Wj V 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, Zy) 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 dZdZddd fd
Z fdZd 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 xZ S )$r )scriptstylexmpiframenoembednoframes)textareatitleTF)convert_charrefs scriptingc ` t | || _ || _ | j y N)super__init__r r reset)selfr r __class__s $/usr/lib64/python3.12/html/parser.pyr zHTMLParser.__init__v s- 0 "
c d| _ d| _ t | _ d | _ d| _ d| _ g | _ d| _ d| _ t | - y )N z???Tr ) rawdatalasttaginteresting_normal interesting
cdata_elem_support_cdata
_escapable_pending _pending_len_parse_thresholdr r )r r s r r zHTMLParser.reset sR ? - "
!
r c | xj t | z
c_ | j | j k r | j j | y | j s| xj
|z
c_ nc| j j | | xj
dj
| j z
c_ | j j d| _ t | j
}| j d t | j
|k rd| _ y t | j
| _ y Nr r r ) r% lenr& r$ appendr joincleargoahead)r datans r feedzHTMLParser.feed s
SY& t 4 4 4 MM &== $
$$T*
66
##% !D DLL!A LLO4<< 1$()% ),DLL(9%r c | j rO| xj dj | j z
c_ | j j d| _ | j d y r( )r$ r r+ r, r% r- r s r closezHTMLParser.close sJ ' == LLBGGDMM 2 2L MM ! !D Qr Nc | j S r )_HTMLParser__starttag_textr2 s r get_starttag_textz HTMLParser.get_starttag_text s 7###r escapablec |j | _ || _ | j dk( rt j d | _ y |rT| j sHt j d| j z t j t j z | _ y t j d| j z t j t j z | _ y )N plaintextz\Zz&|])z])) lowerr! r# recompiler r
IGNORECASEASCII)r elemr8 s r set_cdata_modezHTMLParser.set_cdata_mode s **,# ??k ) !zz% 0D
t 4 4 !zz*Dt*V*,--*@ BD "zz*BT__*T*,--*@ BD r c 6 t | _ d | _ d| _ y )NT)r r r! r# r2 s r clear_cdata_modezHTMLParser.clear_cdata_mode s - r c || _ y r )r" )r flags r _set_support_cdataz HTMLParser._set_support_cdata s #r c | j }d}t | }||k r2| j rq| j se|j d| }|dk r|j dt
||dz
}|dk\ r't j d j || sn|}n?| j j || }|r|j }n| j rnx|}||k rJ| j r*| j r | j t ||| n| j ||| | j || }||k( rn|j } |d| r=t" j% || r| j' | } n |d| r| j) | } nt |d| r| j+ | } nY |d| r| j- | } n> |d | r| j/ | } n#|d
z |k s|r| j d |d
z } nnA| dk rl|sn7t" j% || rnN |d| rK|d z |k( r| j d n*t0 j% || rn| j3 ||d z d n |d| rF|}d D ]' }
|j5 |
|d
z s|t |
z } n | j3 ||d
z | n |d| r$| j6 r| j9 ||dz d n~|||dz j; dk( r| j= ||d z d nM |d | r| j3 ||d z d n, |d| r| j? ||d z d n tA d |} | j || }n |d| rtB j% || }|rY|jE d d } | jG | |jI } |d| d
z
s| d
z
} | j || }d||d v r,| j |||d z | j ||d z }n |d| rtJ j% || }|rW|jE d
} | jM | |jI } |d| d
z
s| d
z
} | j || }tN j% || }|rE|rB|jE ||d k( r,|jI } | |k r|} | j ||d
z }n7|d
z |k r'| j d | j ||d
z }nn||k r2|ra||k r\| j r*| j r | j t ||| n| j ||| | j || }||d | _ y )Nr <&" z [\t\n\r\f ;]z >w O A ((//;
A A1u((T__$$Xgal%;<$$WQq\2q!$AAvu ++J#q! %%gq1 ++A.A a( ))!,A * **1-A a(
a(A a( 33A6A!eq[C$$S)AAq5 #))'15 #D!, q5A: , ,T 2 ' - -gq 9 ! / /!
>#FA. &8F & / /! < !S[ 0 % '9 ++GAaCN;#K38K8K ))'!A#$-8 1Q3 --/;> ((1 7#D!, ++GAaCDM:#D!, wqst}5 ,-F GGANN1a(D!$
gq1 ;;=2.D''- A%c1Q3/ E q!,A gabk) ((1Q3 8 NN1ac 2C#!3 ;;q>D))$/ A%c1Q3/ E q!,A "((!4u{{} ; !IIK 6 !A NN1a!e 4!eq[ $$S) q!a%0As !ez 1q5$$ '!A,!78 1 .q!$A qr{ r c | j }|||dz dk( r| j | S |||dz dk( rC| j r7|j d|dz }|dk ry| j ||dz | |dz S |||dz j d k( r7|j d
|dz }|dk( ry| j
||d z | |d z S |||dz d
k( ra|j d
|dz }|dk ry||d z
dk( r | j ||dz |d z
|d z S | j ||d z | |d z S | j | S )NrN rK rQ rO z]]>r rS rP rR r rL r zV #%%a( (
Qqs^{
*t/B/B UAaC(A1u gac1o .q5L
Qqs^
!
!
#{
2LLac*E{ WQqS / 07N
Qqs^u
$ S!A#&A1uqs|s"!!'!A#qs"34 q5L ##GAaCO4q5L++A. .r c | j }t j ||dz }|s t j ||dz }|sy|r'|j }| j
||dz | |j S )NrN rS )r commentcloserX commentabruptcloser^ rY re rn )r rr reportr r^ rs s r ra zHTMLParser.parse_comment sr ,, ##GQqS1&,,Wac:E
A !Q 0yy{r c | j }|j d|dz }|dk( ry|r| j ||dz | |dz S )Nr rL rS r )r rU re )r rr r r poss r rz z HTMLParser.parse_bogus_comment sP ,, ll3!$ "9 !C 0 1Qwr c | j }t j ||dz }|sy|j }| j ||dz | |j }|S )NrL rS )r picloserX rY ri rn r rr r r^ rs s r rb zHTMLParser.parse_pi sV ,,w!, KKM wqsA' IIKr c d | _ | j | }|dk r|S | j }||| | _ g }t j ||dz }|j }|j
d j x| _ }||k rt j || }|sn|j
ddd \ } }
} |
sd } n,| d d dcxk( r| dd k( sn | d d dcxk( r| dd k( rn n| dd } | r t | } |j | j | f |j }||k r||| j } | dvr| j ||| |S | j d r| j || |S | j! || || j" v s| j$ r|d
k( s|d k( r| j' |d
|S || j( v r| j' |d
|S )Nr r rL rP 'rS ")r />r noscriptr: Fr7 T)r5 check_for_whole_start_tagr tagfind_tolerantr^ rn rl r; r attrfind_tolerantr r* striprZ rf handle_startendtaghandle_starttagCDATA_CONTENT_ELEMENTSr rA RCDATA_CONTENT_ELEMENTS)
r rr endposr attrsr^ ru tagmattrnamerest attrvaluern s
r r_ zHTMLParser.parse_starttag s $ //2 A:M,, &q 0 &&w!4 IIK "[[^ 1 1 33 s&j!''3A()1a(8 %HdI 2A $8)BC.82A #723 7 %aO $Y / LL(..*I6 7A &j a %%' k ! WQv . /M << # #C /
e ,t222 C:$5{"##C5#9
444##C4#8
r c | j }t j ||dz }|j }||dz
dk7 ry|S )Nr r rS )r locatetagendr^ rn r s r r z$HTMLParser.check_for_whole_start_tag sD ,, ""7AaC0 IIK 1Q3<3 r c | j }|j d|dz dk ryt j || s$||dz |dz dk( r|dz S | j | S t
j ||dz }|j
}||dz
dk7 ryt j ||dz }|j d j }| j | | j |S )Nr rL r rS rP r ) r rU rd r^ rz r rn r rl r;
handle_endtagrC )r rr r r^ rs r s r r` zHTMLParser.parse_endtag s ,, <<QqS !A % +qs1Q3 3&s
//22 ""7AaC0 IIK 1Q3<3 !&&w!4kk!n""$ 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 zHTMLParser.handle_starttag r c y r r )r r s r r zHTMLParser.handle_endtag r r c y r r r rw s r rm zHTMLParser.handle_charref r r c y r r r s r rp zHTMLParser.handle_entityref r r c y r r r r. s r rZ zHTMLParser.handle_data# r r c y r r r s r re zHTMLParser.handle_comment' r r c y r r )r decls r rh zHTMLParser.handle_decl+ r r c y r r r s r ri zHTMLParser.handle_pi/ r r c y r r r s r rg zHTMLParser.unknown_decl2 r r )T)r )!__name__
__module__ __qualname__r r r r r0 r3 r5 r6 rA rC rF r- rc ra rz rb r_ r r` r r r rm rp rZ re rh ri rg
__classcell__)r s @r r r Z s 0 Y 3+/5
:< O$ 16
B
# G#X /D .d<
r )r< _markupbasehtmlr __all__r= r rq ro rk r] rd r r} r~ r VERBOSEr r locatestarttagend_tolerant endendtag
endtagfind
ParserBaser r r r r s[ "
. RZZ'
RZZ
%
BJJ> ?
"**@
Arzz+&
RZZ
%
"**S/rzz(# RZZ' 2:: QR BJJ ZZ rzz
ZZ
(RZZ ) ZZ
BJJsO
RZZ>
?
Y
'' Y
r