o
    ۩Zh7                     @   sr  d Z ddlZddlZddlZddlZddlmZmZm	Z	m
Z
mZmZmZmZmZ ddlmZ ddlZddlmZmZmZ ddlmZ ddlmZ ddlmZmZ dd	lmZmZm Z m!Z! dd
l"m#Z#m$Z$m%Z% ddl&m'Z' e(  e)e*Z+e,dZ-dee.e/f de.fddZ0d3dede1dee. ddfddZ2	d4dedede3ddfddZ4		d5dededee. de3ddf
ddZ5				d6dede.d ed!ee6 d"e.d#e3dee. d$ee. ddfd%d&Z7e%d'Z8e%d(Z9de.d"e.d$e.ddfd)d*Z:					d7dede.d e
e6 d!ee6 d"e.d#e3dee. d$ee. de3ddfd+d,Z;defd-d.Z<d3d/eee.  ddfd0d1Z=e*d2kr7e=  dS dS )8z#Extract pdf structure in XML format    N)	Any	ContainerDictIterableListOptionalTextIOUnioncast)ArgumentParser)PDFDocumentPDFNoOutlinesPDFXRefFallback)PDFPage)	PDFParser)PDFObjectNotFoundPDFValueError)	PDFStream	PDFObjRefresolve1stream_value)	PSKeyword	PSLiteralLIT)isnumberz&[\000-\037&<>()"\042\047\134\177-\377]sreturnc                 C   s*   t | trt| d}n| }tdd |S )Nzlatin-1c                 S   s   dt | d S )Nz&#%d;r   )ordgroup)m r    K/var/www/html/lang_env/lib/python3.10/site-packages/../../../bin/dumppdf.py<lambda>   s    zescape.<locals>.<lambda>)
isinstancebytesstrESC_PATsub)r   usr    r    r!   escape   s   
r)   outobjcodecc                 C   s  |d u r|  d d S t|tr?|  dt|  | D ]\}}|  d|  |  d t| | |  d q|  d d S t|trc|  dt|  |D ]}t| | |  d qO|  d	 d S t|ttfry|  d
t|t	|f  d S t|t
r|dkr|  |  d S |dkr|  |  d S |  d t| |j |  d |dkr| }|  dt|t	|f  |  d d S t|tr|  d|j  d S t|tr|  d|j  d S t|tr|  d|j  d S t|r|  d|  d S t|)Nz<null />z<dict size="%d">
z<key>%s</key>
z<value>z	</value>
z</dict>z<list size="%d">

z</list>z<string size="%d">%s</string>rawbinaryz<stream>
<props>
z

</props>
textz<data size="%d">%s</data>
z	</stream>z<ref id="%d" />z<keyword>%s</keyword>z<literal>%s</literal>z<number>%s</number>)writer#   dictlenitemsdumpxmllistr%   r$   r)   r   Zget_rawdataget_dataattrsr   objidr   namer   r   	TypeError)r*   r+   r,   kvdatar    r    r!   r5   !   sd   








	





r5   Fdocshow_fallback_xrefc                 C   sj   |j D ]}t|tr|r| d t| |  | d qtdd |j D }|r3|s3d}t| d S )Nz
<trailer>
z
</trailer>

c                 s   s    | ]}t |tV  qd S N)r#   r   ).0xrefr    r    r!   	<genexpr>j   s    zdumptrailers.<locals>.<genexpr>zThis PDF does not have an xref. Use --show-fallback-xref if you want to display the content of a fallback xref that contains all objects.)	xrefsr#   r   r1   r5   Zget_trailerallloggerwarning)r*   r?   r@   rC   Zno_xrefsmsgr    r    r!   dumptrailersb   s   



rJ   c           	      C   s   t  }| d |jD ]K}| D ]D}||v rq|| z ||}|d u r)W q| d|  t| ||d | d W q tyU } ztd|  W Y d }~qd }~ww qt	| || | d d S )Nz<pdf>z<object id="%d">
r,   z
</object>

znot found: %rz</pdf>)
setr1   rE   
get_objidsaddgetobjr5   r   printrJ   )	r*   r?   r,   r@   visitedrC   r9   r+   er    r    r!   dumpallobjsu   s,   




rS    outfpfnameobjidspagenospassworddumpall
extractdirc                    sx  t |d}t|}	t|	| dd tt dD }
dtdtf fdd}z  }| 	d	 |D ]o\}}}}}d }|rI||}|
|d
 j
 }n(|rq|}t|trq|d}|rqt|dkrq|drq||d }|
|d
 j
 }t|}| 	d|| |d ur| 	d t| | | 	d |d ur| 	d|  | 	d q2| 	d W n	 ty   Y nw |	  |  d S )Nrbc                 S   s   i | ]\}}|j |qS r    )Zpageid)rB   pagenopager    r    r!   
<dictcomp>   s    zdumpoutline.<locals>.<dictcomp>   destr   c                    s`   t | ttfrt | } nt | trt | j} t | tr%| d } t | tr.| 	 } | S )ND)
r#   r%   r$   r   Zget_destr   r:   r2   r   resolve)ra   r?   r    r!   resolve_dest   s   


z!dumpoutline.<locals>.resolve_destz<outlines>
r   Sz/'GoTo'rb   z"<outline level="{!r}" title="{}">
z<dest>z</dest>
z<pageno>%r</pageno>
z</outline>
z</outlines>
)openr   r   	enumerater   create_pagesobjectr   Zget_outlinesr1   r9   r#   r2   getreprr)   formatr5   r   close)rU   rV   rW   rX   rY   rZ   r,   r[   fpparserZpagesre   Zoutlinesleveltitlera   aser]   actionsubtyper   r    rd   r!   dumpoutline   sL   








rw   ZFilespecZEmbeddedFilec           
         s   dt dtttf dd f fdd}t| dA}t|}t|| t } jD ](}|	 D ]!} 
|}	||vrNt|	trN|	dtu rN|| |||	 q-q'W d    d S 1 s[w   Y  d S )Nr9   r+   r   c                    s   t j|dptt|d }|d dp |d d} |j}t	|t
s4d| }t||dturAtd| t jd| |f }t j|rXtd| td	|  t jt j|d
d t|d}||  |  d S )NZUFFZEFz:unable to process PDF: reference for %r is not a PDFStreamTypez>unable to process PDF: reference for %r is not an EmbeddedFilez%.6d-%szfile exists: %rzextracting: %rT)exist_okwb)ospathbasenamerk   r
   r$   decoderO   r9   r#   r   r   LITERAL_EMBEDDEDFILEjoinexistsIOErrorrP   makedirsdirnamerg   r1   r7   rn   )r9   r+   filenameZfilereffileobj	error_msgr}   r*   r?   r[   r    r!   extract1   s0   &

z!extractembedded.<locals>.extract1r\   ry   )intr   r%   r   rg   r   r   rL   rE   rM   rO   r#   r2   rk   LITERAL_FILESPECrN   )
rV   rY   r[   r   ro   rp   Zextracted_objidsrC   r9   r+   r    r   r!   extractembedded   s*   $





r   c	                 C   s   t |d}	t|	}
t|
|}|r!|D ]}||}t| ||d q|rMtt|D ]"\}}||v rL|rF|jD ]}t	|}t| ||d q7q*t| |j
 q*|rVt| ||| |sb|sb|sbt| || |	  |dvro| d d S )Nr\   rK   )r.   r/   r-   )rg   r   r   rO   r5   rh   r   ri   contentsr   r8   rS   rJ   rn   r1   )rU   rV   rW   rX   rY   rZ   r,   r[   r@   ro   rp   r?   r9   r+   r]   r^   r    r    r!   dumppdf   s2   




r   c                  C   sd  t tdd} | jdtd ddd | jddd	d
tjd | jdddddd |  }|jdddddd |jddtdd | jddd}|jdt	d ddd |jddtd d |jd!d"td#d |jd$d%ddd&d |jd'dd(d) |jd*d+td,d-d. | jd/d0d}|jd1d2td3d4d. | }|jd5d6ddd7d |jd8d9ddd:d |jd;d<ddd=d | S )>NT)descriptionadd_helpfiles+zOne or more paths to PDF files.)typedefaultnargshelpz	--versionz-vversionzpdfminer.six v{})ru   r   z--debugz-dF
store_truezUse debug logging level.)r   ru   r   z--extract-tocz-TzExtract structure of outlinez--extract-embeddedz-EzExtract embedded files)r   r   ParserzUsed during PDF parsing)r   z--page-numbersz0A space-seperated list of page numbers to parse.z	--pagenosz-pzA comma-separated list of page numbers to parse. Included for legacy applications, use --page-numbers for more idiomatic argument entry.z	--objectsz-iz1Comma separated list of object numbers to extractz--allz-az3If the structure of all objects should be extractedz--show-fallback-xrefzAdditionally show the fallback xref. Use this if the PDF has zero or only invalid xref's. This setting is ignored if --extract-toc or --extract-embedded is used.)ru   r   z
--passwordz-PrT   z,The password to use for decrypting PDF file.)r   r   r   OutputzUsed during output generation.z	--outfilez-o-zJPath to file where output is written. Or "-" (default) to write to stdout.z--raw-streamz-rz%Write stream objects without encodingz--binary-streamz-bz)Write stream objects with binary encodingz--text-streamz-tz"Write stream objects as plain text)
r   __doc__add_argumentr%   rm   pdfminer__version__add_mutually_exclusive_groupadd_argument_groupr   )rp   Zprocedure_parserZparse_paramsZoutput_paramsZcodec_parserr    r    r!   create_parser!  s   
r   argvc           	      C   sB  t  }|j| d}|jrt tj |jdkrtj	}nt
|jd}|jr2dd |jdD }ng }|jr@dd |jD }n|jrOd	d |jdD }nt }|j}|jr[d
}n|jrad}n|jrgd}nd }|jD ].}|jrt||||||j|d d ql|jrt|||jd qlt||||||j|d |jd	 ql|  d S )N)argsr   wc                 S   s   g | ]}t |qS r    r   rB   xr    r    r!   
<listcomp>      zmain.<locals>.<listcomp>,c                 S   s   h | ]}|d  qS r`   r    r   r    r    r!   	<setcomp>  r   zmain.<locals>.<setcomp>c                 S   s   h | ]}t |d  qS r   r   r   r    r    r!   r     s    r.   r/   r0   )rY   rZ   r,   r[   )rY   r[   )rY   rZ   r,   r[   r@   )r   
parse_argsdebuglogging	getLoggersetLevelDEBUGoutfilesysstdoutrg   objectssplitZpage_numbersrX   rL   rY   Z
raw_streamZbinary_streamZtext_streamr   Zextract_tocrw   rF   Zextract_embeddedr   r   r@   rn   )	r   rp   r   rU   rW   rX   rY   r,   rV   r    r    r!   main  sb   


r   __main__rA   )F)NF)rT   FNN)rT   FNNF)>r   r   Zos.pathr|   rer   typingr   r   r   r   r   r   r   r	   r
   argparser   r   Zpdfminer.pdfdocumentr   r   r   Zpdfminer.pdfpager   Zpdfminer.pdfparserr   Zpdfminer.pdftypesr   r   r   r   r   r   Zpdfminer.psparserr   r   r   Zpdfminer.utilsr   basicConfigr   __name__rG   compiler&   r%   r$   r)   rj   r5   boolrJ   rS   r   rw   r   r   r   r   r   r   r    r    r    r!   <module>   s   ,

 B

 	
=0	

%v
A
