
    i>                        d dl Z d Ze j                  j                  dg d      d        Ze j                  j                  dddg      d	        Ze j                  j                  dd
dg      d        Ze j                  j                  dg d      d        Ze j                  j                  dg d      d        Ze j                  j                  dddg      d        Z	e j                  j                  dg d      e j                  j                  dddg      d               Z
e j                  j                  dg d      d        Ze j                  j                  ddd g      d!        Ze j                  j                  dg d"      d#        Zd$ Ze j                  j                  dg d%      d&        Ze j                  j                  d'd(d)d*gfd+d,d-gfd.g d/fg      d0        Ze j                  j                  d1d2d3g      d4        Zy)5    Nc                     d} | |      }t        |      dk(  sJ |d   j                  dk(  sJ d} | |      }t        |      dk(  sJ |d   j                  dk(  sJ y )	Nzdon't giggle      zn'tzi said don't!      !lentexten_tokenizerr   tokenss      t/var/www/vps2.regionflexible.com/Desarrollo/venv/lib/python3.12/site-packages/spacy/tests/lang/en/test_exceptions.py+test_en_tokenizer_handles_basic_contractionr      sp    D$Fv;!!9>>U"""D$Fv;!!9>>S       r   )z`ain'tz"isn'tzcan't!c                 4     | |      }t        |      dk(  sJ y )Nr   r
   r   s      r   1test_en_tokenizer_handles_basic_contraction_punctr          $Fv;!r   ztext_poss,text)zRobin'sRobin)zAlexis'sAlexisc                      | |      }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  dk(  sJ y )N   r   r   z'sr	   )r   	text_possr   r   s       r   *test_en_tokenizer_handles_poss_contractionr      sL     )$Fv;!!9>>T!!!!9>>T!!!r   zschools'zAlexis'c                      | |      }t        |      dk(  sJ |d   j                  |j                  d      d   k(  sJ |d   j                  dk(  sJ y )Nr   r   'r   r
   r   splitr   s      r   &test_en_tokenizer_splits_trailing_aposr       sW    $Fv;!!9>>TZZ_Q////!9>>S   r   )z'emznothin'zol'c                 \     | |      }t        |      dk(  sJ |d   j                  |k(  sJ y )Nr   r   r	   r   s      r   'test_en_tokenizer_doesnt_split_apos_excr"   '   s4    $Fv;!!9>>T!!!r   )zwe'llzYou'llzthere'llzthis'llzthose'llc                      | |      }t        |      dk(  sJ |d   j                  |j                  d      d   k(  sJ |d   j                  dk(  sJ y )Nr   r   r   r   'llr   r   s      r   (test_en_tokenizer_handles_ll_contractionr%   .   sW    $Fv;!!9>>TZZ_Q////!9>>U"""r   ztext_lower,text_title)zcan'tzCan't)zain'tzAin'tc                 "    | |      } | |      }|d   j                   |d   j                   j                         k(  sJ |d   j                   |d   j                   j                         k(  sJ |d   j                   |d   j                   k(  sJ y Nr   r   )r   titlelower)r   
text_lower
text_titletokens_lowertokens_titles        r   (test_en_tokenizer_handles_capitalizationr.   6   s      
+L
+L?<?#7#7#=#=#?????<?#7#7#=#=#?????<?#7#7777r   pron)IYouHeSheItWeTheycontractionr$   z'dc                 j     | ||z         }|d   j                   |k(  sJ |d   j                   |k(  sJ y r'   )r   )r   r/   r7   r   s       r   "test_en_tokenizer_keeps_title_caser9   A   s?     $,-F!9>>T!!!!9>>[(((r   exc)IllillHellhellWellwellc                 4     | |      }t        |      dk(  sJ y Nr   r   )r   r:   r   s      r   $test_en_tokenizer_excludes_ambiguousrC   I   s    #Fv;!r   zwo_punct,w_punct)zWe'vez`We've)zcouldn'tz	couldn't)c                 d     | |      }t        |      dk(  sJ  | |      }t        |      dk(  sJ y )Nr   r   r   )r   wo_punctw_punctr   s       r   &test_en_tokenizer_splits_defined_punctrG   O   s<     (#Fv;!'"Fv;!r   )ze.g.zp.m.Jan.zDec.zInc.c                 4     | |      }t        |      dk(  sJ y rB   r   r   s      r   test_en_tokenizer_handles_abbrrJ   Y   r   r   c                 `    d} | |      }t        |      dk(  sJ |d   j                  dk(  sJ y )NzIt's mediocre i.e. bad.   r   zi.e.r	   r   s      r   %test_en_tokenizer_handles_exc_in_textrM   _   s9    $D$Fv;!!9>>V###r   )1amz12a.m.z11p.m.4pmc                 4     | |      }t        |      dk(  sJ y )Nr   r   r   s      r   test_en_tokenizer_handles_timesrQ   f   r   r   z
text,normszI'miamzshan'tshallnotz.Many factors cause cancer 'cause it is complex)manyfactorscausecancerbecauseitiscomplexc                 \     | |      }|D cg c]  }|j                    c}|k(  sJ y c c}w )Nnorm_)r   r   normsr   tokens        r   !test_en_tokenizer_norm_exceptionsrc   l   s.     $F%+,EEKK,555,s   )z	text,norm)rH   January)z'cuzrZ   c                 <     | |      }|d   j                   |k(  sJ y )Nr   r_   )r   r   normr   s       r   !test_en_lex_attrs_norm_exceptionsrg   |   s"    $F!9??d"""r   )pytestr   markparametrizer   r   r    r"   r%   r.   r9   rC   rG   rJ   rM   rQ   rc   rg    r   r   <module>rl      sr   ! !CD E
 +-CD"" *i!89! :! !:;" <" !WX# Y# 02DE88 !NO6) 7 P)  NO P
 ,.GH !IJ K
$ !CD E
 	d	GU#$<T	

6
6
 ':<O&PQ# R#r   