a
´iÌR ã @ sÚ d dl Z d dlZd dlmZ dgZe d¡Ze d¡Ze d¡Ze d¡Z e d¡Z
e d ¡Z e d
¡Z e d ¡Z
e d ¡Ze d
¡Ze de j¡Ze de j¡Ze de j¡Ze d
¡Ze d¡ZG dd„ dejƒZdS )é N)ÚunescapeÚ
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
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#c @ sü e Zd ZdZdZdddœdd„Zdd „ Zd
d „ Zd d
„ ZdZ dd„ Z
ddœdd„Z dd„ Z d=dd„Z
dd„ Zdd„ Zd>d d „Zd?d d „Zd!d"„ Zd#d$„ Zd%d&„ Zd'd(„ Zd)d*„ Zd+d,„ Zd-d.„ Zd/d0„ Zd1d2„ Zd3d4„ Zd5d6„ Z d7d8„ Z d9d:„ Z d;d<„ Z dS )@r )ZscriptÚstyleZxmpZiframeZnoembedZnoframes)ZtextareaÚtitleTF)Úconvert_charrefsÚ scriptingc C s || _ || _| ¡ d S ©N)r r Úreset)Úselfr r © r ú#/usr/lib64/python3.9/html/parser.pyÚ__init__v s
zHTMLParser.__init__c C s4 d| _ d| _t| _d | _d| _d| _tj | ¡ d S )NÚ z???T)
ÚrawdataÚlasttagÚinteresting_normalÚ interestingÚ
cdata_elemÚ_support_cdataÚ
_escapableÚ _markupbaseÚ
ParserBaser
©r r r r
r
„ s zHTMLParser.resetc C s | j | | _ | d¡ d S )Nr )r Úgoahead©r Údatar r r
ÚfeedŽ s zHTMLParser.feedc C s | d¡ d S )Né )r r r r r
Úclose— s zHTMLParser.closeNc C s | j S r )Ú_HTMLParser__starttag_textr r r r
Úget_starttag_text s z HTMLParser.get_starttag_text©Ú escapablec C sp | ¡ | _|| _| jdkr(t d¡| _nD|rP| jsPt d| j tjtjB ¡| _n t d| j tjtjB ¡| _d S )NÚ plaintextz\Zz&|])z])) Úlowerr r ÚreÚcompiler r Ú
IGNORECASEÚASCII)r Úelemr# r r r
Úset_cdata_mode¡ s
ÿ
ÿzHTMLParser.set_cdata_modec C s t | _d | _d| _d S )NT)r r r r r r r r
Úclear_cdata_mode s zHTMLParser.clear_cdata_modec C s
|| _ d S r )r )r Úflagr r r
Ú_set_support_cdata² s z HTMLParser._set_support_cdatac C s4 | j }d}t|ƒ}||k rÌ| jrv| jsv| d|¡}|dk r | dt||d ƒ¡}|dkrpt d¡ ||¡spqÌ|}n*| j
||¡}|r’| ¡ }n| jrœqÌ|}||k rÞ| jrÌ| j rÌ|
t|||… ƒ¡ n|
|||… ¡ | ||¡}||kröqÌ|j}|d|ƒrr r: r7 r9 r r3 r zÚ
ri Znoscriptr$ Fr" T)r Úcheck_for_whole_start_tagr Útagfind_tolerantrE rU rS r% r Úattrfind_tolerantr ÚappendÚstripZgetposÚcountr< r> rB rM Úhandle_startendtagÚhandle_starttagÚCDATA_CONTENT_ELEMENTSr r+ ÚRCDATA_CONTENT_ELEMENTS)r rY Úendposr ÚattrsrE r\ ÚtagÚmÚattrnameÚrestZ attrvaluerU ÚlinenoÚoffsetr r r
rF — sh
& ÿ
ÿ
ÿ ÿÿþ zHTMLParser.parse_starttagc C s6 | j }t ||d ¡}| ¡ }||d dkr2dS |S )Nr r r: )r Ú locatetagendrE rU rf r r r
rk Ð s z$HTMLParser.check_for_whole_start_tagc C s´ | j }| d|d ¡dk r dS t ||¡sT||d |d … dkrJ|d S | |¡S t ||d ¡}| ¡ }||d dkr€dS t ||d ¡}| d¡ ¡ }|
|¡ | ¡ |S )Nr r3 r r: r7 r ) r r= rK rE r` r} rU rl rS r% Ú
handle_endtagr, )r rY r rE r[ rw r r r
rG Ú s
zHTMLParser.parse_endtagc C s | ||¡ | |¡ d S r )rr r~ ©r rw rv r r r
rq ø s z HTMLParser.handle_startendtagc C s d S r r r r r r
rr ý s zHTMLParser.handle_starttagc C s d S r r )r rw r r r
r~ s zHTMLParser.handle_endtagc C s d S r r ©r r^ r r r
rT s zHTMLParser.handle_charrefc C s d S r r r€ r r r
rW s zHTMLParser.handle_entityrefc C s d S r r r r r r
rB
s zHTMLParser.handle_datac C s d S r r r r r r
rL s zHTMLParser.handle_commentc C s d S r r )r Zdeclr r r
rO s zHTMLParser.handle_declc C s d S r r r r r r
rP s zHTMLParser.handle_pic C s d S r r r r r r
rN s zHTMLParser.unknown_decl)T)T)r ) Ú__name__Ú
__module__Ú __qualname__rs rt r r
r r r r! r+ r, r. r rJ rH r` rI rF rk rG rq rr r~ rT rW rB rL rO rP rN r r r r
r Z s<
"
9
)r& r Zhtmlr Ú__all__r' r rX rV rR rD rK re ra rb rl ÚVERBOSErm r} Zlocatestarttagend_tolerantZ endendtagZ
endtagfindr r r r r r
Ú s2
õ
óò