o
     ¢i>  ã                   @   sš  d dl Z dd„ Ze j dg d¢¡dd„ ƒZe j dd	d
g¡dd„ ƒZe j dddg¡dd„ ƒZe j dg d¢¡dd„ ƒZe j dg d¢¡dd„ ƒZe j dddg¡dd„ ƒZ	e j dg d¢¡e j ddd g¡d!d"„ ƒƒZ
e j d#g d$¢¡d%d&„ ƒZe j d'd(d)g¡d*d+„ ƒZe j dg d,¢¡d-d.„ ƒZd/d0„ Ze j dg d1¢¡d2d3„ ƒZe j d4d5d6d7gfd8d9d:gfd;g d<¢fg¡d=d>„ ƒZe j d?d@dAg¡dBdC„ ƒZdS )Dé    Nc                 C   s`   d}| |ƒ}t |ƒdksJ ‚|d jdksJ ‚d}| |ƒ}t |ƒdks%J ‚|d jdks.J ‚d S )	Nzdon't giggleé   é   zn'tzi said don't!é   é   ú!©ÚlenÚtext©Úen_tokenizerr	   Útokens© r   úW/home/ubuntu/.local/lib/python3.10/site-packages/spacy/tests/lang/en/test_exceptions.pyÚ+test_en_tokenizer_handles_basic_contraction   s   r   r	   )z`ain'tz"isn'tzcan't!c                 C   ó   | |ƒ}t |ƒdksJ ‚d S )Nr   ©r   r
   r   r   r   Ú1test_en_tokenizer_handles_basic_contraction_punct   ó   r   ztext_poss,text)zRobin'sÚRobin)zAlexis'sÚAlexisc                 C   s@   | |ƒ}t |ƒdksJ ‚|d j|ksJ ‚|d jdksJ ‚d S )Né   r   r   z'sr   )r   Ú	text_possr	   r   r   r   r   Ú*test_en_tokenizer_handles_poss_contraction   s   r   zschools'zAlexis'c                 C   sJ   | |ƒ}t |ƒdksJ ‚|d j| d¡d ksJ ‚|d jdks#J ‚d S )Nr   r   ú'r   ©r   r	   Úsplitr
   r   r   r   Ú&test_en_tokenizer_splits_trailing_apos   ó   r   )z'emznothin'zol'c                 C   s.   | |ƒ}t |ƒdksJ ‚|d j|ksJ ‚d S )Nr   r   r   r
   r   r   r   Ú'test_en_tokenizer_doesnt_split_apos_exc'   s   r   )zwe'llzYou'llzthere'llzthis'llzthose'llc                 C   sJ   | |ƒ}t |ƒdksJ ‚|d j| d¡d ksJ ‚|d jdks#J ‚d S )Nr   r   r   r   ú'llr   r
   r   r   r   Ú(test_en_tokenizer_handles_ll_contraction.   r   r    ztext_lower,text_title)zcan'tzCan't)zain'tzAin'tc                 C   sd   | |ƒ}| |ƒ}|d j |d j  ¡ ksJ ‚|d j |d j  ¡ ks$J ‚|d j |d j ks0J ‚d S ©Nr   r   )r	   ÚtitleÚlower)r   Ú
text_lowerÚ
text_titleÚtokens_lowerÚtokens_titler   r   r   Ú(test_en_tokenizer_handles_capitalization6   s
   r(   Úpron)ÚIÚYouÚHeÚSheÚItÚWeÚTheyÚcontractionr   z'dc                 C   s4   | || ƒ}|d j |ksJ ‚|d j |ksJ ‚d S r!   )r	   )r   r)   r1   r   r   r   r   Ú"test_en_tokenizer_keeps_title_caseA   s   r2   Úexc)ÚIllÚillÚHellÚhellÚWellÚwellc                 C   r   ©Nr   r   )r   r3   r   r   r   r   Ú$test_en_tokenizer_excludes_ambiguousI   r   r;   zwo_punct,w_punct)zWe'vez`We've)zcouldn'tz	couldn't)c                 C   s4   | |ƒ}t |ƒdksJ ‚| |ƒ}t |ƒdksJ ‚d S )Nr   r   r   )r   Úwo_punctÚw_punctr   r   r   r   Ú&test_en_tokenizer_splits_defined_punctO   s   r>   )ze.g.zp.m.úJan.zDec.zInc.c                 C   r   r:   r   r
   r   r   r   Útest_en_tokenizer_handles_abbrY   r   r@   c                 C   s2   d}| |ƒ}t |ƒdksJ ‚|d jdksJ ‚d S )NzIt's mediocre i.e. bad.é   r   zi.e.r   r
   r   r   r   Ú%test_en_tokenizer_handles_exc_in_text_   s   rB   )Ú1amz12a.m.z11p.m.Ú4pmc                 C   r   )Nr   r   r
   r   r   r   Útest_en_tokenizer_handles_timesf   r   rE   z
text,normszI'mÚiÚamzshan'tÚshallÚnotz.Many factors cause cancer 'cause it is complex)ÚmanyÚfactorsÚcauseÚcancerÚbecauseÚitÚisÚcomplexc                 C   s"   | |ƒ}dd„ |D ƒ|ksJ ‚d S )Nc                 S   s   g | ]}|j ‘qS r   ©Únorm_)Ú.0Útokenr   r   r   Ú
<listcomp>y   s    z5test_en_tokenizer_norm_exceptions.<locals>.<listcomp>r   )r   r	   Únormsr   r   r   r   Ú!test_en_tokenizer_norm_exceptionsl   s   rX   z	text,norm)r?   ÚJanuary)z'cuzrN   c                 C   s   | |ƒ}|d j |ksJ ‚d S )Nr   rR   )r   r	   Únormr   r   r   r   Ú!test_en_lex_attrs_norm_exceptions|   s   r[   )Úpytestr   ÚmarkÚparametrizer   r   r   r   r    r(   r2   r;   r>   r@   rB   rE   rX   r[   r   r   r   r   Ú<module>   sX    
ÿ



ÿ

ÿ




þýþ
