
    +sgj                    8    d dl mZ d dlZddlmZ  G d d      Zy)    )annotationsN   )InputExamplec                      e Zd ZdZd ZddZy)PairedFilesReaderz/Reads in the a Pair Dataset, split in two filesc                    || _         y )N)	filepaths)selfr	   s     b/var/www/html/venv/lib/python3.12/site-packages/sentence_transformers/readers/PairedFilesReader.py__init__zPairedFilesReader.__init__   s	    "    c           	        g }| j                   D ]I  }|j                  d      rt        j                  |dd      nt        |d      }|j	                  |       K g }d}|sg }|D ],  }|j                         }|dk(  rd} n|j	                  |       . |r	 |S |j	                  t        t        t        |            |d	             |d
kD  rt        |      |k\  r	 |S |s|S )Nz.gzrtzutf-8)encodingF Tr   )guidtextslabelr   )	r	   endswithgzipopenappendreadliner   strlen)	r
   max_examplesfInsfilepathfInexampleseofr   texts	            r   get_exampleszPairedFilesReader.get_examples   s     	H $$U+ 		(D7;(W5 
 KK	 E #||~2:CT"#  	 OOLc#h-.@UVWXaCM\$A% $ r   N)r   )__name__
__module____qualname____doc__r   r#    r   r   r   r      s    9#r   r   )
__future__r   r   r   r   r   r(   r   r   <module>r*      s    "  % %r   