
    i'                     <   d dl Z g dZg dZg dZe j                  j                  dg d      d        Ze j                  j                  de      e j                  j                  dg d	      d
               Ze j                  j                  de      e j                  j                  dg d	      d               Ze j                  j                  de      e j                  j                  ddg      e j                  j                  dg d	      d                      Z	e j                  j                  de      e j                  j                  ddg      e j                  j                  dg d	      d                      Z
e j                  j                  de      e j                  j                  dg d	      d               Ze j                  j                  de      e j                  j                  dg d	      d               Ze j                  j                  ddg      d        Ze j                  j                  ddg      d        Ze j                  j                  de      e j                  j                  dg d	      d               Ze j                  j                  de      e j                  j                  ddg      e j                  j                  dg d	      d                      Ze j                  j                  dg d      d        Zd Zd Zy)     N)([{*))]}r   ))r   r   )r   r   )r   r	   )r   r   text)r   z((<c                 F     | |      }t        |      t        |      k(  sJ y )Nlenuk_tokenizerr
   tokenss      s/var/www/vps2.regionflexible.com/Desarrollo/venv/lib/python3.12/site-packages/spacy/tests/lang/uk/test_tokenizer.py$test_uk_tokenizer_handles_only_punctr      s"    $Fv;#d)###    punct)u   Приветu   Привітu   Ґелґотатиu   З'єднанняu   Єдністьu   їхніc                      | ||z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y N   r      r   r
   r   r   r
   r   s       r   #test_uk_tokenizer_splits_open_punctr      sP    
 %$,'Fv;!!9>>U"""!9>>T!!!r   c                      | ||z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y r   r   r   s       r   $test_uk_tokenizer_splits_close_punctr      sP    
 $,'Fv;!!9>>T!!!!9>>U"""r   	punct_add`c                      | ||z   |z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y N   r   r   r   r   r   r   r   r
   r   s        r   ,test_uk_tokenizer_splits_two_diff_open_punctr%   $   sl     %)+d23Fv;!!9>>U"""!9>>Y&&&!9>>T!!!r   'c                      | ||z   |z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y r"   r   r$   s        r   -test_uk_tokenizer_splits_two_diff_close_punctr(   1   sk     $,23Fv;!!9>>T!!!!9>>U"""!9>>Y&&&r   c                      | ||z   |z   |z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y )N   r   r#   r   r   s       r   (test_uk_tokenizer_splits_same_open_punctr+   >   sZ    
 %%-%/$67Fv;!!9>>U"""!9>>T!!!r   c                      | ||z   |z   |z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y )Nr*   r   r   r   r   s       r   )test_uk_tokenizer_splits_same_close_punctr-   I   sZ    
 $,.67Fv;!!9>>T!!!!9>>U"""r   u	   'Тестc                 \     | |      }t        |      dk(  sJ |d   j                  dk(  sJ y )Nr   r   r&   r   r   s      r   )test_uk_tokenizer_splits_open_appostropher/   T   s4    $Fv;!!9>>S   r   u
   Тест''c                 d     | |      }t        |      dk(  sJ  | d      }t        |      dk(  sJ y )Nr   z''r   r   )r   r
   r   tokens_puncts       r   )test_uk_tokenizer_splits_double_end_quoter2   [   s;    $Fv;!%L|!!!r   zpunct_open,punct_closec                      | ||z   |z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y r"   r   )r   
punct_openpunct_closer
   r   s        r   )test_uk_tokenizer_splits_open_close_punctr6   c   sl     *t+k9:Fv;!!9>>Z'''!9>>T!!!!9>>[(((r   zpunct_open2,punct_close2)r    r&   c                     | ||z   |z   |z   |z         }t        |      dk(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ |d   j                  |k(  sJ y )N   r   r   r   r#   r*   r   )r   r4   r5   punct_open2punct_close2r
   r   s          r    test_uk_tokenizer_two_diff_punctr;   q   s     +
2T9KG,VWFv;!!9>>[(((!9>>Z'''!9>>T!!!!9>>[(((!9>>\)))r   )u   Привет.u   Привіт.u   Ґелґотати.u   З'єднання.u   Єдність.u	   їхні.c                 <     | |      }|d   j                   dk(  sJ y )Nr   .r
   r   s      r   %test_uk_tokenizer_splits_trailing_dotr?      s$     $F!9>>S   r   c                 X    d} | |      }|t        |      dz
     j                  dk(  sJ y )Nu+   (Раз, два, три, проверка).r   r=   r   r   s      r   'test_uk_tokenizer_splits_bracket_periodrA      s2    8D$F#f+/"''3...r   c                 h    d} | |      }|d   j                   dk(  sJ |d   j                   dk(  sJ y )Nu?   Хлібі́в не було́. Хлібі́в не було́.r   u
   було́r#   r=   r>   r   s      r   *test_uk_tokenizer_handles_final_diacriticsrC      s=    LD$F!9>>\)))!9>>S   r   )pytest
PUNCT_OPENPUNCT_CLOSEPUNCT_PAIREDmarkparametrizer   r   r   r%   r(   r+   r-   r/   r2   r6   r;   r?   rA   rC    r   r   <module>rK      s%   !
"? !12$ 3$
 *-
u" ." +.
u# /# *-se,
u" - .
" +.se,
u' - /
' *-
u" ." +.
u# /# +/! 0! ,0" 1" 1<@
u) A) 1<@3j\B
u	* C A
	* 
{!!
/!r   