U
    .Rí`¼  ã                   @   s   d dl mZ G dd„ dƒZdS )é    )Úngramsc                   @   s   e Zd Zddd„ZdS )ÚCorpusé   ú<s>ú</s>c                 C   sZ  d| _ tƒ | _tƒ | _tƒ | _|D ]Ä}|D ]4}| j |d¡d | j|< | j |d¡d | j|< q*| j |d¡d | j|< | j |d¡d | j|< t|g| |g dƒ}|D ]:}|  j d7  _ || jkrÚ| j|  d7  < qªd| j|< qªq"t| j ¡ ƒ||g }	|	D ]R}
|	D ]F}|  j |7  _ | j |
|fd¡| | j|
|f< | j|  |7  < �q
�qd S )Nr   r   é   )	ÚnÚdictZ
vocabularyZunigramsZbigramsÚgetr   ÚlistÚkeys)ÚselfZcorpusÚalphaÚstart_symbolZ
end_symbolÚsentenceÚwordZgramsZgramZ_vocabularyÚwZw2© r   ú8C:\Users\PC\Pytho\Projects\NLP\brownclustering\corpus.pyÚ__init__   s,    
 zCorpus.__init__N)r   r   r   )Ú__name__Ú
__module__Ú__qualname__r   r   r   r   r   r      s   r   N)Z	nltk.utilr   r   r   r   r   r   Ú<module>   s   