3
	jQc              2   @   s  d dl Z d dlZd dlZd dlZd dlZd dlZyd dlmZ W n ek
rZ   ejZY nX ee	krndd Z
ndd Z
dd Z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)d*d+d,d-d.d/d0d1)Zd2d3 ZG d4d5 d5eZd6d7 ZG d8d9 d9ejd:d;d<gZG d=d> d>eZG d?d@ d@ZG dAdB dBZG dCdD dDeZG dEdF dFeZG dGdH dHZG dIdJ dJZdKdL ZG dMdN dNZdS )O    N)UserDictc             C   s   | S )N )sr   r   //usr/lib/python3/dist-packages/PIL/PdfParser.py
make_bytes   s    r   c             C   s
   | j dS )Nzus-ascii)encode)r   r   r   r   r      s    c             C   s   t j| jd S )N	utf_16_be)codecsBOM_UTF16_BEr   )r   r   r   r   encode_text   s    r   u   ˘u   ˇu   ˆu   ˙u   ˝u   ˛u   ˚u   ˜u   •u   †u   ‡u   …u   —u   –u   ƒu   ⁄u   ‹u   ›u   −u   ‰u   „u   “u   ”u   ‘u   ’u   ‚u   ™u   ﬁu   ﬂu   Łu   Œu   Šu   Ÿu   Žu   ıu   łu   œu   šu   žu   €))                                                                                                                           c             C   sd   | d t tj tjkr0| t tjd  jdS ttkrLdjdd | D S djdd | D S d S )Nr    c             s   s   | ]}t jt||V  qd S )N)PDFDocEncodinggetord).0byter   r   r   	<genexpr>L   s    zdecode_text.<locals>.<genexpr>c             s   s   | ]}t j|t|V  qd S )N)r7   r8   chr)r:   r;   r   r   r   r<   N   s    )lenr	   r
   decodestrbytesjoin)br   r   r   decode_textH   s
    rD   c               @   s   e Zd ZdZdS )PdfFormatErrorzXAn error that probably indicates a syntactic or semantic error in the PDF file structureN)__name__
__module____qualname____doc__r   r   r   r   rE   Q   s   rE   c             C   s   | st |d S )N)rE   )Z	conditionZerror_messager   r   r   check_format_conditionV   s    rJ   c               @   s4   e Zd Zdd Zdd Zdd Zdd Zd	d
 ZdS )IndirectReferencec             C   s   d|  S )Nz%s %s Rr   )selfr   r   r   __str__\   s    zIndirectReference.__str__c             C   s   | j  jdS )Nzus-ascii)rM   r   )rL   r   r   r   	__bytes___   s    zIndirectReference.__bytes__c             C   s$   |j | j ko"|j| jko"|j| jkS )N)	__class__	object_id
generation)rL   otherr   r   r   __eq__b   s    zIndirectReference.__eq__c             C   s
   | |k S )Nr   )rL   rR   r   r   r   __ne__e   s    zIndirectReference.__ne__c             C   s   t | j| jfS )N)hashrP   rQ   )rL   r   r   r   __hash__h   s    zIndirectReference.__hash__N)rF   rG   rH   rM   rN   rS   rT   rV   r   r   r   r   rK   [   s
   rK   ZIndirectReferenceTuplerP   rQ   c               @   s   e Zd Zdd ZdS )IndirectObjectDefc             C   s   d|  S )Nz	%s %s objr   )rL   r   r   r   rM   m   s    zIndirectObjectDef.__str__N)rF   rG   rH   rM   r   r   r   r   rW   l   s   rW   c               @   sL   e Zd Zdd Zdd Zdd Zdd Zd	d
 Zdd Zdd Z	dd Z
dS )	XrefTablec             C   s    i | _ i | _ddi| _d| _d S )Nr   i   F)existing_entriesnew_entriesdeleted_entriesreading_finished)rL   r   r   r   __init__r   s    
zXrefTable.__init__c             C   s2   | j r|| j|< n
|| j|< || jkr.| j|= d S )N)r\   rZ   rY   r[   )rL   keyvaluer   r   r   __setitem__x   s
    

zXrefTable.__setitem__c             C   s*   y
| j | S  tk
r$   | j| S X d S )N)rZ   KeyErrorrY   )rL   r^   r   r   r   __getitem__   s    
zXrefTable.__getitem__c             C   s   || j kr0| j | d d }| j |= || j|< nR|| jkrX| j| d d }|| j|< n*|| jkrn| j| }ntdt| d d S )N   z
object ID z+ cannot be deleted because it doesn't exist)rZ   r[   rY   
IndexErrorr@   )rL   r^   rQ   r   r   r   __delitem__   s    


zXrefTable.__delitem__c             C   s   || j kp|| jkS )N)rY   rZ   )rL   r^   r   r   r   __contains__   s    zXrefTable.__contains__c             C   s.   t t| jj t| jj B t| jj B S )N)r>   setrY   keysrZ   r[   )rL   r   r   r   __len__   s    zXrefTable.__len__c             C   s*   t | jj t | jj  t | jj B S )N)rg   rY   rh   r[   rZ   )rL   r   r   r   rh      s    zXrefTable.keysc             C   s`  t t| jj t| jj B }t t| jj }|j }|jd x|rZd }xPt|D ]<\}}|d ksx|d |kr~|}q\|d | }||d  }P q\W |}d }|jtd|d t	|f  x|D ]}	|	| jkr|jtd| j|	   q|j
d}
t|	|
kd|	|
f  y|d }W n tk
r6   d}Y nX |jtd|| j|	 f  qW qHW |S )Ns   xref
rc   z%d %d
r   z%010d %05d n 
z>expected the next deleted object ID to be %s, instead found %sz%010d %05d f 
)sortedrg   rZ   rh   r[   tellwrite	enumerater   r>   poprJ   rd   )rL   frh   Zdeleted_keysZ	startxrefprevindexr^   Zcontiguous_keysrP   Zthis_deleted_object_idZnext_in_linked_listr   r   r   rl      s4     





$zXrefTable.writeN)rF   rG   rH   r]   r`   rb   re   rf   ri   rh   rl   r   r   r   r   rX   q   s   rX   c               @   sl   e Zd Zdd Zdd Zdd Zdd Zd	d
 Zedd Z	e
edde
dd dD  Zdd ZeZdS )PdfNamec             C   s6   t |tr|j| _nt |tr&|| _n|jd| _d S )Nzus-ascii)
isinstancerr   namerA   r   )rL   rt   r   r   r   r]      s
    


zPdfName.__init__c             C   s   | j jdS )Nzus-ascii)rt   r?   )rL   r   r   r   name_as_str   s    zPdfName.name_as_strc             C   s    t |tr|j| jkp|| jkS )N)rs   rr   rt   )rL   rR   r   r   r   rS      s    zPdfName.__eq__c             C   s
   t | jS )N)rU   rt   )rL   r   r   r   rV      s    zPdfName.__hash__c             C   s   dt | j S )NzPdfName(%s))reprrt   )rL   r   r   r   __repr__   s    zPdfName.__repr__c             C   s   | t j|S )N)	PdfParserinterpret_name)klassdatar   r   r   from_pdf_stream   s    zPdfName.from_pdf_stream!      c             c   s   | ]}t |V  qd S )N)r9   )r:   cr   r   r   r<      s    zPdfName.<genexpr>z#%/()<>[]{}c             C   s   t tkrNtd}x|| jD ]0}t|| jkr6|j| q|jdt|  qW n@td}x6| jD ],}|| jkrx|j| q^|jtd|  q^W t|S )N   /s   #%02Xz#%02X)	r@   rA   	bytearrayrt   r9   allowed_charsappendextendr   )rL   resultrC   r   r   r   rN      s    
zPdfName.__bytes__N)rF   rG   rH   r]   ru   rS   rV   rw   classmethodr|   rg   ranger   rN   rM   r   r   r   r   rr      s    rr   c               @   s   e Zd Zdd ZeZdS )PdfArrayc             C   s   ddj dd | D  d S )Ns   [     c             s   s   | ]}t |V  qd S )N)pdf_repr)r:   xr   r   r   r<      s    z%PdfArray.__bytes__.<locals>.<genexpr>s    ])rB   )rL   r   r   r   rN      s    zPdfArray.__bytes__N)rF   rG   rH   rN   rM   r   r   r   r   r      s   r   c               @   s8   e Zd Zdd Zdd Zdd Zdd Zeekr4eZ	d	S )
PdfDictc             O   s   t j| f|| d S )N)r   r]   )rL   argskwargsr   r   r   r]      s    zPdfDict.__init__c             C   sN   |dkr.t tdr"tj| || qJ|| j|< nt|trB|jd}|| |< d S )Nr{   __setattr__zus-ascii)hasattrr   r   __dict__rs   r@   r   )rL   r^   r_   r   r   r   r      s    


zPdfDict.__setattr__c             C   sl   y| | }W nD t k
rP   y| |jd }W n t k
rJ   t|Y nX Y nX t|trdt|S |S d S )Nzus-ascii)ra   r   AttributeErrorrs   rA   rD   )rL   r^   r_   r   r   r   __getattr__  s    
zPdfDict.__getattr__c             C   sr   t d}xV| j D ]J\}}|d kr$qt|}|jd |jtt| |jd |j| qW |jd t|S )Ns   <<   
r   s   
>>)r   itemsr   r   rA   rr   )rL   outr^   r_   r   r   r   rN     s    


zPdfDict.__bytes__N)
rF   rG   rH   r]   r   r   rN   r@   rA   rM   r   r   r   r   r      s   r   c               @   s.   e Zd Zdd Zeekr"dd Zndd ZdS )	PdfBinaryc             C   s
   || _ d S )N)r{   )rL   r{   r   r   r   r]   !  s    zPdfBinary.__init__c             C   s   ddj dd | jD  S )Nz<%s>r6   c             s   s   | ]}d t | V  qdS )z%02XN)r9   )r:   rC   r   r   r   r<   &  s    z$PdfBinary.__str__.<locals>.<genexpr>)rB   r{   )rL   r   r   r   rM   %  s    zPdfBinary.__str__c             C   s   t ddjdd | jD  S )Nz<%s>r6   c             s   s   | ]}d | V  qdS )z%02XNr   )r:   rC   r   r   r   r<   *  s    z&PdfBinary.__bytes__.<locals>.<genexpr>)r   rB   r{   )rL   r   r   r   rN   )  s    zPdfBinary.__bytes__N)rF   rG   rH   r]   r@   rA   rM   rN   r   r   r   r   r      s   
r   c               @   s   e Zd Zdd Zdd ZdS )	PdfStreamc             C   s   || _ || _d S )N)
dictionarybuf)rL   r   r   r   r   r   r]   .  s    zPdfStream.__init__c             C   s   y| j j}W n tk
r"   | jS X |dkrjy| j j}W n tk
rT   | j j}Y nX tj| jt|dS t	dt
| j j d S )Ns   FlateDecode)bufsizez$stream filter %s unknown/unsupported)r   ZFilterr   r   ZDLLengthzlibZ
decompressintNotImplementedErrorrv   )rL   filterZexpected_lengthr   r   r   r?   2  s    zPdfStream.decodeN)rF   rG   rH   r]   r?   r   r   r   r   r   -  s   r   c             C   s  | dkrdS | dkrdS | d kr$dS t | tsLt | tsLt | tsLt | trTt| S t | trlt| jdS t | t	rtt| S t | t
rtt| S ttkrt | tsttkrt | trtt| S t | t rd| jdd	jdd
jdd d S t| S d S )NTs   trueFs   falses   nullzus-ascii   (   \s   \\s   \(   )s   \))rs   rr   r   r   r   rA   r   r@   r   dictlistZunicoder   r   replace)r   r   r   r   r   A  s&    (


$$r   c                @   s  e Zd ZdZdrddZdd Zd	d
 Zdd Zdd Zdd Z	dd Z
dd Zdd Zdd Zdd ZdsddZdd Zdd  Zd!d" Zed#d$ Zd%d& Zdtd'd(Zd)Zd*Zd+Zd,Zed- Zed. Zd/Zd0Zee e Ze j!ed1 e d2 e d3 e d4 e d5 e d6 e j"Z#e j!ed1 e d7 e d3 e d4 e d5 e e j"Z$d8d9 Z%d:d; Z&e j!eZ'e j!ed< e d= Z(e j!ed> Z)e j!ed? e Z*e+d@dA Z,e j!dBZ-e+dudDdEZ.e j!edF e d= Z/e j!edG e d= Z0e j!edH e d= Z1e j!edI e d= Z2e j!edJ e d= Z3e j!edK Z4e j!edL Z5e j!edM e dN Z6e j!edO Z7e j!edP e dP e dQ e d= Z8e j!edP e dP e dR e d= Z9e j!edS e d= Z:e j!dTe dU e dV Z;e j!edW Z<e j!edX e d= Z=e+dwdZd[Z>e j!d\Z?d]d^d_d`dadbdcdddedfdTdTd=d=dgdge@d]d^e@d_d`e@dadbe@dcdde@dedfe@dTdTe@d=d=e@dgdgiZAe+dhdi ZBe j!edj e ZCe j!ed4 e d4 e e ZDe j!dkZEdldm ZFdydndoZGdzdpdqZHdS ){rx   zwBased on http://www.adobe.com/content/dam/acom/en/devnet/acrobat/pdfs/PDF32000_2008.pdf
    Supports PDF up to 1.4
    Nr   rbc             C   s  |r|rt d|| _|| _|| _|| _d| _d| _|d k	rZ|d krZt|| | _}d| _|d k	r| j| | _}d| _| rt	|dr|j
| _i | _|r| j  nTd | _| _t | _d | _t | _d | _i | _g | _g | _d | _d | _i | _t | _d| j_|r| j  d S )Nz4specify buf or f or filename, but not both buf and fFTrt   r   )RuntimeErrorfilenamer   ro   start_offsetshould_close_bufshould_close_fileopenget_buf_from_filer   rt   cached_objectsread_pdf_info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_dictrX   
xref_tabler\   seek_end)rL   r   ro   r   r   moder   r   r   r]   ]  sD    
zPdfParser.__init__c             C   s   | S )Nr   )rL   r   r   r   	__enter__  s    zPdfParser.__enter__c             C   s   | j   dS )NF)close)rL   exc_type	exc_value	tracebackr   r   r   __exit__  s    zPdfParser.__exit__c             C   s   | j   | j  d S )N)	close_bufr   )rL   r   r   r   start_writing  s    zPdfParser.start_writingc             C   s.   y| j j  W n tk
r"   Y nX d | _ d S )N)r   r   r   )rL   r   r   r   r     s
    zPdfParser.close_bufc             C   s2   | j r| j  | jd k	r.| jr.| jj  d | _d S )N)r   r   ro   r   r   )rL   r   r   r   r     s
    
zPdfParser.closec             C   s   | j jdtj d S )Nr   )ro   seekosSEEK_END)rL   r   r   r   r     s    zPdfParser.seek_endc             C   s   | j jd d S )Ns	   %PDF-1.4
)ro   rl   )rL   r   r   r   write_header  s    zPdfParser.write_headerc             C   s   | j jd|f jd d S )Nz%% %s
zutf-8)ro   rl   r   )rL   r   r   r   r   write_comment  s    zPdfParser.write_commentc             C   sl   | j   | j| jj | _| jd| _| j  | j| jtd| jd | j| jtdt	| j
| j
d | jS )Nr   s   Catalog)TypeZPagess   Pages)r   ZCountZKids)del_rootnext_object_idro   rk   r   r   rewrite_pages	write_objrr   r>   r   )rL   r   r   r   write_catalog  s    

zPdfParser.write_catalogc             C   s  g }xt | jD ]\}}| j| }| j|j= |j|td  || jkrJqi }x |j D ]\}}|||j	 < qXW | j
|d< | jd|}x(t | jD ]\}	}
|
|kr|| j|	< qW qW xB|D ]:}x4|r| j| }|j| jkr| j|j= |jdd }qW qW g | _d S )Ns   ParentParent)N)rm   r   r   r   rP   r   rr   r   r   ru   r   
write_pager8   )rL   Zpages_tree_nodes_to_deleteiZpage_refZ	page_infoZstringified_page_infor^   r_   Znew_page_refjZcur_page_refZpages_tree_node_refZpages_tree_noder   r   r   r     s,    






zPdfParser.rewrite_pagesc             C   s   |r| j   || _| jr(| jd | j| _| jj| j}t| j}| j|d}| j	d k	r`| j	|d< | jrp| j|d< || _	| jjdt
t| td|   d S )N)s   Roots   Sizes   Prevs   Infos   trailer
z
startxref
%d
%%%%EOF)r   r   r   r   r   r   rl   ro   r>   r   rA   r   r   )rL   Znew_root_refZ
start_xrefZnum_entriesr   r   r   r   write_xref_and_trailer  s    



z PdfParser.write_xref_and_trailerc             O   sL   t |tr| j| }d|kr(td|d< d|kr:| j|d< | j|f||S )Nr   s   Pager   )rs   r   r   rr   r   r   )rL   refobjsdict_objr   r   r   r     s    


zPdfParser.write_pagec             O   s   | j }|d kr| j|j }n|j |jf| j|j< |jtt|  |j	dd }|d k	rft
||d< |rx|jt| x|D ]}|jt| q~W |d k	r|jd |j| |jd |jd |S )Nstreamr   s   stream
s   
endstream
s   endobj
)ro   r   rk   rQ   r   rP   rl   rA   rW   rn   r>   r   )rL   r   r   r   ro   r   objr   r   r   r     s$    




zPdfParser.write_objc             C   s.   | j d krd S | j| j j= | j| jd j= d S )Ns   Pages)r   r   rP   r   )rL   r   r   r   r     s    
zPdfParser.del_rootc             C   sT   t | dr| j S t | dr$| j S ytj| j dtjdS  tk
rN   dS X d S )N	getbuffergetvaluer   )access    )r   r   r   mmapfilenoZACCESS_READ
ValueError)ro   r   r   r   r     s    

zPdfParser.get_buf_from_filec             C   s   t | j| _| j| j | _| j  | jd | _| jjdd | _	t
| j| j| _| j	d krdt
 | _nt
| j| j	| _td| jkd t| jd dkd td| jkd tt| jd td	 | jd | _| j| j| _| j| j| _| jd d  | _d S )
Ns   Roots   Infos   Typez/Type missing in Roots   Catalogz/Type in Root is not /Catalogs   Pagesz/Pages missing in Rootz+/Pages in Root is not an indirect reference)r>   r   r   r   r   read_trailerr   r   r8   r   r   read_indirectr   r   rJ   rs   rK   r   r   linearize_page_treer   r   )rL   r   r   r   r     s"    

zPdfParser.read_pdf_infoc             C   sX   yt t| jj d d}W n tk
r:   t dd}Y nX |d k	rT|df| j|j< |S )Nrc   r   )rK   maxr   rh   r   rP   )rL   offsetZ	referencer   r   r   r   %  s    zPdfParser.next_object_ids   [][()<>{}/%]s$   [][()<>{}/%\000\011\012\014\015\040]s   [\000\011\012\014\015\040]s#   [\000\011\012\014\015\0400-9a-fA-F]   *   +s   [\000\011\014\015\040]*s   [\r\n]+s   trailers   \<\<(.*\>\>)s	   startxrefs   ([0-9]+)s   %%EOF   $s   \<\<(.*?\>\>)c             C   s   t | jd }|| jk r| j}| jj| j|}t|d |}x$|r`|}| jj| j|j d }q>W |sj|}|jd}t|jd| _	| j
|| _t | _| j| j	d d| jkr| j| jd  d S )Ni @  ztrailer end not found   rc      )xref_section_offsets   Prev)r>   r   r   re_trailer_endsearchrJ   startgroupr   r   interpret_trailerr   rX   r   read_xref_tableread_prev_trailer)rL   Zsearch_start_offsetmZ
last_matchtrailer_datar   r   r   r   =  s$    



zPdfParser.read_trailerc             C   sv   | j |d}| jj| j||d  }t|d |jd}tt|jd|kd | j|}d|krr| j|d  d S )N)r   i @  zprevious trailer not foundrc   r   zGxref section offset in previous trailer doesn't match what was expecteds   Prev)	r   re_trailer_prevr   r   rJ   r   r   r   r   )rL   r   Ztrailer_offsetr   r   r   r   r   r   r   R  s    


zPdfParser.read_prev_trailers   /([!-$&'*-.0-;=?-Z\\^-z|~]+)(?=r   s   \<\<s   \>\>c             C   s   i }d}x|| j j||}|sV| jj||}t|o>|j t|kdt||d    P | j|jd}| j	||j \}}|||< q
W td|kot
|d td td|kot
|d td |S )Nr   z+name not found in trailer, remaining data: rc   s   Sizez&/Size not in trailer or not an integers   Rootz1/Root not in trailer or not an indirect reference)re_namematchre_dict_endrJ   endr>   rv   ry   r   	get_valuers   r   rK   )rz   r   Ztrailerr   r   r^   r_   r   r   r   r   a  s    *zPdfParser.interpret_trailers   ([^#]*)(#([0-9a-fA-F]{2}))?Fc             C   sr   d}xR| j j|D ]B}|jdrF||jdtj|jdjd 7 }q||jd7 }qW |rf|jdS t|S d S )Nr      rc   zus-asciizutf-8)re_hashes_in_namefinditerr   r   fromhexr?   rA   )rz   rawZas_textrt   r   r   r   r   ry   t  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   *)\>s   \(s   ([-+]?[0-9]+)s   R(?=s   obj(?=s	   endobj(?=r   s	   %[^\r\n]*s   )*s   stream\r?\ns   endstream(?=rc   c             C   s  |dkrdS | j j||}|r&|j }| jj||}|rtt|jddkd tt|jddkd t|d kp|tt|jdt|jdkd | j||j |d d\}}|d kr|d fS | j	j||}t|d ||j fS t| d	 | j
j||}|r^tt|jddkd
 tt|jddkd tt|jdt|jd|j fS | jj||}|r|j }i }| jj||}xv|s| j|||d d\}}|d kr|d fS | j|||d d\}	}|	||< |d kr|d fS | jj||}qW |j }| jj||}|ryt|d }
W n0 tttfk
r`   td|jdd  Y nX ||j |j |
  }| jj||j |
 }t|d |j }tt||}nt|}||fS | jj||}|rL|j }g }| jj||}xN|s>| j|||d d\}	}|j|	 |d kr,|d fS | jj||}qW ||j fS | jj||}|rld |j fS | jj||}|rd|j fS | jj||}|rd|j fS | jj||}|rt| j|jd|j fS | jj||}|rt|jd|j fS | j j||}|r0t!|jd|j fS | j"j||}|rt#dd |jdD }t$|d dkr||jt%d t#j&|j'd|j fS | j(j||}|r| j)||j S tdt*|||d   d S )Nr   rc   z<indirect object definition: object ID must be greater than 0r   z;indirect object definition: generation must be non-negativez2indirect object definition different than expected)max_nestingz(indirect object definition end not foundz$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   g | ]}|d kr|qS )s   0123456789abcdefABCDEFr   )r:   rC   r   r   r   
<listcomp>  s    z'PdfParser.get_value.<locals>.<listcomp>   0zus-asciizunrecognized object:     )NN)+
re_commentr   r   re_indirect_def_startrJ   r   r   rK   r   re_indirect_def_endre_indirect_referencere_dict_startr   re_stream_start	TypeErrorra   r   rE   r8   re_stream_endr   r   re_array_startre_array_endr   re_nullre_truere_falser   rr   ry   re_intre_realfloatre_string_hexr   r>   r9   r  r?   re_string_litget_literal_stringrv   )rz   r{   r   expect_indirectr  r   objectr   r^   r_   Z
stream_lenZstream_dataZ
hex_stringr   r   r   r     s    *
&




zPdfParser.get_valuesF   (\\[nrtbf()\\])|(\\[0-9]{1,3})|(\\(\r\n|\r|\n))|(\r\n|\r|\n)|(\()|(\))   nr      r      t   	   b      f   r   c             C   s  d}t  }x| jj||D ]}|j|||j   |jdrZ|j| j|jdd   n|jdr|jt|jddd  d nt|jdrnh|jdr|jd nR|jdr|jd	 |d7 }n4|jd
r|dkrt	||j
 fS |jd |d8 }|j
 }qW tdd S )Nr   rc   r      r      r      r      r   zunfinished literal string)r   re_lit_str_tokenr   r   r   r   escaped_charsr   r   rA   r   rE   )rz   r{   r   Znesting_depthr   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  d}| j j| j|| j }t|d |j }x| jj| j|}|sNt|d P d}|j }t|jd}t|jd}xt	||| D ]}| j
j| j|}t|d |j }|jdd	k}t|jd}	|st|jd|	f}
t|| jkp| j| |
kd
 |
| j|< qW q.W |S )NFzxref section start not foundzxref subsection start not foundTrc   r   zxref entry not foundr   r$  z)xref entry duplicated (and not identical))re_xref_section_startr   r   r   rJ   r   re_xref_subsection_startr   r   r   re_xref_entryr   )rL   r   Zsubsection_foundr   r   Zfirst_objectZnum_objectsr   Zis_freerQ   Z	new_entryr   r   r   r   "  s0    


zPdfParser.read_xref_tablec             C   sh   | j |d  \}}t||d kd|d |d ||f  | j| j|| j t| |dd }|| j|< |S )Nr   rc   zgexpected to find generation %s for object ID %s in xref table, instead found generation %s at offset %s)r  r  )r   rJ   r   r   r   rK   r   )rL   r   r  r   rQ   r_   r   r   r   r   <  s    "
zPdfParser.read_indirectc             C   sn   |d kr| j }t|d dkd g }xD|d D ]8}| j|}|d dkrT|j| q.|j| j|d q.W |S )Ns   Types   Pagesz%/Type of page tree node is not /Pagess   Kidss   Page)node)r   rJ   r   r   r   r   )rL   r/  r   ZkidZ
kid_objectr   r   r   r   D  s    
zPdfParser.linearize_page_tree)NNNr   r   )N)N)F)Nr0  r0  )r0  )N)IrF   rG   rH   rI   r]   r   r   r   r   r   r   r   r   r   r   r   r   r   r   staticmethodr   r   r   Z	delimiterZdelimiter_or_wsZ
whitespaceZwhitespace_or_hexZwhitespace_optionalZwhitespace_mandatoryZwhitespace_optional_no_nlZnewline_onlynewlinerecompileDOTALLr   r   r   r   Zre_whitespace_optionalr   r  r   r   r   r   ry   r  r  r  r  r  r  r  r  r  r  r	  r
  r  r  r  r   r*  r9   r+  r  r,  r-  r.  r   r   r   r   r   r   r   rx   X  s   
&
	

62


&&_


rx   )r	   collectionsr   r   r3  r   r   ImportErrorr@   rA   r   r   r7   rD   r   rE   rJ   
namedtuplerK   rW   rX   rr   r   r   r   r   r   r   rx   r   r   r   r   <module>   s   
	L/-