python (3.11.7)
eB d Z ddlZddlZddlmZ dgZ ej d Z ej d Z ej d Z ej d Z
ej d Z ej d
Z ej d Z
ej d Z ej d
Z ej dej Z ej d
Z ej d Z G d dej ZdS )z A parser for HTML and XHTML. 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#</\s*([a-zA-Z][-.a-zA-Z0-9:_]*)\s*>c e Zd ZdZdZdddZd Zd Zd Zd Z d
Z
d Z d Z d
Z
d Zd dZd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Zd Z d Z d S ) r aE 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). If convert_charrefs is
True the character references are converted automatically to the
corresponding Unicode character (and self.handle_data() is no
longer split in chunks), otherwise they are passed by calling
self.handle_entityref() or self.handle_charref() with the string
containing respectively the named or numeric reference as the
argument.
)scriptstyleT)convert_charrefsc < || _ | dS )zInitialize and reset this instance.
If convert_charrefs is True (the default), all character references
are automatically converted to the corresponding Unicode characters.
N)r reset)selfr s ?/BuggyBox/python/3.11.7/bootstrap/lib/python3.11/html/parser.py__init__zHTMLParser.__init__V s !1
c d| _ d| _ t | _ d| _ t
j | dS )z1Reset this instance. Loses all unprocessed data. z???N)rawdatalasttaginteresting_normal interesting
cdata_elem _markupbase
ParserBaser r s r
r zHTMLParser.reset_ s< - $$T*****r c N | j |z | _ | d dS )zFeed data to the parser.
Call this as often as you want, with as little or as much text
as you want (may include '\n').
r N)r goaheadr datas r
feedzHTMLParser.feedg s% |d* Qr c 0 | d dS )zHandle any buffered data. N)r r s r
closezHTMLParser.closep s Qr Nc | j S )z)Return full source of start tag: '<...>'.)_HTMLParser__starttag_textr s r
get_starttag_textz HTMLParser.get_starttag_textv s ##r c | | _ t j d| j z t j | _ d S )Nz </\s*%s\s*>)lowerr recompileIr )r elems r
set_cdata_modezHTMLParser.set_cdata_modez s4 **,, :nt&FMMr c , t | _ d | _ d S N)r r r r s r
clear_cdata_modezHTMLParser.clear_cdata_mode~ s - r c | j }d}t | }||k r| j r}| j sv| d| }|dk rY| dt
||dz
}|dk r*t j d || sn|}n=| j
|| }|r| }n | j rnZ|}||k rV| j r2| j s+| t ||| n | ||| | || }||k rn|j } |d| rt || r| | } n |d| r| | } n |d| r| | } nj |d| r| | } nH |d | r| | } n&|d
z |k r| d |d
z } nn| dk r|sn| d |d
z } | dk r%| d|d
z } | dk r|d
z } n| d
z
} | j r2| j s+| t ||| n | ||| | || }n# |d | rt. || }|rq| d
d }
| |
| } |d| d
z
s| d
z
} | || }d||d v r9| |||d
z | ||d
z }nI |d| r5t6 || }|rj| d
}
| |
| } |d| d
z
s| d
z
} | || }kt: || }|rX|rU| ||d k r5| } | |k r|} | ||d
z }n@|d
z |k r/| d | ||d
z }nn||k |ry||k rs| j sl| j r2| j s+| t ||| n | ||| | || }||d | _ d S )Nr <&" z[\s;]z</<!--z<?z<!r r z&# ;) r lenr r findrfindmaxr' r( searchr start handle_datar updatepos
startswith starttagopenmatchparse_starttag parse_endtag
parse_commentparse_piparse_html_declarationcharrefgrouphandle_charrefend entityrefhandle_entityref
incomplete) r rJ r injampposrA r? knames r
r zHTMLParser.goahead s) ,
LL!ee$
T_
LLa((q55 %]]3Aqt = =F ! Jx 0 0 7 7 H H $ A (//;;
AA A1uu( 3 3$$Xgacl%;%;<<<<$$WQqS\222q!$$AAvvu +Jz#q!! J
6 %%gq11 ++A..AA Za(( ))!,,AA Z**
**1--AA Za((
a((AA Za(( 33A66AA!eq[[$$S)))AAAq55 S!a%00A1uu #LLa!e 4 4 q55 !AAQ, 7T_ 7 (('!A#,)?)?@@@@ ((1 666NN1a((D!$$ +
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[[ $$S))) q!a%00AAO !eeV %1q55 5$
/T_
/ '!A#,!7!78888 1 ...q!$$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 z<![ z <!doctyper r5 r4 r )r rD parse_marked_sectionr& r8 handle_declparse_bogus_comment)r rN r gtposs r
rF z!HTMLParser.parse_html_declaration s , 1QqS5>V # #%%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 z</>r )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!<<I
71QqS5>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____doc__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' r htmlr __all__r( r rM rK rG r@ ra commentcloserh ri VERBOSEry r} r~ r r r r r
<module>r sm " "
. 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 
Browse
Usage
Versions