Update app.py

app.py

@@ -11,14 +11,24 @@ class AIChatbot:
         self.database_url = database_url
         self.conversation_history = []
         
-        # Profanity filter - list of bad words to filter
+        # Profanity filter - list of bad words to filter (English and Tagalog)
         self.bad_words = {
+            # English bad words
             'damn', 'hell', 'crap', 'suck', 'idiot', 'fool', 'jerk', 'loser', 'dumb', 'trash',
             'butt', 'freak', 'nut', 'moron', 'dummy', 'bozo', 'twit', 'dope', 'dumbass', 
             'poophead', 'jerkoff', 'bugger', 'wanker', 'tosser', 'bastard', 'scum', 'slime',
             'creep', 'brat', 'dweeb', 'goon', 'booby', 'puke', 'vomit', 'dung', 'sap', 
             'clutz', 'knob', 'prick', 'ass', 'shit', 'fuck', 'cock', 'tits', 'pussy', 
-            'cunt', 'slut', 'bitch', 'whore', 'skank'
+            'cunt', 'slut', 'bitch', 'whore', 'skank', 'stupid',
+            # Tagalog bad words
+            'gago', 'putangina', 'hayop', 'lintik', 'walang', 'hiya', 'bobo', 'leche', 
+            'punyeta', 'sira', 'ulo', 'bwisit', 'pakshet', 'tarantado', 'ulol', 'buwisit', 
+            'hudas', 'kupal', 'shet', 'tae', 'tanga', 'tangina'
+        }
+        
+        # Tagalog phrases (multi-word profanity)
+        self.bad_phrases = {
+            'walang hiya', 'sira ulo'
         }
         
         # Simple conversation patterns
@@ -77,7 +87,17 @@ class AIChatbot:
         # Normalize message: convert to lowercase
         message_lower = message.lower()
         
-        # First, normalize common obfuscation characters
+        # First, check for bad phrases (multi-word profanity like "walang hiya", "sira ulo")
+        for phrase in self.bad_phrases:
+            # Remove punctuation and check if phrase exists in message
+            phrase_clean = re.sub(r'[^\w\s]', '', phrase)
+            message_clean_phrase = re.sub(r'[^\w\s]', '', message_lower)
+            # Check if phrase appears in message (with flexible spacing)
+            phrase_pattern = r'\b' + r'\s+'.join(re.escape(word) for word in phrase_clean.split()) + r'\b'
+            if re.search(phrase_pattern, message_clean_phrase, re.IGNORECASE):
+                return True
+        
+        # Normalize common obfuscation characters
         # Replace common character substitutions (numbers/symbols) with letters
         obfuscation_map = {
             '0': 'o', '1': 'i', '3': 'e', '4': 'a', '5': 's',