Your IP : 216.73.216.160
�
��fMc@sJdZddlZddlZddlZejd�Zejd�Zejd�Zejd�Zejd�Z ejd�Z
ejd �Zejd
�Zejd�Z
ejd�Zejd
�Zejdej�Zejdej�Zejd�Zejd�ZGdd�de�ZGdd�dej�ZdS(uA parser for HTML and XHTML.iNu[&<]u
&[a-zA-Z#]u%&([a-zA-Z][-.a-zA-Z0-9]*)[^a-zA-Z0-9]u)&#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]u <[a-zA-Z]u>u--\s*>u(([a-zA-Z][-.a-zA-Z0-9:_]*)(?:\s|/(?!>))*u$([a-zA-Z][^
/>]*)(?:\s|/(?!>))*uJ\s*([a-zA-Z_][-.:a-zA-Z_0-9]*)(\s*=\s*(\'[^\']*\'|"[^"]*"|[^\s"\'=<>`]*))?u]((?<=[\'"\s/])[^\s/>][^\s/=>]*)(\s*=+\s*(\'[^\']*\'|"[^"]*"|(?![\'"])[^>\s]*))?(?:\s|/(?!>))*u�
<[a-zA-Z][-.a-zA-Z0-9:_]* # tag name
(?:\s+ # whitespace before attribute name
(?:[a-zA-Z_][-.:a-zA-Z0-9_]* # attribute name
(?:\s*=\s* # value indicator
(?:'[^']*' # LITA-enclosed value
|\"[^\"]*\" # LIT-enclosed value
|[^'\">\s]+ # bare value
)
)?
)
)*
\s* # trailing whitespace
uF
<[a-zA-Z][^\t\n\r\f />\x00]* # tag name
(?:[\s/]* # optional whitespace before attribute name
(?:(?<=['"\s/])[^\s/>][^\s/=>]* # attribute name
(?:\s*=+\s* # value indicator
(?:'[^']*' # LITA-enclosed value
|"[^"]*" # LIT-enclosed value
|(?!['"])[^>\s]* # bare value
)
(?:\s*,)* # possibly followed by a comma
)?(?:\s|/(?!>))*
)*
)?
\s* # trailing whitespace
u#</\s*([a-zA-Z][-.a-zA-Z0-9:_]*)\s*>cBs5|EeZdZdZddd�Zdd�ZdS(uHTMLParseErroru&Exception raised for all parse errors.cCs'||_|d|_|d|_dS(Nii(umsgulinenouoffset(uselfumsguposition((u0/opt/alt/python33/lib64/python3.3/html/parser.pyu__init__Ps
uHTMLParseError.__init__cCsW|j}|jdk r,|d|j}n|jdk rS|d|jd}n|S(Nu, at line %du, column %di(umsgulinenouNoneuoffset(uselfuresult((u0/opt/alt/python33/lib64/python3.3/html/parser.pyu__str__Vs uHTMLParseError.__str__N(NN(u__name__u
__module__u__qualname__u__doc__uNoneu__init__u__str__(u
__locals__((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuHTMLParseErrorMsuHTMLParseErrorcBsd|EeZdZdZd:Zd;dd�Zdd�Zdd �Zd
d�Z dd
�Z
d9Zdd�Z
dd�Zdd�Zdd�Zdd�Zddd�Zdd�Zdd�Zdd �Zd!d"�Zd#d$�Zd%d&�Zd'd(�Zd)d*�Zd+d,�Zd-d.�Zd/d0�Zd1d2�Zd3d4�Zd5d6�Z d7d8�Z!d9S(<u
HTMLParseru�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.
uscriptustylecCs6|rtjdtdd�n||_|j�dS(u�Initialize and reset this instance.
If strict is set to False (the default) the parser will parse invalid
markup, otherwise it will raise an error. Note that the strict mode
is deprecated.
uThe strict mode is deprecated.u
stackleveliN(uwarningsuwarnuDeprecationWarningustrictureset(uselfustrict((u0/opt/alt/python33/lib64/python3.3/html/parser.pyu__init__us
uHTMLParser.__init__cCs8d|_d|_t|_d|_tjj|�dS(u1Reset this instance. Loses all unprocessed data.uu???N( urawdataulasttaguinteresting_normaluinterestinguNoneu
cdata_elemu_markupbaseu
ParserBaseureset(uself((u0/opt/alt/python33/lib64/python3.3/html/parser.pyureset�s
uHTMLParser.resetcCs!|j||_|jd�dS(u�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').
iN(urawdataugoahead(uselfudata((u0/opt/alt/python33/lib64/python3.3/html/parser.pyufeed�suHTMLParser.feedcCs|jd�dS(uHandle any buffered data.iN(ugoahead(uself((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuclose�suHTMLParser.closecCst||j���dS(N(uHTMLParseErrorugetpos(uselfumessage((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuerror�suHTMLParser.errorcCs|jS(u)Return full source of start tag: '<...>'.(u_HTMLParser__starttag_text(uself((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuget_starttag_text�suHTMLParser.get_starttag_textcCs2|j�|_tjd|jtj�|_dS(Nu</\s*%s\s*>(uloweru
cdata_elemureucompileuIuinteresting(uselfuelem((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuset_cdata_mode�suHTMLParser.set_cdata_modecCst|_d|_dS(N(uinteresting_normaluinterestinguNoneu
cdata_elem(uself((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuclear_cdata_mode�s uHTMLParser.clear_cdata_modec
Cs|j}d}t|�}x�||kr�|jj||�}|rT|j�}n|jraPn|}||kr�|j|||��n|j||�}||kr�Pn|j}|d|�rqt j
||�r�|j|�}n�|d|�r|j|�}n�|d|�r-|j
|�}n�|d|�rN|j|�}ng|d|�r�|jrx|j|�}q�|j|�}n+|d|kr�|jd�|d}nP|dkr\|s�Pn|jr�|jd�n|jd |d�}|dkr8|jd|d�}|dkrB|d}qBn
|d7}|j|||��n|j||�}q|d
|�rMtj
||�}|r�|j�dd�} |j| �|j�}|d|d�s�|d}n|j||�}qq�d||d�krI|j|||d��|j||d�}nPq|d
|�rtj
||�}|r�|jd�} |j| �|j�}|d|d�s�|d}n|j||�}qntj
||�}|re|ra|j�||d�kra|jr'|jd�qa|j�}||krH|}n|j||d�}nPq�|d|kr�|jd
�|j||d�}q�PqqW|r�||kr�|jr�|j|||��|j||�}n||d�|_dS(Niu<u</u<!--u<?u<!iuEOF in middle of constructu>u&#iu;u&u#EOF in middle of entity or char refi����(urawdataulenuinterestingusearchustartu
cdata_elemuhandle_datau updateposu
startswithustarttagopenumatchuparse_starttaguparse_endtagu
parse_commentuparse_piustrictuparse_declarationuparse_html_declarationuerrorufinducharrefugroupuhandle_charrefuendu entityrefuhandle_entityrefu
incomplete(
uselfuendurawdatauiunumatchuju
startswithukuname((u0/opt/alt/python33/lib64/python3.3/html/parser.pyugoahead�s�
"
uHTMLParser.goaheadcCs�|j}|||d�dkr0|j|�S|||d�dkrW|j|�S|||d�j�dkr�|jd|d�}|d
kr�dS|j||d |��|dS|j|�SdS(Niu<!--iu<![i u <!doctypeu>iii����i����(urawdatau
parse_commentuparse_marked_sectionulowerufinduhandle_decluparse_bogus_comment(uselfuiurawdataugtpos((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuparse_html_declarations
u!HTMLParser.parse_html_declarationicCs[|j}|jd|d�}|dkr/dS|rS|j||d|��n|dS(Nu>iii����i����(urawdataufinduhandle_comment(uselfuiureporturawdataupos((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuparse_bogus_comment)s uHTMLParser.parse_bogus_commentcCs`|j}tj||d�}|s)dS|j�}|j||d|��|j�}|S(Niii����(urawdataupicloseusearchustartu handle_piuend(uselfuiurawdataumatchuj((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuparse_pi5s uHTMLParser.parse_picCsd|_|j|�}|dkr(|S|j}|||�|_g}|jrltj||d�}ntj||d�}|j�}|j d�j
�|_}x'||kr�|jr�tj||�}nt
j||�}|s�Pn|j ddd�\} }
}|
s d}ns|dd�dkoK|dd�kns�|dd�dko{|d
d�knr�|dd�}n|r�|j|�}n|j| j
�|f�|j�}q�W|||�j�}|dkr�|j�\}
}d |jkrO|
|jjd �}
t|j�|jjd �}n|t|j�}|jr�|jd
|||�dd�f�n|j|||��|S|jd�r�|j||�n/|j||�||jkr|j|�n|S(Niiiiu'u"u>u/>u
u junk characters in start tag: %rii����i����i����(u>u/>(uNoneu_HTMLParser__starttag_textucheck_for_whole_start_tagurawdataustrictutagfindumatchutagfind_tolerantuendugroupulowerulasttaguattrfinduattrfind_tolerantuunescapeuappendustripugetposucountulenurfinduerroruhandle_datauendswithuhandle_startendtaguhandle_starttaguCDATA_CONTENT_ELEMENTSuset_cdata_mode(uselfuiuendposurawdatauattrsumatchukutagumuattrnameurestu attrvalueuendulinenouoffset((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuparse_starttagAs^ 00 "uHTMLParser.parse_starttagcCsk|j}|jr'tj||�}ntj||�}|r[|j�}|||d�}|dkrs|dS|dkr�|jd|�r�|dS|jd|�r�dS|jr�|j||d�|jd�n||kr�|S|dSn|dkrdS|dkrd
S|jr@|j||�|jd �n||krP|S|dSnt d
��dS(Niu>u/u/>iumalformed empty start taguu6abcdefghijklmnopqrstuvwxyz=/ABCDEFGHIJKLMNOPQRSTUVWXYZumalformed start taguwe should not get here!i����i����i����(
urawdataustrictulocatestarttagendumatchulocatestarttagend_tolerantuendu
startswithu updateposuerroruAssertionError(uselfuiurawdataumujunext((u0/opt/alt/python33/lib64/python3.3/html/parser.pyucheck_for_whole_start_tag}s> u$HTMLParser.check_for_whole_start_tagcCs�|j}tj||d�}|s)dS|j�}tj||�}|s1|jdk rw|j|||��|S|j r�|j
d|||�f�ntj||d�}|s�|||d�dkr�|dS|j|�Sn|j
d�j�}|jd|j��}|j|�|dS|j
d�j�}|jdk r�||jkr�|j|||��|Sn|j|j��|j�|S(Niubad end tag: %riiu</>u>i����(urawdatau endendtagusearchuendu
endtagfindumatchu
cdata_elemuNoneuhandle_dataustrictuerrorutagfind_tolerantuparse_bogus_commentugroupulowerufindu
handle_endtaguclear_cdata_mode(uselfuiurawdataumatchugtposu namematchutagnameuelem((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuparse_endtag�s: !
uHTMLParser.parse_endtagcCs!|j||�|j|�dS(N(uhandle_starttagu
handle_endtag(uselfutaguattrs((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuhandle_startendtag�suHTMLParser.handle_startendtagcCsdS(N((uselfutaguattrs((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuhandle_starttag�suHTMLParser.handle_starttagcCsdS(N((uselfutag((u0/opt/alt/python33/lib64/python3.3/html/parser.pyu
handle_endtag�suHTMLParser.handle_endtagcCsdS(N((uselfuname((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuhandle_charref�suHTMLParser.handle_charrefcCsdS(N((uselfuname((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuhandle_entityref�suHTMLParser.handle_entityrefcCsdS(N((uselfudata((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuhandle_data�suHTMLParser.handle_datacCsdS(N((uselfudata((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuhandle_comment�suHTMLParser.handle_commentcCsdS(N((uselfudecl((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuhandle_decl�suHTMLParser.handle_declcCsdS(N((uselfudata((u0/opt/alt/python33/lib64/python3.3/html/parser.pyu handle_pi�suHTMLParser.handle_picCs$|jr |jd|f�ndS(Nuunknown declaration: %r(ustrictuerror(uselfudata((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuunknown_decl�s uHTMLParser.unknown_declcCs8d|kr|Sdd�}tjd||dtj�S(Nu&c
SsB|j�d}yx|ddkr�|dd�}|ddkrht|dd�jd�d�}nt|jd��}t|�SWntk
r�d|SYn�Xdd lm}||kr�||S|jd�r�d
|Sx[tdt |��D]<}|d|�|kr�||d|�||d�Sq�Wd
|SdS(
Niu#iuxuXu;iu&#(uhtml5u&i(uxuX(
ugroupsuinturstripuchru
ValueErroru
html.entitiesuhtml5uendswithurangeulen(usucuhtml5ux((u0/opt/alt/python33/lib64/python3.3/html/parser.pyureplaceEntities�s&%
$u,HTMLParser.unescape.<locals>.replaceEntitiesu&&(#?[xX]?(?:[0-9a-fA-F]+;|\w{1,32};?))uflags(ureusubuASCII(uselfusureplaceEntities((u0/opt/alt/python33/lib64/python3.3/html/parser.pyuunescape�s
uHTMLParser.unescapeN(uscriptustyleF("u__name__u
__module__u__qualname__u__doc__uCDATA_CONTENT_ELEMENTSuFalseu__init__uresetufeeducloseuerroruNoneu_HTMLParser__starttag_textuget_starttag_textuset_cdata_modeuclear_cdata_modeugoaheaduparse_html_declarationuparse_bogus_commentuparse_piuparse_starttagucheck_for_whole_start_taguparse_endtaguhandle_startendtaguhandle_starttagu
handle_endtaguhandle_charrefuhandle_entityrefuhandle_datauhandle_commentuhandle_declu handle_piuunknown_decluunescape(u
__locals__((u0/opt/alt/python33/lib64/python3.3/html/parser.pyu
HTMLParser_s:
i<+*u
HTMLParser(u__doc__u_markupbaseureuwarningsucompileuinteresting_normalu
incompleteu entityrefucharrefustarttagopenupicloseucommentcloseutagfindutagfind_tolerantuattrfinduattrfind_tolerantuVERBOSEulocatestarttagendulocatestarttagend_tolerantu endendtagu
endtagfindu ExceptionuHTMLParseErroru
ParserBaseu
HTMLParser(((u0/opt/alt/python33/lib64/python3.3/html/parser.pyu<module>s0
?>