
    5bhN                        	 d dl Z d dlZd dlmZ dgZ e j
                  d      Z e j
                  d      Z e j
                  d      Z e j
                  d      Z	 e j
                  d      Z
 e j
                  d	      Z e j
                  d
      Z e j
                  d      Z e j
                  d      Z e j
                  d      Z e j
                  de j                         Z e j
                  de j                         Z e j
                  de j                         Z e j
                  d
      Z e j
                  d      Z G d dej,                        Zy)    N)unescape
HTMLParserz[&<]z
&[a-zA-Z#]z%&([a-zA-Z][-.a-zA-Z0-9]*)[^a-zA-Z0-9]z)&#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]z	<[a-zA-Z]z
</[a-zA-Z]>z--!?>z-?>z0([a-zA-Z][^\t\n\r\f />]*)(?:[\t\n\r\f ]|/(?!>))*a{  
  (
    (?<=['"\t\n\r\f /])[^\t\n\r\f />][^\t\n\r\f /=>]*  # attribute name
   )
  ([\t\n\r\f ]*=[\t\n\r\f ]*        # value indicator
    ('[^']*'                        # LITA-enclosed value
    |"[^"]*"                        # LIT-enclosed value
    |(?!['"])[^>\t\n\r\f ]*         # bare value
    )
   )?
  (?:[\t\n\r\f ]|/(?!>))*           # possibly followed by a space
a  
  [a-zA-Z][^\t\n\r\f />]*           # tag name
  [\t\n\r\f /]*                     # optional whitespace before attribute name
  (?:(?<=['"\t\n\r\f /])[^\t\n\r\f />][^\t\n\r\f /=>]*  # attribute name
    (?:[\t\n\r\f ]*=[\t\n\r\f ]*    # value indicator
      (?:'[^']*'                    # LITA-enclosed value
        |"[^"]*"                    # LIT-enclosed value
        |(?!['"])[^>\t\n\r\f ]*     # bare value
       )
     )?
    [\t\n\r\f /]*                   # possibly followed by a space
   )*
   >?
aF  
  <[a-zA-Z][^\t\n\r\f />\x00]*       # tag name
  (?:[\s/]*                          # optional whitespace before attribute name
    (?:(?<=['"\s/])[^\s/>][^\s/=>]*  # attribute name
      (?:\s*=+\s*                    # value indicator
        (?:'[^']*'                   # LITA-enclosed value
          |"[^"]*"                   # LIT-enclosed value
          |(?!['"])[^>\s]*           # bare value
         )
        \s*                          # possibly followed by a space
       )?(?:\s|/(?!>))*
     )*
   )?
  \s*                                # trailing whitespace
z#</\s*([a-zA-Z][-.a-zA-Z0-9:_]*)\s*>c                        e Zd Z	 dZdZdd fd
Z fdZd Zd Zd	Z	d
 Z
dddZd Zd"dZd Zd Zd"dZd#dZd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd  Zd! Z xZ S )$r   )scriptstyle)textareatitleT)convert_charrefsc                R    	 t         |           || _        | j                          y N)super__init__r   reset)selfr   	__class__s     5/opt/python-3.12.12/usr/lib/python3.12/html/parser.pyr   zHTMLParser.__init__s   s&    	
 	 0

    c                     	 d| _         d| _        t        | _        d | _        d| _        d| _        t        | !          y )N z???T)	rawdatalasttaginteresting_normalinteresting
cdata_elem_support_cdata
_escapabler   r   )r   r   s    r   r   zHTMLParser.reset}   s;    ?-"r   c                 P    	 | j                   |z   | _         | j                  d       y )Nr   )r   goaheadr   datas     r   feedzHTMLParser.feed   s$    	
 ||d*Qr   c                 (    	 | j                  d       y )N   )r   r   s    r   closezHTMLParser.close   s    'Qr   Nc                     	 | j                   S r   )_HTMLParser__starttag_textr%   s    r   get_starttag_textzHTMLParser.get_starttag_text   s    7###r   F	escapablec                v   |j                         | _        || _        |rT| j                  sHt	        j
                  d| j                  z  t        j                  t        j                  z        | _        y t	        j
                  d| j                  z  t        j                  t        j                  z        | _        y )Nz&|</%s(?=[\t\n\r\f />])z</%s(?=[\t\n\r\f />]))	lowerr   r   r   recompile
IGNORECASEASCIIr   )r   elemr+   s      r   set_cdata_modezHTMLParser.set_cdata_mode   s    **,#T22!zz*Dt*V*,--*@ BD  "zz*BT__*T*,--*@ BDr   c                 6    t         | _        d | _        d| _        y )NT)r   r   r   r   r%   s    r   clear_cdata_modezHTMLParser.clear_cdata_mode   s    -r   c                     	 || _         y r   )r   )r   flags     r   _set_support_cdatazHTMLParser._set_support_cdata   s    		 #r   c                    | j                   }d}t        |      }||k  r2| j                  rq| j                  se|j	                  d|      }|dk  r|j                  dt        ||dz
              }|dk\  r't        j                  d      j                  ||      sn|}n?| j                  j                  ||      }|r|j                         }n| j                  rnx|}||k  rJ| j                  r*| j                  r| j                  t        |||              n| j                  |||        | j                  ||      }||k(  rn|j                   } |d|      r=t"        j%                  ||      r| j'                  |      }	n |d|      r| j)                  |      }	nt |d|      r| j+                  |      }	nY |d|      r| j-                  |      }	n> |d	|      r| j/                  |      }	n#|d
z   |k  s|r| j                  d       |d
z   }	nnA|	dk  rl|sn7t"        j%                  ||      rnN |d|      rK|dz   |k(  r| j                  d       n*t0        j%                  ||      rn| j3                  ||dz   d         n |d|      rF|}dD ]'  }
|j5                  |
|dz         s|t        |
      z  } n | j3                  ||dz   |        n |d|      r$| j6                  r| j9                  ||dz   d         n~|||dz    j;                         dk(  r| j=                  ||dz   d         nM |d	|      r| j3                  ||dz   d         n, |d|      r| j?                  ||dz   d         ntA        d      |}	| j                  ||	      }n |d|      rtB        j%                  ||      }|rY|jE                         dd }| jG                  |       |jI                         }	 |d|	d
z
        s|	d
z
  }	| j                  ||	      }d||d  v r,| j                  |||dz           | j                  ||dz         }n |d|      rtJ        j%                  ||      }|rW|jE                  d
      }| jM                  |       |jI                         }	 |d|	d
z
        s|	d
z
  }	| j                  ||	      }tN        j%                  ||      }|rE|rB|jE                         ||d  k(  r,|jI                         }	|	|k  r|}	| j                  ||d
z         }n7|d
z   |k  r'| j                  d       | j                  ||d
z         }nn||k  r2|ra||k  r\| j                  r*| j                  r| j                  t        |||              n| j                  |||        | j                  ||      }||d  | _         y )Nr   <&"   z[\t\n\r\f ;]z</<!--z<?z<!r$      )z--!z---   	<![CDATA[   	   	<!doctypezwe should not get here!z&#;)(r   lenr   r   findrfindmaxr.   r/   searchr   startr   handle_datar   	updatepos
startswithstarttagopenmatchparse_starttagparse_endtagparse_commentparse_piparse_html_declaration
endtagopenhandle_commentendswithr   unknown_declr-   handle_decl	handle_piAssertionErrorcharrefgrouphandle_charrefend	entityrefhandle_entityref
incomplete)r   ra   r   injampposrQ   rO   ksuffixnames               r   r   zHTMLParser.goahead   s{   ,,L!e$$T__LLa(q5 %]]3Aqt=F!JJ7>>wOA((//;AA1u((T__$$Xgal%;<$$WQq\2q!$AAvu ++J#q!%%gq1++A.Aa())!,A***1-Aa(a(Aa(33A6A!eq[C$$S)AAq5#))'15#D!,q5A: ,,T2'--gq9  !//!>#FA.&8F&//!< !S[ 0 % '9 ++GAaCN;#K38K8K))'!A#$-8 1Q3--/;>((17#D!,++GAaCDM:#D!,wqst}5,-FGGANN1a(D!$gq1 ;;=2.D''-		A%c1Q3/Eq!,Agabk)((1Q38 NN1ac2C#!3 ;;q>D))$/		A%c1Q3/Eq!,A"((!4u{{};!IIK6 !A NN1a!e4!eq[ $$S)q!a%0As !ez 1q5$$  '!A,!78  1.q!$Aqr{r   c                    | j                   }|||dz    dk(  r| j                  |      S |||dz    dk(  rC| j                  r7|j                  d|dz         }|dk  ry| j	                  ||dz   |        |dz   S |||dz    j                         d	k(  r7|j                  d
|dz         }|dk(  ry| j                  ||dz   |        |dz   S |||dz    dk(  ra|j                  d
|dz         }|dk  ry||dz
     dk(  r| j	                  ||dz   |dz
          |dz   S | j                  ||dz   |        |dz   S | j                  |      S )Nr@   r=   rC   rA   z]]>r   rE   rB   rD   r   r>   r$   z<![])	r   rT   r   rH   rZ   r-   r[   rX   parse_bogus_comment)r   re   r   rg   gtposs        r   rV   z!HTMLParser.parse_html_declarationE  s   ,, 1QqS>V#%%a((Qqs^{*t/B/BUAaC(A1ugac1o.q5LQqs^!!#{2LLac*E{WQqS/07NQqs^u$S!A#&A1uqs|s"!!'!A#qs"34 q5L ##GAaCO4q5L++A..r   c                     | j                   }t        j                  ||dz         }|st        j	                  ||dz         }|sy|r'|j                         }| j                  ||dz   |        |j                         S )Nr@   rE   )r   commentcloserK   commentabruptcloserQ   rL   rX   ra   )r   re   reportr   rQ   rg   s         r   rT   zHTMLParser.parse_commentg  sr    ,,##GQqS1&,,Wac:EA!Q0yy{r   c                     | j                   }|j                  d|dz         }|dk(  ry|r| j                  ||dz   |        |dz   S )Nr   r>   rE   r$   )r   rH   rX   )r   re   rs   r   poss        r   rn   zHTMLParser.parse_bogus_commentv  sP    ,, ll3!$"9!C 01Qwr   c                     | j                   }t        j                  ||dz         }|sy|j                         }| j	                  ||dz   |        |j                         }|S )Nr>   rE   )r   picloserK   rL   r\   ra   r   re   r   rQ   rg   s        r   rU   zHTMLParser.parse_pi  sV    ,,w!,KKMwqsA'IIKr   c                    d | _         | j                  |      }|dk  r|S | j                  }||| | _         g }t        j	                  ||dz         }|j                         }|j                  d      j                         x| _        }||k  rt        j	                  ||      }|sn|j                  ddd      \  }	}
}|
sd }n,|d d dcxk(  r|dd  k(  sn |d d dcxk(  r|dd  k(  rn n|dd }|rt        |      }|j                  |	j                         |f       |j                         }||k  r||| j                         }|dvr| j                  |||        |S |j                  d	      r| j                  ||       |S | j!                  ||       || j"                  v r| j%                  |       |S || j&                  v r| j%                  |d
       |S )Nr   r$   r>   rB   'rE   ")r   />r|   Tr*   )r(   check_for_whole_start_tagr   tagfind_tolerantrQ   ra   r_   r-   r   attrfind_tolerantr   appendstriprM   rY   handle_startendtaghandle_starttagCDATA_CONTENT_ELEMENTSr3   RCDATA_CONTENT_ELEMENTS)r   re   endposr   attrsrQ   ri   tagmattrnamerest	attrvaluera   s                r   rR   zHTMLParser.parse_starttag  s     $//2A:M,,&q0  &&w!4IIK"[[^1133s&j!''3A()1a(8%HdI 	2A$8)BC.82A#7237%aO	$Y/	LL(..*I67A &j a%%'k!WQv./M<<##C/    e,d111##C(  444##C4#8r   c                     | j                   }t        j                  ||dz         }|j                         }||dz
     dk7  ry|S )Nr$   r   rE   )r   locatetagendrQ   ra   rx   s        r   r}   z$HTMLParser.check_for_whole_start_tag  sD    ,,""7AaC0IIK1Q3<3r   c                    | j                   }|j                  d|dz         dk  ryt        j                  ||      s$||dz   |dz    dk(  r|dz   S | j	                  |      S t
        j                  ||dz         }|j                         }||dz
     dk7  ryt        j                  ||dz         }|j                  d      j                         }| j                  |       | j                          |S )Nr   r>   r   rE   rB   r$   )r   rH   rW   rQ   rn   r   ra   r~   r_   r-   handle_endtagr5   )r   re   r   rQ   rg   r   s         r   rS   zHTMLParser.parse_endtag  s     ,,<<QqS!A%+qs1Q33&s
//22""7AaC0IIK1Q3<3 !&&w!4kk!n""$3r   c                 J    | j                  ||       | j                  |       y r   )r   r   r   r   r   s      r   r   zHTMLParser.handle_startendtag  s     S%(3r   c                      y r    r   s      r   r   zHTMLParser.handle_starttag      r   c                      y r   r   )r   r   s     r   r   zHTMLParser.handle_endtag  r   r   c                      y r   r   r   rk   s     r   r`   zHTMLParser.handle_charref  r   r   c                      y r   r   r   s     r   rc   zHTMLParser.handle_entityref  r   r   c                      y r   r   r    s     r   rM   zHTMLParser.handle_data  r   r   c                      y r   r   r    s     r   rX   zHTMLParser.handle_comment  r   r   c                      y r   r   )r   decls     r   r[   zHTMLParser.handle_decl  r   r   c                      y r   r   r    s     r   r\   zHTMLParser.handle_pi  r   r   c                      y r   r   r    s     r   rZ   zHTMLParser.unknown_decl
  r   r   )T)r$   )!__name__
__module____qualname__r   r   r   r   r"   r&   r(   r)   r3   r5   r8   r   rV   rT   rn   rU   rR   r}   rS   r   r   r   r`   rc   rM   rX   r[   r\   rZ   __classcell__)r   s   @r   r   r   Z   s    * 13+/  O$ 16 B
# G#X/D		,`< 
r   )r.   _markupbasehtmlr   __all__r/   r   rd   rb   r^   rP   rW   rw   rq   rr   r~   VERBOSEr   r   locatestarttagend_tolerant	endendtag
endtagfind
ParserBaser   r   r   r   <module>r      s[   " 
   .  RZZ' RZZ%
BJJ>?	
"**@
Arzz+&RZZ%

"**S/rzz(#RZZ'  2::QR BJJ   ZZ  rzz  ZZ  (RZZ ) ZZ  BJJsO	RZZ>?
q'' qr   