Hazm 0.12.1

Fixed

• Normalizer: Resolved a TypeError: argument of type 'NoneType' is not iterable in the token_spacing method. This occurred when correct_spacing was enabled but seperate_mi was disabled, causing the internal verb list to remain uninitialized as None.
• WordTokenizer: Resolved an AttributeError where internal markers like before_verbs and after_verbs were missing if the tokenizer was initialized with join_verb_parts=False. These attributes are now safely initialized as empty sets by default.