import os import re from confusables import is_confusable, confusable_regex SCAM_BRANDS_FILE = os.path.join(os.path.dirname(__file__), 'scam_brands.txt') def load_scam_brands(): """Load the list of scam brands from the text file.""" with open(SCAM_BRANDS_FILE, encoding='utf-8') as f: return [line.strip().strip('"') for line in f if line.strip()] SCAM_BRANDS = load_scam_brands() def find_confusable_brand(message): """ Check if the message contains a confusable/homoglyph variant of any scam brand. Returns the matched brand if found, otherwise None. Does not return the brand if the match is an exact (byte-for-byte, case-sensitive) match. """ for brand in SCAM_BRANDS: # Build a regex that matches the brand or any confusable variant regex_string = confusable_regex(brand, include_character_padding=False) regex = re.compile(regex_string) for match in regex.finditer(message): matched_text = match.group(0) # Skip if the matched text is exactly the same as the brand (case-sensitive) if matched_text.strip().lower() == brand.lower().strip(): continue else: print(f"matched_text: {matched_text.lower().strip()} brand: {brand.lower().strip()}") return brand return None def test_find_confusable_brand(): """ Test the find_confusable_brand function with example messages. """ test_cases = [ "This is a message from Amazοn support.", # Greek omicron instead of o "Your Αpple account has been locked.", # Greek capital alpha instead of a "Contact S0ciété Générale for more info.", # Zero instead of O "Welcome to Netflix!", "This is a message from a random sender.", "Bonjour, c'est le livreur votre colis ne rentrait pas dans la boite aux lettres merci de choisir un point relais sur : https://mondiaIrelais-expedition.com", "599915 est votre code de vérification Leboncoin." ] for msg in test_cases: result = find_confusable_brand(msg) if result: print(f"[ALERT] Message: '{msg}' => Confusable brand detected: {result}") else: print(f"[OK] Message: '{msg}' => No confusable brand detected.") if __name__ == "__main__": test_find_confusable_brand()