p( 7 T d dl Z d dlZd dlmZ dgZ 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 de j Z e j de j Z e j de j Z e j d
Z e j d Z G d dej ZdS ) 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
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 e Zd Z dZdZddddZd Zd Zd Zd
Z d Z
dd d
Z d Z d#dZ
d Zd Zd#dZd$dZd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Z d Z d! Z d" Z d
S )%r )scriptstylexmpiframenoembednoframes)textareatitleTF)convert_charrefs scriptingc L || _ || _ | d S N)r r reset)selfr r s 2/opt/alt/python311/lib64/python3.11/html/parser.py__init__zHTMLParser.__init__v s) !1 "
c d| _ d| _ t | _ d | _ d| _ d| _ t j | d S )N z???T)
rawdatalasttaginteresting_normal interesting
cdata_elem_support_cdata
_escapable _markupbase
ParserBaser r s r r zHTMLParser.reset sN ? - " $$T*****r c P | j |z | _ | d d S )Nr )r goaheadr datas r feedzHTMLParser.feed s*
|d* Qr c 2 | d d S )N )r% r# s r closezHTMLParser.close s ' Qr Nc | j S r )_HTMLParser__starttag_textr# s r get_starttag_textz HTMLParser.get_starttag_text s 7##r escapablec | | _ || _ | j dk rt j d | _ d S |rB| j s;t j d| j z t j t j z | _ d S t j d| j z t j t j z | _ d S )N plaintextz\Zz&|])z])) lowerr r recompiler r
IGNORECASEASCII)r elemr0 s r set_cdata_modezHTMLParser.set_cdata_mode s **,,# ?k ) ) !z% 0 0D
Bt 4 B !z*Dt*V*,-*@ B BD "z*BT_*T*,-*@ B BD r c : t | _ d | _ d| _ d S )NT)r r r r r# s r clear_cdata_modezHTMLParser.clear_cdata_mode s - r c || _ d S r )r )r flags r _set_support_cdataz HTMLParser._set_support_cdata s #r c
| j }d}t | }||k rI| j r}| j sv| d| }|dk rY| dt
||dz
}|dk r*t j d || sn|}n=| j
|| }|r| }n | j rn|}||k rV| j r2| j r+|
t ||| n |
||| | || }||k rn|j } |d| rt" || r| | } n |d| r| | } n |d| r| | } nl |d| r| | } nJ |d | r| | } n(|d
z |k s|r|
d |d
z } nn| dk r|snt" || rn |d| r_|d z |k r|
d nt0 || rnd| ||d z d nB |d| rU|}d D ]/}
| |
|d
z r|t |
z } n0| ||d
z | n |d| r(| j r!| ||dz d n|||dz dk r!| ||d z d ni |d | r!| ||d z d n< |d| r!| ||d z d ntA d |} | || }n# |d| rtB || }|rq|" d d } | # | |$ } |d| d
z
s| d
z
} | || }d||d v r9|
|||d z | ||d z }nI |d| r5tJ || }|rj|" d
} | & | |$ } |d| d
z
s| d
z
} | || }tN || }|rX|rU|" ||d k r5|$ } | |k r|} | ||d
z }n@|d
z |k r/|
d | ||d
z }nn||k I|rr||k rl| j r2| j r+|
t ||| n |
||| | || }||d | _ d S )Nr <&" z [\t\n\r\f ;]z >w O O $ A (//;;
AA A1uu( 3T_ 3$$Xgacl%;%;<<<<$$WQqS\222q!$$AAvvu +Jz#q!! \
6 %%gq11 ++A..AA Za(( ))!,,AA Z**
**1--AA Za((
a((AA Za(( 33A66AA!eq[[C[$$S)))AAAq55 #))'155 H #D!,, H q5A:: , ,T 2 2 2 2 ' - -gq 9 9 ? ! / /!
> > > >#FA.. H &8 & &F & / /! < < & !S[[ 0 % & ++GAaCEN;;;;#K33
H8K
H ))'!A#$$-8888 1Q3 --//;>> ((1 7777#D!,, H ++GAaCDDM::::#D!,, H wqstt}5555 ,-F G GGANN1a((D!$$ +
6
gq11 ;;==2.D''--- A%:c1Q3// " E q!,,A gabbk)) ((1Q3 888 NN1ac 2 2C##
6!33 ;;q>>D))$/// A%:c1Q3// " E q!,,A "((!44 5u{{}} ; ; !IIKK 66 !A NN1a!e 4 4!eq[[ $$S))) q!a%00AAs !eez %1q55$
/
/ '!A#,!7!78888 1 ...q!$$A qrr{ r c ( | j }|||dz dk r| | S |||dz dk rM| j rF| d|dz }|dk rdS | ||dz | |dz S |||dz d k rF| d
|dz }|dk rdS | ||d z | |d z S |||dz d
k ry| d
|dz }|dk rdS ||d z
dk r$| ||dz |d z
n | ||d z | |d z S | | S )NrF rC rI rG z]]>r rK rH rJ r rD r* zV # #%%a(( (
QqsU^{
*
*t/B
* UAaC((A1uur gac1fo . . .q5L
QqsU^
!
!
#
#{
2
2LLac**E{{r WQqSY / 0 0 07N
QqsU^u
$
$ S!A#&&A1uurqs|s""!!'!A#qs("34444##GAaCFO444q5L++A.. .r c ( | j }t ||dz }|s"t ||dz }|sdS |r4| }| ||dz | | S )NrF rK )r commentcloserQ commentabruptcloserW rR r^ rg )r rk reportr rW rm s r rZ zHTMLParser.parse_commentp s , ##GQqS11 &,,Wac::E
r 1
A !Q 0 0 0yy{{r r* c | j }| d|dz }|dk rdS |r | ||dz | |dz S )Nr rD rK r* )r rN r^ )r rk ry r poss r rt z HTMLParser.parse_bogus_comment sb , ll3!$$ "992 2 !C 0 1 1 1Qwr c | j }t ||dz }|sdS | }| ||dz | | }|S )NrD rK )r picloserQ rR rb rg r rk r rW rm s r r[ zHTMLParser.parse_pi sj ,w!,, 2 KKMM wqsAv''' IIKKr c d | _ | | }|dk r|S | j }||| | _ g }t ||dz }| }| d x| _ }||k rt || }|sn| ddd \ } }
} |
sd } nI| d d dcxk r| dd k s"n | d d dcxk r| dd k r
n n
| dd } | rt | } | | | f | }||k ||| } | dvr |
||| |S | d r| || nj| || || j v s| j r|d
k s|d k r| |d
n || j v r| |d
|S )Nr r* rD rH 'rK ")r />r noscriptr2 Fr/ T)r- check_for_whole_start_tagr tagfind_tolerantrW rg re r3 r attrfind_tolerantr appendstriprS r_ handle_startendtaghandle_starttagCDATA_CONTENT_ELEMENTSr r9 RCDATA_CONTENT_ELEMENTS)
r rk endposr attrsrW ro tagmattrnamerest attrvaluerg s
r rX zHTMLParser.parse_starttag s $ //22 A::M, &qx 0 &&w!44 IIKK "[[^^ 1 1 3 33 s&jj!''33A
()1a(8(8 %HdI
, 2A2 $8888)BCC.88882A2 #7777233 77777 %adO
0 $Y / / LL(..**I6 7 7 7A &jj ah %%'' k ! ! WQvX . / / /M <<
9 # #C / / / / e , , ,t222 3$':$5$5{""##C5#9999444##C4#888
r c | j }t ||dz }| }||dz
dk rdS |S )Nr* r rK )r locatetagendrW rg r~ s r r z$HTMLParser.check_for_whole_start_tag sL , ""7AaC00 IIKK 1Q3<3 2r c F | j }| d|dz dk rdS t || s.||dz |dz dk r|dz S | | S t
||dz }| }||dz
dk rdS t ||dz }| d }|
| | |S )Nr rD r rK rH r* ) r rN r] rW rt r rg r re r3
handle_endtagr; )r rk r rW rm r s r rY zHTMLParser.parse_endtag s , <<QqS ! !A % %2 ++ 3qs1Q3w 3&&s
//222 ""7AaC00 IIKK 1Q3<3 2 !&&w!44kk!nn""$$ 3 r c \ | || | | d S r )r r r r r s r r z HTMLParser.handle_startendtag s2 S%((( 3 r c d S r r s r r zHTMLParser.handle_starttag r c d S r r )r r s r r zHTMLParser.handle_endtag r r c d S r r r rq s r rf zHTMLParser.handle_charref r r c d S r r r s r ri zHTMLParser.handle_entityref r r c d S r r r& s r rS zHTMLParser.handle_data r r c d S r r r& s r r^ zHTMLParser.handle_comment
r r c d S r r )r decls r ra zHTMLParser.handle_decl r r c d S r r r& s r rb zHTMLParser.handle_pi r r c d S r r r& s r r` zHTMLParser.unknown_decl r r )T)r* ) __name__
__module__ __qualname__r r r r r( r+ r- r. r9 r; r> r% r\ rZ rt r[ rX r rY r r r rf ri rS r^ ra rb r` r r r r r Z s 0 Y 3+/5 + + + O$ $ $ 16
B
B
B
B
B
# # # # G# G# G#X / / /D . . .d <
r )r4 r! htmlr __all__r5 r rj rh rd rV r] r} rw rx r VERBOSEr r locatestarttagend_tolerant endendtag
endtagfindr" r r r r r s "
. RZ''
RZ
%
%
BJ> ? ?
"*@
A
Arz+&&
RZ
%
%
"*S//rz(## RZ'' 2: QRR BJ Z rz
Z
(RZ ) Z
BJsOO
RZ>
?
?
|
|
|
|
|
' |
|
|
|
|
r