o
     Pj                  	   @  sT  d dl mZ d dlZd dlZd dlZd dlZd dlZd dlZd dlZd dl	Z	d dl
mZmZ ddlmZ dZerHd dl
mZ ejeeB ef ZnejZd}ddZi ddddddddddddddddddd d!d"d#d$d%d&d'd(d)d*d+d,d-d.d/i d0d1d2d3d4d5d6d7d8d9d:d;d<d=d>d?d@dAdBdCdDdEdFdGdHdIdJdKdLdMdNdOdPdQdRdSdTdUdVdWdXdYZd~d[d\ZG d]d^ d^eZddcddZG dedf dfeZG dgdh dheZG didj djeZG dkdl dlZG dmdn dnZG dodp dpe e Z!G dqdr dreZ"G dsdt dtZ#G dudv dvZ$ddydzZ%G d{d| d|Z&dS )    )annotationsN)Any
NamedTuple   )	ImageFileF)IOsstrreturnbytesc                 C  s   t j| d S )N	utf_16_be)codecsBOM_UTF16_BEencode)r    r   B/home/thesage/.local/lib/python3.10/site-packages/PIL/PdfParser.pyencode_text      r         u   ˘   u   ˇ   u   ˆ   u   ˙   u   ˝   u   ˛   u   ˚   u   ˜   u   •   u   †   u   ‡   u   …   u   —   u   –   u   ƒ   u   ⁄   u   ‹   u   ›   u   −   u   ‰   u   „   u   “   u   ”   u   ‘   u   ’   u   ‚   u   ™   u   ﬁ   u   ﬂ   u   Ł   u   Œ   u   Š   u   Ÿu   Žu   ıu   łu   œu   šu   žu   €)                     bc                 C  sD   | d t tj tjkr| t tjd  dS ddd | D S )Nr    c                 s  s     | ]}t |t|V  qd S N)PDFDocEncodinggetchr).0byter   r   r   	<genexpr>O   s    zdecode_text.<locals>.<genexpr>)lenr   r   decodejoin)r>   r   r   r   decode_textK   s   rJ   c                   @  s   e Zd ZdZdS )PdfFormatErrorz\An error that probably indicates a syntactic or semantic error in the
    PDF file structureN)__name__
__module____qualname____doc__r   r   r   r   rK   R   s    rK   	conditionboolerror_messageNonec                 C  s   | st |d S r@   )rK   )rP   rR   r   r   r   check_format_conditionY   s   rT   c                   @  s   e Zd ZU ded< ded< dS )IndirectReferenceTupleint	object_id
generationN)rL   rM   rN   __annotations__r   r   r   r   rU   ^   s   
 rU   c                   @  s>   e Zd ZdddZdddZdddZdddZdddZdS )IndirectReferencer
   r	   c                 C     | j  d| j dS )N z RrW   rX   selfr   r   r   __str__d      zIndirectReference.__str__r   c                 C  s   |   dS Nus-ascii)r`   r   r^   r   r   r   	__bytes__g   s   zIndirectReference.__bytes__otherobjectrQ   c                 C  s6   | j |j urdS t|tsJ |j| jko|j| jkS )NF)	__class__
isinstancerZ   rW   rX   r_   re   r   r   r   __eq__j   s   zIndirectReference.__eq__c                 C  s
   | |k S r@   r   ri   r   r   r   __ne__p      
zIndirectReference.__ne__rV   c                 C  s   t | j| jfS r@   )hashrW   rX   r^   r   r   r   __hash__s   r   zIndirectReference.__hash__Nr
   r	   r
   r   re   rf   r
   rQ   r
   rV   )rL   rM   rN   r`   rd   rj   rk   rn   r   r   r   r   rZ   c   s    



rZ   c                   @     e Zd ZdddZdS )IndirectObjectDefr
   r	   c                 C  r[   )Nr\   z objr]   r^   r   r   r   r`   x   ra   zIndirectObjectDef.__str__Nro   )rL   rM   rN   r`   r   r   r   r   rt   w       rt   c                   @  s\   e Zd ZdddZdd	d
ZdddZdddZd ddZd!ddZd"ddZ	d#ddZ
dS )$	XrefTabler
   rS   c                 C  s    i | _ i | _ddi| _d| _d S )Nr   i   F)existing_entriesnew_entriesdeleted_entriesreading_finishedr^   r   r   r   __init__}   s   

zXrefTable.__init__keyrV   valuetuple[int, int]c                 C  s6   | j r	|| j|< n|| j|< || jv r| j|= d S d S r@   )rz   rx   rw   ry   r_   r|   r}   r   r   r   __setitem__   s   

zXrefTable.__setitem__c                 C  s*   z| j | W S  ty   | j|  Y S w r@   )rx   KeyErrorrw   r_   r|   r   r   r   __getitem__   s
   zXrefTable.__getitem__c                 C  s   || j v r| j | d d }| j |= || j|< d S || jv r.| j| d d }|| j|< d S || jv r:| j| }d S d| d}t|)Nr   z
object ID z+ cannot be deleted because it doesn't exist)rx   ry   rw   
IndexError)r_   r|   rX   msgr   r   r   __delitem__   s   


zXrefTable.__delitem__rQ   c                 C  s   || j v p	|| jv S r@   )rw   rx   r   r   r   r   __contains__   ra   zXrefTable.__contains__c                 C  s.   t t| j t| j B t| j B S r@   )rG   setrw   keysrx   ry   r^   r   r   r   __len__   s   zXrefTable.__len__set[int]c                 C  s*   t | j t | j  t | j B S r@   )r   rw   r   ry   rx   r^   r   r   r   r      s   zXrefTable.keysf	IO[bytes]c              	   C  sF  t t| j t| j B }t t| j }| }|d |rd }t|D ]\}}|d u s8|d |kr;|}q*|d | }||d  } n|}g }|d|d t|f  |D ]B}	|	| jv rn|d| j|	   q\|	d}
t
|	|
kd|	 d|
  z|d }W n ty   d}Y nw |d|| j|	 f  q\|s$|S )	Ns   xref
r   s   %d %d
r   s   %010d %05d n 
z*expected the next deleted object ID to be z, instead found s   %010d %05d f 
)sortedr   rx   r   ry   tellwrite	enumeraterG   poprT   r   )r_   r   r   deleted_keys	startxrefprevindexr|   contiguous_keysrW   this_deleted_object_idnext_in_linked_listr   r   r   r      sJ    


 zXrefTable.writeNr
   rS   )r|   rV   r}   r~   r
   rS   )r|   rV   r
   r~   )r|   rV   r
   rS   )r|   rV   r
   rQ   rr   )r
   r   )r   r   r
   rV   )rL   rM   rN   r{   r   r   r   r   r   r   r   r   r   r   r   rv   |   s    







rv   c                   @  s|   e Zd ZU ded< d ddZd!d	d
Zd"ddZd#ddZd!ddZe	d$ddZ
eedddd dD  Zd%ddZdS )&PdfNamer   namePdfName | bytes | strr
   rS   c                 C  s:   t |tr|j| _d S t |tr|| _d S |d| _d S rb   )rh   r   r   r   r   )r_   r   r   r   r   r{      s
   


zPdfName.__init__r	   c                 C  s   | j dS rb   )r   rH   r^   r   r   r   name_as_str      zPdfName.name_as_strre   rf   rQ   c                 C  s    t |to
|j| jkp|| jkS r@   )rh   r   r   ri   r   r   r   rj      s   zPdfName.__eq__rV   c                 C  s
   t | jS r@   )rm   r   r^   r   r   r   rn      rl   zPdfName.__hash__c                 C  s   | j j dt| j dS )N())rg   rL   reprr   r^   r   r   r   __repr__      zPdfName.__repr__datac                 C  s   | t |S r@   )	PdfParserinterpret_name)clsr   r   r   r   from_pdf_stream   s   zPdfName.from_pdf_stream!      c                 C  s   h | ]}t |qS r   )ord)rD   cr   r   r   	<setcomp>   s    zPdfName.<setcomp>z#%/()<>[]{}c                 C  s@   t d}| jD ]}|| jv r|| q|d|  qt|S )N   /s   #%02X)	bytearrayr   allowed_charsappendextendr   )r_   resultr>   r   r   r   rd      s   

zPdfName.__bytes__N)r   r   r
   rS   ro   rq   rr   )r   r   r
   r   rp   )rL   rM   rN   rY   r{   r   rj   rn   r   classmethodr   r   ranger   rd   r   r   r   r   r      s   
 




r   c                   @  rs   )PdfArrayr
   r   c                 C  s   dd dd | D  d S )Ns   [     c                 s  s    | ]}t |V  qd S r@   )pdf_repr)rD   xr   r   r   rF         z%PdfArray.__bytes__.<locals>.<genexpr>s    ])rI   r^   r   r   r   rd     s   zPdfArray.__bytes__Nrp   )rL   rM   rN   rd   r   r   r   r   r     ru   r   c                   @  s*   e Zd ZdddZdd
dZdddZdS )PdfDictr|   r	   r}   r   r
   rS   c                 C  s.   |dkrt j| || d S || |d< d S )Nr   rc   )collectionsUserDict__setattr__r   r   r   r   r   r   	  s   zPdfDict.__setattr__str | time.struct_timec              
   C  s  z	| | d }W n ty } zt||d }~ww t|tr$t|}|dr|dr4|dd  }d}t|dkrZ|d }t	|dd d	 }t|d
krZ|t	|dd
 7 }dd t|d  }t
|d t|d  |}|dv r|d	9 }|dkr|d9 }t
t|| }|S )Nrc   DatezD:   Z         <         z%Y%m%d%H%M%S)+-r   )r   r   AttributeErrorrh   r   rJ   endswith
startswithrG   rV   timestrptimegmtimecalendartimegm)r_   r|   r}   erelationshipoffsetformatr   r   r   __getattr__  s2   



zPdfDict.__getattr__r   c                 C  sn   t d}|  D ]%\}}|d u rqt|}|d |tt| |d || q|d t|S )N   <<   
r   s   
>>)r   itemsr   r   r   r   )r_   outr|   r}   r   r   r   rd   *  s   


zPdfDict.__bytes__N)r|   r	   r}   r   r
   rS   )r|   r	   r
   r   rp   )rL   rM   rN   r   r   rd   r   r   r   r   r     s    

r   c                   @  s    e Zd ZdddZddd	Zd
S )	PdfBinaryr   list[int] | bytesr
   rS   c                 C  s
   || _ d S r@   )r   )r_   r   r   r   r   r{   9  rl   zPdfBinary.__init__r   c                 C  s   dd dd | jD  S )Ns   <%s>    c                 s  s    | ]}d | V  qdS )s   %02XNr   rD   r>   r   r   r   rF   =  r   z&PdfBinary.__bytes__.<locals>.<genexpr>)rI   r   r^   r   r   r   rd   <  r   zPdfBinary.__bytes__N)r   r   r
   rS   rp   )rL   rM   rN   r{   rd   r   r   r   r   r   8  s    
r   c                   @  s&   e Zd ZdddZejfdddZdS )	PdfStream
dictionaryr   bufr   r
   rS   c                 C  s   || _ || _d S r@   )r   r   )r_   r   r   r   r   r   r{   A  s   
zPdfStream.__init__
max_lengthrV   c                 C  st   z| j d }W n ty   | j Y S w |dkr.t }|| j|}|jr,d}t||S dt| d}t	|)Ns   Filters   FlateDecodezDecompressed data too largezstream filter z unknown/unsupported)
r   r   r   zlibdecompressobj
decompressunconsumed_tail
ValueErrorr   NotImplementedError)r_   r   filterdobj	plaintextr   r   r   r   rH   E  s   
zPdfStream.decodeN)r   r   r   r   r
   rS   )r   rV   r
   r   )rL   rM   rN   r{   r   	SAFEBLOCKrH   r   r   r   r   r   @  s    
r   r   r   c                 C  s  | du rdS | du rdS | d u rdS t | ttttfrt| S t | ttfr-t| 	dS t | t
jr@dt
d| 	d d	 S t | trKtt| S t | trVtt| S t | tratt| S t | tr~| d
d} | dd} | d	d} d|  d	 S t| S )NTs   trueFs   falses   nullrc   s   (D:z%Y%m%d%H%M%SZ   )   \s   \\   (   \(s   \))rh   r   r   r   r   r   rV   floatr	   r   r   struct_timestrftimedictlistr   r   replace)r   r   r   r   r   V  s0   



r   c                   @  sZ  e Zd ZdZ					ddddZdddZdddZdddZdddZdddZ	ddd Z
dd!d"Zdd$d%Zdd'd(Zdd)d*Z	ddd-d.Zdd4d5Zdd6d7Zdd8d9Zedd<d=Zdd>d?ZdddBdCZdDZdEZdFZdGZedH ZedI ZdJZdKZee e Ze !edL e dM e dN e dO e dP e dQ e j"Z#e !edL e dR e dN e dO e dP e e j"Z$ddSdTZ%	dddXdYZ&e !eZ'e !edZ e d[ Z(e !ed\ Z)e !ed] e Z*e+ddadbZ,e !dcZ-e+dddidjZ.e !edk e d[ Z/e !edl e d[ Z0e !edm e d[ Z1e !edn e d[ Z2e !edo e d[ Z3e !edp Z4e !edq Z5e !edr e ds Z6e !edt Z7e !edu e du e dv e d[ Z8e !edu e du e dw e d[ Z9e !edx e d[ Z:e !dye dz e d{ Z;e !ed| Z<e !ed} e d[ Z=e+		~ddddZ>e !dZ?i dddddddddddydyd[d[dde@dde@dde@dde@dde@dde@dydye@d[d[e@ddZAe+dddZBe !ed e ZCe !edO e dO e e ZDe !dZEdddZFddddZG	ddddZHdS )r   z|Based on
    https://www.adobe.com/content/dam/acom/en/devnet/acrobat/pdfs/PDF32000_2008.pdf
    Supports PDF up to 1.4
    Nr   rbfilename
str | Noner   IO[bytes] | Noner   bytes | bytearray | Nonestart_offsetrV   moder	   r
   rS   c                 C  sB  |r
|r
d}t ||| _|| _|| _|| _d| _d| _|d ur/|d u r/t|| | _}d| _|d urG| || _d| _|sGt	|drG|j
| _i | _|  |  |  |  | jrhz|   W n7 tyg   |    w d | _| _t | _d | _t | _d | _t | _g | _g | _d | _d | _i | _t | _d| j_|r|   d S d S )Nz4specify buf or f or filename, but not both buf and fFTr   r   ) RuntimeErrorr   r   r   r   should_close_bufshould_close_fileopenget_buf_from_filehasattrr   cached_objectsread_pdf_inforK   closefile_size_totalfile_size_thisr   rootroot_refinfoinfo_refpage_tree_rootpages
orig_pages	pages_reflast_xref_section_offsettrailer_dictrv   
xref_tablerz   seek_end)r_   r   r   r   r   r   r   r   r   r   r{   y  sZ   zPdfParser.__init__c                 C  s   | S r@   r   r^   r   r   r   	__enter__  s   zPdfParser.__enter__argsrf   c                 G  s   |    d S r@   )r	  )r_   r  r   r   r   __exit__  r   zPdfParser.__exit__c                 C  s   |    |   d S r@   )	close_bufr  r^   r   r   r   start_writing  s   zPdfParser.start_writingc                 C  s:   t | jtr| j  nt | jtjr| j  d | _d S r@   )rh   r   
memoryviewreleasemmapr	  r^   r   r   r   r    s
   

zPdfParser.close_bufc                 C  s:   | j r|   | jd ur| jr| j  d | _d S d S d S r@   )r  r  r   r  r	  r^   r   r   r   r	    s   

zPdfParser.closec                 C  s"   | j d usJ | j dtj d S )Nr   )r   seekosSEEK_ENDr^   r   r   r   r    s   zPdfParser.seek_endc                 C  s   | j d usJ | j d d S )Ns	   %PDF-1.4
)r   r   r^   r   r   r   write_header  s   zPdfParser.write_headerr   c                 C  s*   | j d usJ | j d| d  d S )Nz% 
)r   r   r   )r_   r   r   r   r   write_comment  s   zPdfParser.write_commentrZ   c                 C  sz   | j d usJ |   | | j  | _| d| _|   | j| jtd| jd | j| jtdt	| j
| j
d | jS )Nr      Catalog)TypePages   Pages)r'  CountKids)r   del_rootnext_object_idr   r  r  rewrite_pages	write_objr   rG   r  r^   r   r   r   write_catalog  s   zPdfParser.write_catalogc                 C  s   g }t | jD ]N\}}| j| }| j|j= ||td  || jvr$qi }| D ]
\}}|||	 < q*| j
|d< | jdi |}t | jD ]\}	}
|
|krT|| j|	< qGq|D ]}|rt| j| }|j| jv rl| j|j= |dd }|s\qXg | _d S )Ns   ParentParentr@   )r   r  r  r  rW   r   r   r  r   r   r  
write_pagerB   )r_   pages_tree_nodes_to_deleteipage_ref	page_infostringified_page_infor|   r}   new_page_refjcur_page_refpages_tree_node_refpages_tree_noder   r   r   r.    s4   







zPdfParser.rewrite_pagesnew_root_refIndirectReference | Nonec                 C  s   | j d usJ |r|   || _| jr| d | j| _| j| j }t| j}| j|d}| j	d ur7| j	|d< | jr?| j|d< || _	| j dt
t| d|   d S )N)   Root   Size   Prev   Infos   trailer
s   
startxref
%d
%%%%EOF)r   r,  r  r  r/  r  r  r   rG   r  r   r   )r_   r=  
start_xrefnum_entriesr  r   r   r   write_xref_and_trailer  s.   




z PdfParser.write_xref_and_trailerrefint | IndirectReference | Noneobjsr   dict_objc                 O  sV   t |tr
| j| n|}d|vrtd|d< d|vr| j|d< | j|g|R i |S )Nr'     Pager1  )rh   rV   r  r   r  r/  )r_   rF  rH  rI  obj_refr   r   r   r2    s   
zPdfParser.write_pagec                 O  s   | j d usJ | j }|d u r| | }n| |jf| j|j< |tt|  |	dd }|d ur:t
||d< |rC|t| |D ]	}|t| qE|d urb|d || |d |d |S )NstreamLengths   stream
s   
endstream
s   endobj
)r   r-  r   rX   r  rW   r   r   rt   r   rG   r   )r_   rF  rH  rI  r   rL  objr   r   r   r/  "  s&   



zPdfParser.write_objc                 C  s.   | j d u rd S | j| j j= | j| jd j= d S )Nr)  )r  r  rW   r  r^   r   r   r   r,  :  s   
zPdfParser.del_rootr   bytes | mmap.mmapc                 C  sR   t | dr	|  S t | dr|  S ztj|  dtjdW S  ty(   Y dS w )N	getbuffergetvaluer   )accessr   )r  rP  rQ  r  filenoACCESS_READr   )r   r   r   r   r  @  s   

zPdfParser.get_buf_from_filec                 C  sB  | j d usJ t| j | _| j| j | _|   t| jdd ud | jd | _	| j	d us0J | jdd | _
t| | j	| _| j
d u rKt | _n	t| | j
| _td| jv d t| jd dkd t| jdd ud	 tt| jd td
 | jd | _| jd usJ | | j| _| | j| _| jd d  | _d S )Nr?  zRoot is missingrB     Typez/Type missing in Rootr&  z/Type in Root is not /Catalogr)  z/Pages missing in Rootz+/Pages in Root is not an indirect reference)r   rG   r
  r   r  read_trailerrT   r  rB   r  r  r   read_indirectr  r  rh   rZ   r  r  linearize_page_treer  r  r^   r   r   r   r  L  s<   

zPdfParser.read_pdf_infor   
int | Nonec                 C  sV   zt t| j d d}W n ty   t dd}Y nw |d ur)|df| j|j< |S )Nr   r   )rZ   maxr  r   r   rW   )r_   r   	referencer   r   r   r-  p  s   zPdfParser.next_object_ids   [][()<>{}/%]s$   [][()<>{}/%\000\011\012\014\015\040]s   [\000\011\012\014\015\040]s#   [\000\011\012\014\015\0400-9a-fA-F]   *   +s   [\000\011\014\040]*s   [\r\n]+s   trailers   <<(.*>>)s	   startxrefs   ([0-9]+)s   %%EOF   $s	   <<(.*?>>)c                 C  s   | j d usJ t| j d }|| jk r| j}| j| j |}t|d ud |}|r9|}| j| j | d }|s)|s=|}|d usCJ |d}t|d| _	| 
|| _t | _| j| j	d d| jv rp| | jd  d S d S )N @  ztrailer end not found   r   r   xref_section_offsetrA  )r   rG   r   re_trailer_endsearchrT   startgrouprV   r  interpret_trailerr  rv   r  read_xref_tableread_prev_trailer)r_   search_start_offsetm
last_matchtrailer_datar   r   r   rV    s,   


zPdfParser.read_trailerrb  processed_offsetslist[int] | Nonec                 C  s   | j d usJ | j|d}| j| j ||d  }t|d ud |d us'J |d}tt|d|kd | |}d|v r_|d u rGg }|| t|d |vd | 	|d | d S d S )	Nra  r_  zprevious trailer not foundr   r   zGxref section offset in previous trailer doesn't match what was expectedrA  ztrailer loop found)
r   rh  re_trailer_prevrd  rT   rf  rV   rg  r   ri  )r_   rb  rn  trailer_offsetrk  rm  r  r   r   r   ri    s,   


zPdfParser.read_prev_trailers   /([!-$&'*-.0-;=?-Z\\^-z|~]+)(?=r   r   s   >>rm  r   dict[bytes, Any]c                 C  s   i }d}	 | j ||}|s-| j||}t|d uo!| t|kdt||d    n%| |d}t	|t
s<J | || \}}|||< |d u rOn|}qtd|v o]t	|d td td|v okt	|d td |S )	Nr   Tz+name not found in trailer, remaining data: r   r@  z&/Size not in trailer or not an integerr?  z1/Root not in trailer or not an indirect reference)re_namematchre_dict_endrT   endrG   r   r   rf  rh   r   	get_valuerV   rZ   )r   rm  trailerr   rk  r|   r}   value_offsetr   r   r   rg    s<   zPdfParser.interpret_trailers   ([^#]*)(#([0-9a-fA-F]{2}))?Frawas_textrQ   str | bytesc                 C  sj   d}| j |D ]!}|dr"||dt|dd 7 }q||d7 }q|r1|dS t|S )Nr      r   rc   zutf-8)re_hashes_in_namefinditerrf  r   fromhexrH   r   )r   rz  r{  r   rk  r   r   r   r     s   
&
zPdfParser.interpret_names   null(?=s   true(?=s   false(?=s   ([-+]?[0-9]+)(?=s)   ([-+]?([0-9]+\.[0-9]*|[0-9]*\.[0-9]+))(?=s   \[   ]s   <(s   *)>r   s   ([-+]?[0-9]+)s   R(?=s   obj(?=s	   endobj(?=r   s	   %[^\r\n]*s   )*s   stream\r?\ns   endstream(?=r   r   *bytes | bytearray | memoryview | mmap.mmapexpect_indirectmax_nestingtuple[Any, int | None]c                 C  s  |dkrdS | j ||}|r| }| j||}|r|tt|ddkd tt|ddkd t|d u pI|tt|dt|dkd | j|| |d d\}}|d u rb|d fS | j	||}t|d ud	 |d usvJ || fS t| d
 | j
||}|rtt|ddkd tt|ddkd tt|dt|d| fS | j||}|rp| }i }| j||}|}	|s|	d usJ | j||	|d d\}
}	|	d u r|d fS | j||	|d d\}}	|||
< |	d u r|d fS | j||	}|r| }	| j||	}|rj|d}|d u s,t|ts6d| d}t|t|| | |  }| j|| | }t|d ud |d us]J | }	tt|||	fS t||	fS | j||}|r| }g }| j||}|}	|s|	d usJ | j||	|d d\}}	|| |	d u r|d fS | j||	}|r|| fS | j||}|rd | fS | j||}|rd| fS | j||}|rd| fS | j||}|rt| |d| fS | j||}|rt|d| fS | j||}|r/t |d| fS | j!||}|rat"dd |dD }t#|d dkrU|t$d t"%|&d| fS | j'||}|rs| (|| S dt)|||d   }t|)Nr   )NNr   z<indirect object definition: object ID must be greater than 0r   z;indirect object definition: generation must be non-negativez2indirect object definition different than expected)r  z(indirect object definition end not foundz$indirect object definition not foundz;indirect object reference: object ID must be greater than 0z:indirect object reference: generation must be non-negatives   Lengthz&bad or missing Length in stream dict (r   zstream end not foundTFc                 s  s    | ]	}|d v r|V  qdS )s   0123456789abcdefABCDEFNr   r   r   r   r   rF     s    z&PdfParser.get_value.<locals>.<genexpr>   0rc   zunrecognized object:     )*
re_commentrt  rv  re_indirect_def_startrT   rV   rf  rZ   rw  re_indirect_def_endre_indirect_referencere_dict_startru  re_stream_startrB   rh   rK   r   re_stream_endr   r   re_array_startre_array_endr   re_nullre_truere_falsers  r   r   re_intre_realr   re_string_hexr   rG   r   r  rH   re_string_litget_literal_stringr   )r   r   r   r  r  rk  rf   object_offsetr   current_offsetr|   r}   
stream_lenr   stream_dataresults
hex_stringr   r   r   rw  ;  s  
&









	zPdfParser.get_valuesF   (\\[nrtbf()\\])|(\\[0-9]{1,3})|(\\(\r\n|\r|\n))|(\r\n|\r|\n)|(\()|(\))   nr      r      t   	   b      f   r   tuple[bytes, int]c                 C  s  d}t  }| j||D ]u}||||   |dr,|| j|dd   nQ|drA|t|ddd  d n<|drGn6|drR|d n+|dra|d	 |d7 }n|d
r}|dkrtt	||
 f  S |d |d8 }|
 }qd}t|)Nr   r   r      r}     r      r      r   zunfinished literal string)r   re_lit_str_tokenr  r   re  rf  escaped_charsr   rV   r   rv  rK   )r   r   r   nesting_depthr   rk  r   r   r   r   r    s.   

 







zPdfParser.get_literal_strings   xrefs+   ([0-9]{10}) ([0-9]{5}) ([fn])( \r| \n|\r\n)c                 C  s&  | j d usJ d}| j| j || j }t|d ud |d us!J | }	 | j| j |}|s8t|d 	 |S d}| }t|d}t|d}t	||| D ]>}| j
| j |}t|d ud |d usjJ | }|dd	k}|st|d}	t|d|	f}
|| jvr|
| j|< qSq&)
NFzxref section start not foundTzxref subsection start not foundr   r   zxref entry not foundr}  r  )r   re_xref_section_startrt  r   rT   rv  re_xref_subsection_startrV   rf  r   re_xref_entryr  )r_   rb  subsection_foundrk  r   first_objectnum_objectsr4  is_freerX   	new_entryr   r   r   rh    sB   

zPdfParser.read_xref_tablec              
   C  s   | j |d  \}}t||d kd|d  d|d  d| d|  | jd us)J | j| j|| j t| |dd }|| j|< |S )Nr   r   zexpected to find generation z for object ID z) in xref table, instead found generation z at offset )r  r  )r  rT   r   rw  r   rZ   r  )r_   rF  r  r   rX   r}   r   r   r   rW  %  s(   

zPdfParser.read_indirectnodePdfDict | Nonelist[IndirectReference]c                 C  sn   |d ur|n| j }t|d dkd g }|d D ]}| |}|d dkr+|| q|| j|d q|S )NrU  r)  z%/Type of page tree node is not /Pagess   KidsrJ  )r  )r  rT   rW  r   r   rX  )r_   r  	page_noder  kid
kid_objectr   r   r   rX  6  s   
zPdfParser.linearize_page_tree)NNNr   r   )r   r   r   r   r   r   r   rV   r   r	   r
   rS   )r
   r   )r  rf   r
   rS   r   )r   r	   r
   rS   )r
   rZ   r@   )r=  r>  r
   rS   )rF  rG  rH  r   rI  r   r
   rZ   )rF  r>  rH  r   rI  r   r
   rZ   )r   r   r
   rO  )r   rY  r
   rZ   )rb  rV   rn  ro  r
   rS   )rm  r   r
   rr  )F)rz  r   r{  rQ   r
   r|  )Nr   )
r   r  r   rV   r  r>  r  rV   r
   r  )r   r  r   rV   r
   r  )rb  rV   r
   rV   )r   )rF  rZ   r  rV   r
   r   )r  r  r
   r  )IrL   rM   rN   rO   r{   r  r  r  r  r	  r  r#  r%  r0  r.  rE  r2  r/  r,  staticmethodr  r  r-  	delimiterdelimiter_or_ws
whitespacewhitespace_or_hexwhitespace_optionalwhitespace_mandatorywhitespace_optional_no_nlnewline_onlynewlinerecompileDOTALLrc  rp  rV  ri  re_whitespace_optionalrs  r  ru  r   rg  r~  r   r  r  r  r  r  r  r  r  r  r  r  r  r  r  r  rw  r  r   r  r  r  r  r  rh  rW  rX  r   r   r   r   r   s  s   
5













$
	
	





 	


!r   )r   r	   r
   r   )r>   r   r
   r	   )rP   rQ   rR   r	   r
   rS   )r   r   r
   r   )'
__future__r   r   r   r   r  r!  r  r   r   typingr   r   r?   r   TYPE_CHECKINGr   r   r	   r   	_DictBaser   rA   rJ   r  rK   rT   rU   rZ   rt   rv   r   r   r   r   r   r   r   r   r   r   r   r   <module>   s    
	
 !"#
-
^)0
