a
    w°Ge?  ã                   @   s    d dl mZ d dlZdd„ ZdS )é    )ÚCounterNc                 C   sZ   dd„ | D ƒ}d  |¡}t d¡}||ƒ}dd„ |D ƒ}d  dd„ t|ƒ d	¡D ƒ¡}|S )
Nc                 S   s   g | ]}t | ¡ ƒd  ‘qS )r   )ÚlistÚkeys)Ú.0Úd© r   úID:\eodhd\seo\seo_app\app\core\internal_funcs\determine_cluster_keyword.pyÚ
<listcomp>   ó    z-determine_cluster_keyword.<locals>.<listcomp>ú Zru_core_news_smc                 S   s   g | ]}|j s|jr|j‘qS r   )Zis_stopZis_alphaÚtext)r   Útokenr   r   r   r	      r
   z, c                 s   s   | ]\}}|V  qd S )Nr   )r   ÚwordÚ_r   r   r   Ú	<genexpr>   r
   z,determine_cluster_keyword.<locals>.<genexpr>é   )ÚjoinÚspacyÚloadr   Úmost_common)Zcluster_valuesZall_keysZcombined_stringZnlpÚdocÚwordsÚkeywordr   r   r   Údetermine_cluster_keyword   s    

r   )Úcollectionsr   r   r   r   r   r   r   Ú<module>   s   