o
    i                     @   sL   d dl Z d dlmZ ddlmZ dd Zdd Zd	d
 Ze jj	dd Z
dS )    N)Chinese   )make_tempdirc                 C   s   |   }t }|j| ||j  ksJ t $}|d }| | t }|j| ||j  ks5J W d    d S 1 s@w   Y  d S )N	tokenizer)to_bytesr   r   
from_bytesr   to_disk	from_disk)zh_tokenizertokenizer_bytesnlpd	file_path r   V/home/ubuntu/.local/lib/python3.10/site-packages/spacy/tests/lang/zh/test_serialize.pyzh_tokenizer_serialize   s   
"r   c                 C      t |  d S Nr   )zh_tokenizer_charr   r   r    test_zh_tokenizer_serialize_char      r   c                 C   r   r   r   )zh_tokenizer_jiebar   r   r   !test_zh_tokenizer_serialize_jieba   r   r   c                 C   s<   ddddidddiid}t |}|  t|j d S )Nr   zspacy.zh.ChineseTokenizerpkuseg)z@tokenizers	segmenterpkuseg_modelmedicine)r   
initialize)r   from_configr   r   r   )zh_tokenizer_pkusegconfigr   r   r   r   2test_zh_tokenizer_serialize_pkuseg_with_processors   s   
r"   )pytestspacy.lang.zhr   utilr   r   r   r   markslowr"   r   r   r   r   <module>   s    