python (3.12.0)
̑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( Zy)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 ZdZdZdd fd
Z fdZd 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 xZ 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 P t | || _ | j y)zInitialize and reset this instance.
If convert_charrefs is True (the default), all character references
are automatically converted to the corresponding Unicode characters.
N)super__init__r reset)selfr __class__s ?/BuggyBox/python/3.12.0/bootstrap/lib/python3.12/html/parser.pyr zHTMLParser.__init__V s! 0
c b d| _ d| _ t | _ d| _ t
| y)z1Reset this instance. Loses all unprocessed data. z???N)rawdatalasttaginteresting_normal interesting
cdata_elemr r
)r r s r r
zHTMLParser.reset` s) -
r c N | j |z | _ | j d y)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.feedh s ||d* Qr c & | j d y)zHandle any buffered data. N)r r s r closezHTMLParser.closeq s
Qr Nc | j S )z)Return full source of start tag: '<...>'.)_HTMLParser__starttag_textr s r get_starttag_textz HTMLParser.get_starttag_textw s ###r c |j | _ t j d| j z t j | _ y )Nz </\s*%s\s*>)lowerr recompileIr )r elems r set_cdata_modezHTMLParser.set_cdata_mode{ s/ **, ::nt&FMr c ( t | _ d | _ y N)r r r r s r clear_cdata_modezHTMLParser.clear_cdata_mode s - r c | j }d}t | }||k rP| 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 rn|}||k rJ| j r*| j s | j t ||| n| j ||| | j || }||k( rn,|j } |d| r[t j# || r| j% | } n |d| r| j' | } nr |d| r| j) | } nW |d| r| j+ | } n< |d | r| j- | } n!|d
z |k r| j d |d
z } nna| dk r|snX|j d |d
z } | dk r |j d|d
z } | dk r |d
z } n| d
z
} | j r*| j s | j t ||| n| j ||| | j || }n |d | rt. j# || }|rY|j1 d
d }
| j3 |
|j5 } |d| d
z
s| d
z
} | j || }d||d v r,| j |||d
z | j ||d
z }n |d| rt6 j# || }|rW|j1 d
}
| j9 |
|j5 } |d| d
z
s| d
z
} | j || }t: j# || }|rE|rB|j1 ||d k( r,|j5 } | |k r|} | j ||d
z }n7|d
z |k r'| j d | j ||d
z }nn||k rP|rm||k rh| j s\| j r*| j s | j t ||| n| j ||| | j || }||d | _ y )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 ,,
L!e$$T__LLa(q5 %]]3Aqt =F ! JJx 0 7 7 H A ((//;
A A1u(( $$Xgal%;<$$WQq\2q!$AAvu ++J#q! %%gq1 ++A.A a( ))!,A * **1-A a(
a(A a( 33A6A!eq[$$S)AAq5 S!a%0A1u #LLa!e 4 q5 !AAQ,,T__ (('!A,)?@ ((1 6NN1a(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[ $$S) q!a%0AO !eV 1q5 $$T__ '!A,!78 1 .q!$A qr{ r c L | j }|||dz dk( r| j | S |||dz dk( r| j | S |||dz j dk( r7|j d|dz }|dk( ry| j ||d z | |d
z S | j
| 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 ,, 1QqS>V #%%a( (
Qqs^u
$,,Q/ /
Qqs^
!
!
#{
2LLac*E{ WQqS / 07N++A. .r c | j }|j d|dz }|dk( ry|r| j ||dz | |dz S )Nr r4 r5 r )r r8 handle_comment)r rN reportr poss r rZ z HTMLParser.parse_bogus_comment sP ,, ll3!$ "9 !C 0 1Qwr c | j }t j ||dz }|sy|j }| j ||dz | |j }|S )Nr4 r5 )r picloser; r< handle_pirJ )r rN r rA rP s r rE zHTMLParser.parse_pi! sV ,,w!, KKM wqsA' IIKr c l 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 r| j% | |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- 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 $Y / LL(..*I6 7A &j a %%' k ! WQv . /M << # #C /
e ,d111##C(
r c H | j }t j || }|rt|j }|||dz }|dk( r|dz S |dk( r6|j d| r|dz S |j d| ry||kD r|S |dz S |dk( ry|dv ry||kD r|S |dz S t d )
Nr r /rf r4 r5 r z6abcdefghijklmnopqrstuvwxyz=/ABCDEFGHIJKLMNOPQRSTUVWXYZzwe should not get here!)r locatestarttagend_tolerantrA rJ r? AssertionError)r rN r rs rP nexts r rg z$HTMLParser.check_for_whole_start_tagY s ,, & , ,Wa 8 A1QqS>Ds{1u s{%%dA. q5L%%c1- q5 H q5Lrz 5 6 1u1u 677r c | j }t j ||dz }|sy|j }t j || }|s| j | j ||| |S t j ||dz }|s!|||dz dk( r|dz S | j | S |j d j }|j d|j }| j | |dz S |j d j }| j %|| j k7 r| j ||| |S | j | | j |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_endtag{ s[ ,, !A#. !,* 5!12 (..w!<I 1QqS>U* Q3J 33A66 ooa(..0G
LLimmo6E w '7N{{1~##% ?? &t& 5!12 4 r c J | j || | j | y r- )rn r r rr rq s r rm z HTMLParser.handle_startendtag s S%( 3 r c y r- r s r rn zHTMLParser.handle_starttag r c y r- r )r rr s r r zHTMLParser.handle_endtag r r c y r- r r rS s r rI zHTMLParser.handle_charref r r c y r- r r s r rL zHTMLParser.handle_entityref r r c y r- r r s r r= zHTMLParser.handle_data r r c y r- r r s r r] zHTMLParser.handle_comment r r c y r- r )r decls r rY zHTMLParser.handle_decl r r c y r- r r s r rb zHTMLParser.handle_pi r r c y 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
__classcell__)r s @r r r > s * 1+/ O$N u#t/* (X 8D%P
r )r r' _markupbasehtmlr __all__r( r rM rK rG r@ ra commentcloserh ri VERBOSEry r} r~
ParserBaser r r r <module>r s "
. RZZ'
RZZ
%
BJJ> ?
"**@
Arzz+&
"**S/rzz)$ 2:: LM BJJ=> (RZZ ) ZZ
BJJsO RZZ>
?
J
'' J
r 
Browse
Usage
Versions