cleaned_paragraphs = []
-    for paragraph in paragraphs:
-        lines = paragraph.split('\n')
-        cleaned_lines = []
-        for line in lines:
-            # Preserve and correctly format headings or bold text
-            if line.strip().startswith('**') and line.strip().endswith('**'):
-                cleaned_line = line.strip()
-            else:
-                # Remove asterisks, special characters, and fix merged text
-                cleaned_line = re.sub(r'\*|\−|\∗', '', line)
-                cleaned_line = re.sub(r'([a-z])([A-Z])', r'\1 \2', cleaned_line)
-
-            # Handle bullet points correctly
-            if cleaned_line.strip().startswith('-'):
-                cleaned_line = '\n' + cleaned_line.strip()
-
-            # Remove extra spaces
-            cleaned_line = re.sub(r'\s+', ' ', cleaned_line).strip()
-            cleaned_lines.append(cleaned_line)
-
-        # Join the lines within each paragraph
-        cleaned_paragraph = '\n'.join(cleaned_lines)
-        cleaned_paragraphs.append(cleaned_paragraph)
-
-    # Join the paragraphs back together
-    cleaned_text = '\n\n'.join(para for para in cleaned_paragraphs if para)
-
+    for a in soup.find_all('a'):
+        a.replace_with(f"[{a.get_text()}]({a['href']})")
+    
+    # Remove unwanted tags but preserve text
+    for tag in ['span', 'i', 'b']:
+        for element in soup.find_all(tag):
+            element.unwrap()  # Remove the tag but keep the content
+    
+    # Get the cleaned text
+    cleaned_text = soup.get_text()
+    
+    # Remove excessive whitespace
+    cleaned_text = re.sub(r'\s+', ' ', cleaned_text).strip()
+    
     return cleaned_text
 def side():        
     with st.sidebar.form(key='feedback_form'):
@@ -631,7 +599,7 @@ if prompt :
                 st.write(full_response)
     
         # Add AI response to chat history
-        st.session_state.chat_history.append({"role": "assistant", "content": full_response})
+        st.session_state.chat_history.append({"role": "assistant", "content": cleaned_text})
         copy_to_clipboard(full_response)