qJ[ 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
e 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--\s*>z+([a-zA-Z][^\t\n\r\f />\x00]*)(?:\s|/(?!>))*z]((?<=[\'"\s/])[^\s/>][^\s/=>]*)(\s*=+\s*(\'[^\']*\'|"[^"]*"|(?![\'"])[^>\s]*))?(?:\s|/(?!>))*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 dZdddZd Zd Zd ZdZd Z d
Z
d Z d Z d
Z
d dZd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Z dS ) r )scriptstyleT)convert_charrefsc > || _ | d S N)r reset)selfr s 8/opt/alt/python-internal/lib64/python3.11/html/parser.py__init__zHTMLParser.__init__V s"
!1
c d| _ d| _ t | _ d | _ t
j | d S )N z???)rawdatalasttaginteresting_normal interesting
cdata_elem _markupbase
ParserBaser r
s r r zHTMLParser.reset_ s? ? - $$T*****r c P | j |z | _ | d d S )Nr )r goaheadr
datas r feedzHTMLParser.feedg s*
|d* Qr c 2 | d d S )N )r r s r closezHTMLParser.closep s ' Qr Nc | j S r )_HTMLParser__starttag_textr s r get_starttag_textz HTMLParser.get_starttag_textv s 7##r c | | _ t j d| j z t j | _ d S )Nz )lowerr recompileIr )r
elems r set_cdata_modezHTMLParser.set_cdata_modez s4 **,, :nt&FMMr c , t | _ d | _ d S r )r r r r s r clear_cdata_modezHTMLParser.clear_cdata_mode~ s - r c | j }d}t | }||k r| j r}| j sv| d| }|dk rY| dt
||dz
}|dk r*t j d || sn|}n=| j
|| }|r| }n | j rnZ|}||k rV| j r2| j s+| t ||| n | ||| | || }||k rn|j } |d| rt || r| | } n |d| r| | } n |d| r| | } nj |d| r| | } nH |d | r| | } n&|d
z |k r| d |d
z } nn| dk r|sn| d |d
z } | dk r%| d|d
z } | dk r|d
z } n| d
z
} | j r2| j s+| t ||| n | ||| | || }n# |d | rt. || }|rq| d
d }
| |
| } |d| d
z
s| d
z
} | || }d||d v r9| |||d
z | ||d
z }nI |d| r5t6 || }|rj| d
}
| |
| } |d| d
z
s| d
z
} | || }kt: || }|rX|rU| ||d k r5| } | |k r|} | ||d
z }n@|d
z |k r/| d | ||d
z }nn||k |ry||k rs| j sl| j r2| j s+| t ||| n | ||| | || }||d | _ d S )Nr <&" z[\s;]z>D))$/// A%:c1Q3// " E q!,,A "((!44 5u{{}} ; ; !IIKK 66 !A NN1a!e 4 4!eq[[ $$S))) q!a%00AAO !eeV %1q55 5$
/T_
/ '!A#,!7!78888 1 ...q!$$A qrr{ r c | j }|||dz dk r| | S |||dz dk r| | S |||dz dk rF| d|dz }|dk rdS | ||d z | |d
z S | | S ) N r3 zV # #%%a(( (
QqsU^u
$
$,,Q// /
QqsU^
!
!
#
#{
2
2LLac**E{{r WQqSY / 0 0 07N++A.. .r r! c | j }| d|dz }|dk rdS |r | ||dz | |dz S )Nr r4 r5 r! )r r8 handle_comment)r
rN reportr poss r rZ z HTMLParser.parse_bogus_comment sb , ll3!$$ "992 2 !C 0 1 1 1Qwr c | j }t ||dz }|sdS | }| ||dz | | }|S )Nr4 r5 )r picloser; r< handle_pirJ )r
rN r rA rP s r rE 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| || n4| || || j v r| | |S )
Nr r! r4 rV 'r5 ")r />rf )r$ check_for_whole_start_tagr tagfind_tolerantrA rJ rH r' r attrfind_tolerantr appendstripr= endswithhandle_startendtaghandle_starttagCDATA_CONTENT_ELEMENTSr, )
r
rN endposr attrsrA rR tagmattrnamerest attrvaluerJ s
r rB zHTMLParser.parse_starttag, sU # //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 $Y / / LL(..**I6 7 7 7A &jj ah %%'' k ! ! WQvX . / / /M << ) # #C / / / / e , , ,d111##C(((
r c | j }t || }|r| }|||dz }|dk r|dz S |dk r@| d| r|dz S | d| rdS ||k r|S |dz S |dk rdS |dv rdS ||k r|S |dz S t d )
Nr! r /rf r4 r5 r z6abcdefghijklmnopqrstuvwxyz=/ABCDEFGHIJKLMNOPQRSTUVWXYZzwe should not get here!)r locatestarttagend_tolerantrA rJ r? AssertionError)r
rN r rs rP nexts r rg z$HTMLParser.check_for_whole_start_tagX s , & , ,Wa 8 8 A1QqS5>Ds{{1u s{{%%dA.. ! q5L%%c1-- 2q55 H q5Lrzzr 5 6 6 r1uu1u 6777r c p | j }t ||dz }|sdS | }t || }|s| j | ||| |S t ||dz }|s+|||dz dk r|dz S | | S |
d }| d| }|
| |dz S |
d }| j *|| j k r | ||| |S |
| | |S )Nr! r5 r4 rV zr )r endendtagr; rJ
endtagfindrA r r= rh rZ rH r' r8
handle_endtagr. )r
rN r rA r[ namematchtagnamer+ s r rC zHTMLParser.parse_endtagz s , !A#.. 2 !,, * 5!1222 (..w!<U** Q3J 33A666 ooa((..00G
LLimmoo66E w ' ' '7N{{1~~##%% ? &t&& 5!1222 4 r c \ | || | | d S r )rn r r
rr rq s r rm z HTMLParser.handle_startendtag s2 S%((( 3 r c d S r r s r rn zHTMLParser.handle_starttag r c d S r r )r
rr s r r zHTMLParser.handle_endtag r r c d S r r r
rS s r rI zHTMLParser.handle_charref r r c d S r r r s r rL zHTMLParser.handle_entityref r r c d S r r r s r r= 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 rY 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 unknown_declzHTMLParser.unknown_decl r r )r! ) __name__
__module__ __qualname__ro r r r r" r$ r% r, r. r rF rZ rE rB rg rC rm rn r rI rL r= r] rY rb r r r r r r > s * 1+/ + + + O$ $ $N N N u# u# u#t/ / /* ( ( (X 8 8 8D% % %P
r )r( r htmlr __all__r) r rM rK rG r@ ra commentcloserh ri VERBOSEry r} r~ r r r r r r sj "
. RZ''
RZ
%
%
BJ> ? ?
"*@
A
Arz+&&
"*S//rz)$$ 2: LMM BJ=> > (RZ ) Z
BJsOO RZ>
?
?
I
I
I
I
I
' I
I
I
I
I
r