Update app.py

app.py

@@ -20,15 +20,65 @@ class AIChatbot:
             'creep', 'brat', 'dweeb', 'goon', 'booby', 'puke', 'vomit', 'dung', 'sap', 
             'clutz', 'knob', 'prick', 'ass', 'shit', 'fuck', 'cock', 'tits', 'pussy', 
             'cunt', 'slut', 'bitch', 'whore', 'skank', 'stupid',
+            'asshole', 'dick', 'douche', 'scumbag', 'slimeball', 'douchebag', 'knobhead',
+            'numskull', 'halfwit', 'nincompoop', 'blockhead', 'dimwit', 'nitwit', 'simpleton',
+            'dunce', 'buffoon', 'doofus', 'clod', 'goober', 'jerkface', 'schmuck', 'scoundrel',
+            'miscreant', 'rat', 'git', 'wazzock', 'pillock', 'prat', 'plonker', 'div', 'bellend',
+            'tosserhead', 'twitbrain', 'sapbrain', 'knucklehead', 'dopey', 'boob', 'dingbat', 'oaf',
+            'ninnyhammer', 'chucklehead', 'saphead', 'pukehead', 'fuckface', 'assface', 'dickhead',
+            'cockhead', 'shithead', 'twatface', 'doucheface', 'bastardface', 'motherfucker', 'shitbag',
+            'cocksucker', 'jackass', 'wankerface', 'tosserface', 'arsehole', 'shitstain', 'assholeface',
+            'prickface', 'dumbfuck', 'fucknut', 'twatwaffle', 'shitbagger', 'dickweed', 'cumdump',
+            'asswipe', 'cockwomble', 'bollocks', 'twat', 'dick', 'fucking',
             # Tagalog bad words
-            'gago', 'putangina', 'hayop', 'lintik', 'walang', 'hiya', 'bobo', 'leche', 
+            'gago', 'putangina', 'putang', 'hayop', 'lintik', 'walang', 'hiya', 'bobo', 'leche', 
             'punyeta', 'sira', 'ulo', 'bwisit', 'pakshet', 'tarantado', 'ulol', 'buwisit', 
-            'hudas', 'kupal', 'shet', 'tae', 'tanga', 'tangina'
+            'hudas', 'kupal', 'shet', 'tae', 'tanga', 'tangina', 'bastos', 'maldita', 'loko',
+            'asar', 'pekpek', 'burat', 'kantot', 'puke', 'kantotin', 'tarantadoin', 'ulolan',
+            'bading', 'bakla', 'unggoy', 'asarin', 'bastusin', 'malditahin', 'buratin', 'pekpekin',
+            'pukein', 'tangain', 'gagoan', 'tarantadohin', 'ina'
         }
         
-        # Tagalog phrases (multi-word profanity)
+        # Bad phrases (multi-word profanity - English and Tagalog)
         self.bad_phrases = {
-            'walang hiya', 'sira ulo', 'walang kwenta', 'walang silbe'
+            # English phrases
+            'fuck you', 'shit you', 'damn you', 'hell you',
+            'you bastard', 'you bitch', 'you dick', 'you prick', 'you cunt', 'you slut', 'you whore',
+            'you jerk', 'you idiot', 'you fool', 'you moron', 'you dumbass', 'you douche', 'you twat',
+            'you bugger', 'you wanker', 'you tosser', 'you poophead', 'you scumbag', 'you slimeball',
+            'you douchebag', 'you knobhead', 'you bozo', 'you twit', 'you dope', 'you numskull',
+            'you halfwit', 'you nincompoop', 'you blockhead', 'you dimwit', 'you nitwit', 'you simpleton',
+            'you dunce', 'you buffoon', 'you doofus', 'you clod', 'you goober', 'you jerkface',
+            'you schmuck', 'you scoundrel', 'you miscreant', 'you rat', 'you puke', 'you vomit',
+            'you dung', 'you ass', 'you tits', 'you pussy', 'you cock', 'you fuckface', 'you assface',
+            'you dickhead', 'you cockhead', 'you shithead', 'you twatface', 'you knobhead', 'you doucheface',
+            'you loser', 'you bastardface', 'you motherfucker', 'you shitbag', 'you cocksucker',
+            'you jackass', 'you wankerface', 'you tosserface', 'you arsehole', 'you asshole', 'you freak', 'you nut',
+            'you scum', 'you creep', 'you brat', 'you dweeb', 'you goon', 'you pukehead', 'you shitstain',
+            'you assholeface', 'you prickface', 'you dumbfuck', 'you fucknut', 'you twatwaffle',
+            'you shitbagger', 'you dickweed', 'you cumdump', 'you asswipe', 'you cockwomble',
+            'you bollocks', 'you wazzock', 'you pillock', 'you plonker', 'you div', 'you bellend',
+            'you twitbrain', 'you motherfucking idiot', 'fuckig stupid',
+            # Tagalog phrases
+            'walang hiya', 'sira ulo', 'walang kwenta', 'walang silbe',
+            'putang ina', 'putang ina ka', 'putang ina mo',
+            'gago ka', 'gago mo', 'gago-gago', 'gago-gago ka', 'gago-gago mo', 'gagoan ka', 'gagoan mo',
+            'tanga ka', 'tanga mo', 'tanga-tanga', 'tanga-tanga ka', 'tanga-tanga mo', 'tangain ka', 'tangain mo', 'tanga-in ka', 'tanga-in mo',
+            'bobo ka', 'bobo mo', 'bobo-bobo', 'bobo-bobo ka', 'bobo-bobo mo', 'bobo-in ka', 'bobo-in mo',
+            'ulol ka', 'ulol mo', 'ulol-ulol', 'ulol-ulol ka', 'ulol-ulol mo', 'ulolan ka', 'ulolan mo', 'ulol-in ka', 'ulol-in mo',
+            'tarantado ka', 'tarantado mo', 'tarantado-tarantado', 'tarantado-tarantado ka', 'tarantado-tarantado mo',
+            'tarantadoin ka', 'tarantadoin mo', 'tarantado-in ka', 'tarantado-in mo', 'tarantadohin ka', 'tarantadohin mo',
+            'bastos ka', 'bastos mo', 'bastusin ka', 'bastusin mo',
+            'maldita ka', 'maldita mo', 'malditahin ka', 'malditahin mo',
+            'loko ka', 'loko mo', 'loko-loko', 'loko-loko ka', 'loko-loko mo',
+            'asar ka', 'asar mo', 'asarin ka', 'asarin mo',
+            'pekpek ka', 'pekpek mo', 'pekpekin ka', 'pekpekin mo',
+            'burat ka', 'burat mo', 'buratin ka', 'buratin mo',
+            'kantot ka', 'kantot mo', 'kantotin ka', 'kantotin mo',
+            'puke ka', 'puke mo', 'pukein ka', 'pukein mo',
+            'bading ka', 'bading mo',
+            'bakla ka', 'bakla mo',
+            'unggoy ka', 'unggoy mo'
         }
         
         # Simple conversation patterns
@@ -87,15 +137,28 @@ class AIChatbot:
         # Normalize message: convert to lowercase
         message_lower = message.lower()
         
-        # First, check for bad phrases (multi-word profanity like "walang hiya", "sira ulo")
+        # First, check for bad phrases (multi-word profanity like "walang hiya", "sira ulo", "gago-gago")
         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)
+            # Replace hyphens with spaces for better matching (handles "gago-gago" as "gago gago")
+            phrase_normalized = phrase.replace('-', ' ')
+            # Remove punctuation but keep spaces, normalize whitespace
+            phrase_clean = re.sub(r'[^\w\s]', '', phrase_normalized)
+            phrase_clean = re.sub(r'\s+', ' ', phrase_clean).strip()
+            
+            # Normalize message similarly - replace hyphens with spaces
+            message_normalized = message_lower.replace('-', ' ')
+            message_clean_phrase = re.sub(r'[^\w\s]', '', message_normalized)
+            message_clean_phrase = re.sub(r'\s+', ' ', message_clean_phrase).strip()
+            
             # 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
+            # Split phrase into words and create pattern that matches with any whitespace
+            phrase_words = phrase_clean.split()
+            if len(phrase_words) > 0:
+                # Create pattern that matches words with one or more spaces between them
+                # Using word boundaries to ensure whole words are matched
+                phrase_pattern = r'\b' + r'\s+'.join(re.escape(word) for word in phrase_words) + 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
@@ -109,8 +172,13 @@ class AIChatbot:
         for char, replacement in obfuscation_map.items():
             normalized = normalized.replace(char, replacement)
         
+        # Replace hyphens with spaces to handle hyphenated words like "gago-gago"
+        normalized = normalized.replace('-', ' ')
+        
         # Remove all non-word characters (except spaces) for word boundary checking
         message_clean = re.sub(r'[^\w\s]', '', normalized)
+        # Normalize multiple spaces to single space
+        message_clean = re.sub(r'\s+', ' ', message_clean).strip()
         words = message_clean.split()
         
         # Check for exact word matches in cleaned message