�
��Xjc@s
dZd
dl
Z
d
dlZejd�
Zejd�
Zejd�
Zejd�
Zejd�
Zejd�
Z	ejd	�
Z
ejd
�
Zejd�
Zejd�
Z
ejd
ej�Zejd�
Zejd�
Zdef
d��YZde
jf
d��YZdS(sA parser for HTML and XHTML.i����Ns[&<]s
&[a-zA-Z#]s%&([a-zA-Z][-.a-zA-Z0-9]*)[^a-zA-Z0-9]s)&#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]s	<[a-zA-Z]t
>s--\s*>s(([a-zA-Z][-.a-zA-Z0-9:_]*)(?:\s|/(?!>))*s[a-zA-Z][^	

 />]*s]((?<=[\'"\s/])[^\s/>][^\s/=>]*)(\s*=+\s*(\'[^\']*\'|"[^"]*"|(?![\'"])[^>\s]*))?(?:\s|/(?!>))*s
  <[a-zA-Z][-.a-zA-Z0-9:_]*          # 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|/(?!>))*
     )*
   )?
  \s*                                # trailing whitespace
s#</\s*([a-zA-Z][-.a-zA-Z0-9:_]*)\s*>tHTMLParseErrorcBs#eZ
dZdd
�
Zd�ZRS(s&Exception raised for all parse errors.cCs3|
st�
|
|_
|d
|_|d|_dS(Nii
(tAssertionErrortmsgtlinenotoffset(tselfRtposition((s"/usr/lib64/python2.7/HTMLParser.pyt__init__9s

	


c
CsW|j}
|j
dk	r,|
d
|j
}
n|jdk	rS|
d|jd}
n|
S(Ns, at line %ds, column %di
(RRtNoneR(Rtresult((s"/usr/lib64/python2.7/HTMLParser.pyt__str__?s
	




N(NN(t__name__t
__module__t__doc__R	RR(((s"/usr/lib64/python2.7/HTMLParser.pyR
6s
t
HTMLParsercBs

eZ
dZdZd�Zd�Zd�Zd�Zd�ZdZ
d�Zd	�Zd
�Z
d�Zd�Zd
d�
Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�ZdZd�Z RS( s�Find tags and other markup and call handler functions.

    Usage:
        p = HTMLParser()
        p.feed(data)
        ...
        p.close()

    Start tags are handled by calling self.handle_starttag() or
    self.handle_startendtag(); end tags by self.handle_endtag().  The
    data between tags is passed from the parser to the derived class
    by calling self.handle_data() with the data as argument (the data
    may be split up in arbitrary chunks).  Entity references are
    passed by calling self.handle_entityref() with the entity
    reference as the argument.  Numeric character references are
    passed to self.handle_charref() with the string containing the
    reference as the argument.
    tscripttstylec


Cs|j�
d
S(s#Initialize and reset this instance.N(
treset(
R((s"/usr/lib64/python2.7/HTMLParser.pyR_sc

CsSd
|_d|_
t|_d|_g|_d|_d|_t	j
j|�

dS(s1Reset this instance.  Loses all unprocessed data.ts???ii
N(trawdatatlasttagtinteresting_normaltinterestingR	t
cdata_elemt_pendingt_pending_lent_parse_thresholdt
markupbaset
ParserBaseR(
R((s"/usr/lib64/python2.7/HTMLParser.pyRcs	
	
	
	
	
	
	
cCs�|jt
|
�
7_|j|jkr:|jj|
�

n�|jsU|j|
7_n2|jj|
�

|jd
j|j�
7_|j2d|_t
|j�
}|jd�

t
|j�
|kr�d|_nt
|j�
|_dS(s�Feed data to the parser.

        Call this as often as you want, with as little or as much text
        as you want (may include '\n').
        Rii
N(RtlenRRtappendRtjointgoahead(Rtdatat
n((s"/usr/lib64/python2.7/HTMLParser.pytfeedns


	



	



c

CsH|jr7|j
d
j|j�
7_
|j2d|_n|jd�

dS(sHandle any buffered data.Rii
N(RRR RR!(
R((s"/usr/lib64/python2.7/HTMLParser.pytclose�s
	



cCst|
|j
���
dS(
N(R
tgetpos(Rtmessage((s"/usr/lib64/python2.7/HTMLParser.pyterror�s
c


Cs|jS(
s)Return full source of start tag: '<...>'.(
t_HTMLParser__starttag_text(
R((s"/usr/lib64/python2.7/HTMLParser.pytget_starttag_text�scCs2|
j�|_
tjd
|j
tj�|_dS(Ns</\s*%s\s*>(tlowerRtretcompilet
IR(Rtelem((s"/usr/lib64/python2.7/HTMLParser.pytset_cdata_mode�s

c

Cst|_
d|_dS(
N(RRR	R(
R((s"/usr/lib64/python2.7/HTMLParser.pytclear_cdata_mode�s
	
c
Cst|j}d
}t
|�
}x||kr|jj||�}|rT|j�}n|jraPn|}||kr�|j|||!�

n|j||�}||kr�Pn|j}|d|�r7t	j
||�r�|j|�
}n�|d|�r	
|j|�
}n�|d|�r*
|j
|�
}nm|d|�rK
|j|�
}nL|d|�rl
|j|�
}n+|d|kr�
|jd�

|d}n
P|d
kr"|
s�
Pn|jd|d�}|d
kr
|jd|d�}|d
kr|d}qn
|d7}|j|||!�

n|j||�}q|d	|�r�tj
||�}|r�|j�d
d!}	|j|	�

|j�}|d|d�s�|d}n|j||�}qqd||kr�|j|d
d
!�

|j|d
�}nPq|d
|�rtj
||�}|r�|jd�
}	|j|	�

|j�}|d|d�sn|d}n|j||�}qntj
||�}|r�|
r�|j�||kr�|jd�

nPq|d|kr|jd
�

|j||d�}qPqd
std�
�
qW|
rc||krc|jrc|j|||!�

|j||�}n|||_dS(Nit
<s</s<!--s<?s<!i
Rs&#ii����t
;t
&s#EOF in middle of entity or char refsinteresting.search() lied(RRRtsearchtstartRthandle_datat	updatepost
startswithtstarttagopentmatchtparse_starttagtparse_endtagt
parse_commenttparse_pitparse_html_declarationtfindtcharreftgroupthandle_charreftendt	entityrefthandle_entityreft
incompleteR(R(
RRERt
iR#R;t
jR9t
ktname((s"/usr/lib64/python2.7/HTMLParser.pyR!�s�
	





	





	































































cCs�|j}||
|
d
!dkr0|j
d�

n||
|
d!dkrT|j|
�
S||
|
d!dkrx|j|
�
S||
|
d!j�d	kr�|jd
|
d�}|dkr�dS|j||
d
|!�

|dS|j|
�
SdS(
Nis<!s+unexpected call to parse_html_declaration()is<!--is<![i	s	<!doctypeRi����i
(RR(R>tparse_marked_sectionR+RAthandle_decltparse_bogus_comment(RRIRtgtpos((s"/usr/lib64/python2.7/HTMLParser.pyR@
s
	











i
cCs|j}||
|
d
!dkr0|j
d�

n|jd|
d
�}|dkrVdS|rw|j||
d
|!�

n|dS(	Nis<!s</s"unexpected call to parse_comment()Ri����i
(s<!s</(RR(RAthandle_comment(RRItreportRtpos((s"/usr/lib64/python2.7/HTMLParser.pyRO
s
	







cCs�|j}||
|
d
!dks,t
d�
�
tj||
d
�}|sLdS|j�}|j||
d
|!�

|j�}|S(Nis<?sunexpected call to parse_pi()i����(RRtpicloseR5R6t	handle_piRE(RRIRR;RJ((s"/usr/lib64/python2.7/HTMLParser.pyR?'
s
	
#






cCs�d|_
|j|
�
}|d
kr(|S|j}||
|!|_
g}tj||
d�}|sotd�
�
|j�}|jd�
j	�|_
}x�||kr�
tj||�}|s�Pn|jddd�\}	}
}|
s�d}nX|d dko
|dkn
s7
|d dko2
|dkn
rG
|dd!}n|r_
|j|�
}n|j
|	j	�|f�

|j�}q�W|||!j�}|dkr+|j�\}
}d|j
kr|
|j
jd�
}
t|j
�
|j
jd�
}n|t|j
�
}|j||
|!�

|S|jd
�
rM|j||�
n/|j||�
||jkr||j|�

n|S(
Nii
s#unexpected call to parse_starttag()iis
'i����t
"Rs/>s

(Rs/>(R	R)tcheck_for_whole_start_tagRttagfindR;RRERCR+RtattrfindtunescapeRtstripR&tcountRtrfindR7tendswiththandle_startendtagthandle_starttagtCDATA_CONTENT_ELEMENTSR0(RRItendposRtattrsR;RKttagt
mtattrnametrestt	attrvalueRERR((s"/usr/lib64/python2.7/HTMLParser.pyR<3
sR
	



	










	
$
$
















cCs�|j}t
j||
�}|r�|j�}|||d
!}|dkrR|d
S|dkr�|jd|�rx|dS|jd|�r�dS|j|
|d
�
|jd�

n|dkr�dS|d	kr�dS||
kr�|S|
d
Sntd
�
�
dS(Ni
Rt
/s/>ii����smalformed empty start tagRs6abcdefghijklmnopqrstuvwxyz=/ABCDEFGHIJKLMNOPQRSTUVWXYZswe should not get here!(RtlocatestarttagendR;RER9R8R(R(RRIRReRJtnext((s"/usr/lib64/python2.7/HTMLParser.pyRWg
s,
	















cCs�
|j}||
|
d
!dks,t
d�
�
tj||
d�}|sLdS|j�}tj||
�}|s!
|jdk	r�|j	||
|!�

|St
j||
d
�}|s�||
|
d!dkr�|
dS|j|
�
Sn|j�j
�}|jd|j��}|j|�

|dS|jd�
j
�}|jdk	ro
||jkro
|j	||
|!�

|Sn|j|�

|j�
|S(	Nis</sunexpected call to parse_endtagi
i����is</>R(RRt	endendtagR5REt
endtagfindR;RR	R7ttagfind_tolerantRORCR+RAt
handle_endtagR1(RRIRR;RPt	namematchttagnameR/((s"/usr/lib64/python2.7/HTMLParser.pyR=�
s8
	
#






















cCs!|j|
|�
|j
|
�

dS(
N(R`Ro(RRdRc((s"/usr/lib64/python2.7/HTMLParser.pyR_�
s

c
CsdS(
N((RRdRc((s"/usr/lib64/python2.7/HTMLParser.pyR`�
s
c
CsdS(
N((RRd((s"/usr/lib64/python2.7/HTMLParser.pyRo�
s
c
CsdS(
N((RRL((s"/usr/lib64/python2.7/HTMLParser.pyRD�
s
c
CsdS(
N((RRL((s"/usr/lib64/python2.7/HTMLParser.pyRG�
s
c
CsdS(
N((RR"((s"/usr/lib64/python2.7/HTMLParser.pyR7�
s
c
CsdS(
N((RR"((s"/usr/lib64/python2.7/HTMLParser.pyRQ�
s
c
CsdS(
N((Rtdecl((s"/usr/lib64/python2.7/HTMLParser.pyRN�
s
c
CsdS(
N((RR"((s"/usr/lib64/python2.7/HTMLParser.pyRU�
s
c
CsdS(
N((RR"((s"/usr/lib64/python2.7/HTMLParser.pytunknown_decl�
s
cs2d
|
kr|
S�f
d�}tj
d||
�S(NR4c
s
|j�d
}yZ|d
dkri|d}|d
d
krSt
|dd�}
nt
|�
}
t|
�
SWntk
r�


d|dSXd	dl}tjdkr�i
d
d6}t_x0|jj	�D]\}}t|�
||<q�Wny�j|SWnt
k
r



d|dSXdS(Nit
#i
t
xt
Xis&#R3i����u
'taposR4(RuRv(tgroupstinttunichrt
ValueErrorthtmlentitydefsRt
entitydefsR	tname2codepointt	iteritemstKeyError(t
st
cR|R}RKt
v(
R(s"/usr/lib64/python2.7/HTMLParser.pytreplaceEntities�
s&




















s#&(#?[xX]?(?:[0-9a-fA-F]+|\w{1,8}));(R,tsub(RR�R�((
Rs"/usr/lib64/python2.7/HTMLParser.pyRZ�
s


(RRN(!RR
RRaRRR$R%R(R	R)R*R0R1R!R@ROR?R<RWR=R_R`RoRDRGR7RQRNRURsR}RZ(((s"/usr/lib64/python2.7/HTMLParser.pyRHs<									^			4	 	(										
(RRR,R-RRHRFRBR:RTtcommentcloseRXRnRYtVERBOSERjRlRmt	ExceptionR
RR(((s"/usr/lib64/python2.7/HTMLParser.pyt<module>
s&