
    'h                     V   d dl Z d dlZd dlZd dlmZ d dlmZ dedefdZdefdZ	e
d	k    rz ee j                  d
k     r ed            e j        d           e j        d         Z ee          Z e	e          Zes ed           dS eD ]\  ZZ ede de            dS dS )    N)	PdfReader)BytesIOpdf_pathreturnc                 `   |                      d          s|                      d          rJt          j        |           }|                                 t	          t          |j                            }nt	          |           }d}|j        D ] }|                                pd}||dz   z  }!|S )uD   PDF에서 전체 텍스트 추출 (URL/파일 경로 둘 다 지원)zhttp://zhttps:// 
)	
startswithrequestsgetraise_for_statusr   r   contentpagesextract_text)r   rreadertextpagets         +/var/www/html/bot/scripts/pdf_diff_names.pyextract_text_from_pdfr      s    9%% %)<)<Z)H)H %L""	719--..8$$D  %2DK    r   c                     ddg}g }|D ][}t          j        ||           D ]C}|                    d          |                    d          }}|                    ||f           D\|S )Nu@   변경\s*전[:：]?\s*([^\s/]+).*변경\s*후[:：]?\s*([^\s/]+)u)   ([가-힣]{2,4})\s*→\s*([가-힣]{2,4})      )refinditergroupappend)r   patternsresultspatmbeforeafters          r   find_name_changesr&      s    K4H G , ,S$'' 	, 	,AGGAJJ

EFNNFE?++++	, Nr   __main__r   z9Usage: python -m scripts.pdf_diff_names <pdf_path_or_url>r   u>   [INFO] 변경 전/후 이름 패턴을 찾지 못했습니다.u   변경 전: u    → 변경 후: )sysr   r   pypdfr   ior   strr   r&   __name__lenargvprintexitr   r   changesr$   r%    r   r   <module>r3      s[   



 				             C C    
C 
 
 
 
 z
s38}}qIJJJx{H  **D%%G CNOOOOO$ 	C 	CMFEEAAA%AABBBB 	C 	Cr   