3
Pf @ sF d dl mZmZmZ d dlmZmZmZ d dlZyd dl m
Z
W n e k
r` d dl m
Z
Y nX ddl
mZ ddl
mZ ddl
mZ dd lmZ dd
l
mZ dd lmZmZmZmZmZmZmZm Z m Z m Z m Z m Z!m"Z"m#Z#m$Z$m%Z% d!ddZ&d"ddZ'dd Z(G dd de)Z*ej+dd Z,dd Z-d#d d Z.G d d d e/Z0dS )$ )absolute_importdivisionunicode_literals)with_metaclassviewkeysPY3N) OrderedDict ) _inputstream)
_tokenizer) treebuilders)Marker)_utils)spaceCharactersasciiUpper2LowerspecialElementsheadingElements
cdataElementsrcdataElements
tokenTypes
tagTokenTypes
namespaces htmlIntegrationPointElements"mathmlTextIntegrationPointElementsadjustForeignAttributesadjustMathMLAttributesadjustSVGAttributesEReparseExceptionetreeTc K s$ t j|}t||d}|j| f|S )z.Parse a string or file-like object into a tree)namespaceHTMLElements)r getTreeBuilder
HTMLParserparse)doc treebuilderr kwargstbp r) !/usr/lib/python3.6/html5parser.pyr# s
r# divc K s, t j|}t||d}|j| fd|i|S )N)r container)r r! r"
parseFragment)r$ r, r% r r& r' r( r) r) r* r- & s
r- c s G fdddt }|S )Nc s e Zd Z fddZdS )z-method_decorator_metaclass..Decoratedc sB x0|j D ]$\}}t|tjr& |}|||< q
W tj| |||S )N)items
isinstancetypes FunctionTypetype__new__)metaZ classnamebasesZ classDictZ
attributeNameZ attribute)functionr) r* r3 . s
z5method_decorator_metaclass..Decorated.__new__N)__name__
__module__ __qualname__r3 r) )r6 r) r* Decorated- s r: )r2 )r6 r: r) )r6 r* method_decorator_metaclass, s r; c @ s e Zd ZdZd+ddZd,dd Zd
d Zed d
Zdd Z dd Z
dd Z dd Z dd Z
dd Zd-dd Zd d Zd d Zd!d" Zd#d$ Zd%d& Zd'd( Zd)d* ZdS ).r" zZHTML parser. Generates a tree structure from a stream of (possibly
malformed) HTMLNFTc sL | _ |dkrtjd}|| _g _t fddt|j D _dS )a
strict - raise an exception when a parse error is encountered
tree - a treebuilder class controlling the type of tree that will be
returned. Built in treebuilders can be accessed through
html5lib.treebuilders.getTreeBuilder(treeType)
Nr c s g | ]\}}|| j fqS r) )tree).0namecls)selfr) r*
M s z'HTMLParser.__init__..) strictr r! r< errorsdict getPhasesr. phases)r@ r< rB r debugr) )r@ r* __init__< s
zHTMLParser.__init__r+ c K sh || _ || _|| _tj|fd| i|| _| j y | j W n$ tk
rb | j | j Y nX d S )Nparser)
innerHTMLModer, scriptingr Z
HTMLTokenizer tokenizerresetmainLoopr )r@ stream innerHTMLr, rK r& r) r) r* _parseP s zHTMLParser._parsec C s | j j d| _g | _g | _d| _| jr| jj | _ | j t
krL| j j | j _
n0| j tkrd| j j| j _
n| j dkr|| j j| j _
n | jd | _| jj | j nd| _ | jd | _d | _d | _d| _d S )NFz no quirks plaintext
beforeHtmlinitialT)r< rM
firstStartTagrC log
compatModerJ r, lowerrP r rL rcdataStatestater rawtextStateplaintextStaterF phaseinsertHtmlElementresetInsertionModeZ lastPhaseZbeforeRCDataPhase
framesetOK)r@ r) r) r* rM ^ s*
zHTMLParser.resetc C s t | dsdS | jjjd jS )zThe name of the character encoding
that was used to decode the input stream,
or :obj:`None` if that is not determined yet.
rL Nr )hasattrrL rO charEncodingr> )r@ r) r) r* documentEncoding s
zHTMLParser.documentEncodingc C sJ |j dkr6|jtd kr6d|jko4|jd jtdkS |j|j ftkS d S )Nzannotation-xmlmathmlencoding text/htmlapplication/xhtml+xml)rf rg )r> namespacer
attributes translater r )r@ elementr) r) r* isHTMLIntegrationPoint s
z!HTMLParser.isHTMLIntegrationPointc C s |j |jftkS )N)rh r> r )r@ rk r) r) r* isMathMLTextIntegrationPoint s z'HTMLParser.isMathMLTextIntegrationPointc C sz t d }t d }t d }t d }t d }t d }t d }x| j D ]}d } |}
x|
d k r |
} | jjrx| jjd nd } | r| jnd } | r| jnd }
|
d }||kr| j|
d
|
jd i d }
qVt| jjd ksl| | jj ksl| j
| r ||kr|d
t ddgksl|||fksl| t d krP|
dkrP||krP|d
dksl| j
| rt||||fkrt| j}n
| jd }||kr|j|
}
qV||kr|j|
}
qV||kr|j|
}
qV||kr|j|
}
qV||kr|j|
}
qV||krV|j|
}
qVW ||krD| d rD| d rD| jdd
| d
i qDW d}g }x8|rt|j| j | jj }|r>| j|ks>tq>W d S )N
CharactersZSpaceCharactersStartTagEndTagCommentZDoctype
ParseErrorr r2 datadatavarsr r> ZmglyphZ
malignmarkrd zannotation-xmlsvginForeignContent selfClosingselfClosingAcknowledgedz&non-void-element-with-trailing-solidusT)r normalizedTokensr< openElementsrh r>
parseErrorgetlendefaultNamespacerm frozensetr rl r] rF processCharactersprocessSpaceCharactersprocessStartTag
processEndTagprocessCommentprocessDoctypeappend
processEOFAssertionError)r@ ZCharactersTokenZSpaceCharactersTokenZ
StartTagTokenZ EndTagTokenZ CommentTokenZ DoctypeTokenZParseErrorTokentokenZ
prev_token new_token currentNodeZcurrentNodeNamespaceZcurrentNodeNamer2 r] Z reprocessrF r) r) r* rN sp
zHTMLParser.mainLoopc c s x| j D ]}| j|V qW d S )N)rL normalizeToken)r@ r r) r) r* rz s zHTMLParser.normalizedTokensc O s | j |ddf|| | jj S )a Parse a HTML document into a well-formed tree
stream - a filelike object or string containing the HTML to be parsed
The optional encoding parameter must be a string that indicates
the encoding. If specified, that encoding will be used,
regardless of any BOM or later declaration (such as in a meta
element)
scripting - treat noscript elements as if javascript was turned on
FN)rQ r< Z getDocument)r@ rO argsr& r) r) r* r# s zHTMLParser.parsec O s | j |df|| | jj S )a2 Parse a HTML fragment into a well-formed tree fragment
container - name of the element we're setting the innerHTML property
if set to None, default to 'div'
stream - a filelike object or string containing the HTML to be parsed
The optional encoding parameter must be a string that indicates
the encoding. If specified, that encoding will be used,
regardless of any BOM or later declaration (such as in a meta
element)
scripting - treat noscript elements as if javascript was turned on
T)rQ r< Z getFragment)r@ rO r r& r) r) r* r- s zHTMLParser.parseFragmentXXX-undefined-errorc C s@ |d kr i }| j j| jjj ||f | jr rP r rh r rF r] )r@ ZlastZnewModesnodeZnodeNameZ new_phaser) r) r* r_ ! sB
z HTMLParser.resetInsertionModec C sR |dks t | jj| |dkr.| jj| j_n | jj| j_| j| _| j d | _dS )zYGeneric RCDATA/RAWTEXT Parsing algorithm
contentType - RCDATA or RAWTEXT
RAWTEXTRCDATAtextN)r r )
r r<
insertElementrL r[ rZ rY r]
originalPhaserF )r@ r Z contentTyper) r) r* parseRCDataRawtextM s z HTMLParser.parseRCDataRawtext)NFTF)Fr+ F)r N)r7 r8 r9 __doc__rH rQ rM propertyrc rl rm rN rz r# r- r| r r r r r r_ r r) r) r) r* r" 8 s&
"
C
,r" c s" dd }dd }G dd dt || | G dd d }G d d
d
}G fd d d }G fd
dd }G fddd }G fddd }G fddd } G fddd }
G fddd } G fddd } G fdd d }
G fd d d }G fd d d }G fd!d"d" }G fd#d$d$ }G fd%d&d& }G fd'd(d( }G fd)d*d* }G fd+d,d, }G fd-d.d. }G fd/d0d0 }G fd1d2d2 }G fd3d4d4 }||||||| |
| | |
||||||||||||d5S )6Nc s( t dd tj D fdd}|S )z4Logger that records which phase processes each tokenc s s | ]\}}||fV qd S )Nr) )r= keyvaluer) r) r* c s z)getPhases..log..c s j jdrt|dkr|d }yd|d i}W n Y nX |d tkr\|d |d< | jjj| jjjj | jj j
j | j
j j |f | f||S | f||S d S )NZprocessr r2 r> ) r7
startswithr~ r rI rV r rL rZ r] __class__)r@ r r& r info)r6
type_namesr) r* wrappedf s
z'getPhases..log..wrapped)rD r r. )r6 r r) )r6 r r* rV a s zgetPhases..logc S s | r t |S tS d S )N)r; r2 )Z
use_metaclassZmetaclass_funcr) r) r* getMetaclassz s z getPhases..getMetaclassc @ sX e Zd ZdZdd Zdd Zdd Zdd Zd
d Zd d
Z dd Z
dd Z dd Z dS )zgetPhases..PhasezNBase class for helper object that implements each phase of processing
c S s || _ || _d S )N)rI r< )r@ rI r< r) r) r* rH s z!getPhases..Phase.__init__c S s t d S )N)NotImplementedError)r@ r) r) r* r s z#getPhases..Phase.processEOFc S s | j j|| j jd d S )Nr ry )r<
insertCommentr{ )r@ r r) r) r* r s z'getPhases..Phase.processCommentc S s | j jd d S )Nzunexpected-doctype)rI r| )r@ r r) r) r* r s z'getPhases..Phase.processDoctypec S s | j j|d d S )Nrs )r<
insertText)r@ r r) r) r* r s z*getPhases..Phase.processCharactersc S s | j j|d d S )Nrs )r< r )r@ r r) r) r* r s z/getPhases..Phase.processSpaceCharactersc S s | j |d |S )Nr> )startTagHandler)r@ r r) r) r* r s z(getPhases..Phase.processStartTagc S sl | j j r"|d dkr"| j jd x<|d j D ],\}}|| jjd jkr0|| jjd j|< q0W d| j _d S )Nr> r z
non-html-rootrs r F)rI rU r| r. r< r{ ri )r@ r attrr r) r) r* startTagHtml s z%getPhases..Phase.startTagHtmlc S s | j |d |S )Nr> )
endTagHandler)r@ r r) r) r* r s z&getPhases..Phase.processEndTagN)
r7 r8 r9 r rH r r r r r r r r r) r) r) r* Phase s
r c @ sL e Zd Zdd Zdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dS )z getPhases..InitialPhasec S s d S )Nr) )r@ r r) r) r* r s z6getPhases..InitialPhase.processSpaceCharactersc S s | j j|| j j d S )N)r< r document)r@ r r) r) r* r s z.getPhases..InitialPhase.processCommentc 8 S s |d }|d }|d }|d }|dks@|d k s@|d k rL|dkrL| j jd |d krXd}| jj| |dkrv|jt}| s|d dk s|jdJ s|dKk s|jdL r|d k s| r|j dDk rdE| j _n*|jdM s|jdNr |d k r dH| j _| j j dI | j _
d S )ONr> publicIdsystemIdcorrectr zabout:legacy-compatzunknown-doctype *+//silmaril//dtd html pro v0r11 19970101//4-//advasoft ltd//dtd html 3.0 aswedit + extensions//*-//as//dtd html 3.0 aswedit + extensions// -//ietf//dtd html 2.0 level 1// -//ietf//dtd html 2.0 level 2//&-//ietf//dtd html 2.0 strict level 1//&-//ietf//dtd html 2.0 strict level 2// -//ietf//dtd html 2.0 strict//-//ietf//dtd html 2.0//-//ietf//dtd html 2.1e//-//ietf//dtd html 3.0// -//ietf//dtd html 3.2 final//-//ietf//dtd html 3.2//-//ietf//dtd html 3//-//ietf//dtd html level 0//-//ietf//dtd html level 1//-//ietf//dtd html level 2//-//ietf//dtd html level 3//"-//ietf//dtd html strict level 0//"-//ietf//dtd html strict level 1//"-//ietf//dtd html strict level 2//"-//ietf//dtd html strict level 3//-//ietf//dtd html strict//-//ietf//dtd html//(-//metrius//dtd metrius presentational//5-//microsoft//dtd internet explorer 2.0 html strict//.-//microsoft//dtd internet explorer 2.0 html//0-//microsoft//dtd internet explorer 2.0 tables//5-//microsoft//dtd internet explorer 3.0 html strict//.-//microsoft//dtd internet explorer 3.0 html//0-//microsoft//dtd internet explorer 3.0 tables//#-//netscape comm. corp.//dtd html//*-//netscape comm. corp.//dtd strict html//*-//o'reilly and associates//dtd html 2.0//3-//o'reilly and associates//dtd html extended 1.0//;-//o'reilly and associates//dtd html extended relaxed 1.0//N-//softquad software//dtd hotmetal pro 6.0::19990601::extensions to html 4.0//E-//softquad//dtd hotmetal pro 4.0::19971010::extensions to html 4.0//$-//spyglass//dtd html 2.0 extended//+-//sq//dtd html 2.0 hotmetal + extensions//--//sun microsystems corp.//dtd hotjava html//4-//sun microsystems corp.//dtd hotjava strict html// -//w3c//dtd html 3 1995-03-24// -//w3c//dtd html 3.2 draft// -//w3c//dtd html 3.2 final//-//w3c//dtd html 3.2// -//w3c//dtd html 3.2s draft// -//w3c//dtd html 4.0 frameset//#-//w3c//dtd html 4.0 transitional//(-//w3c//dtd html experimental 19960712//&-//w3c//dtd html experimental 970421//-//w3c//dtd w3 html//-//w3o//dtd w3 html 3.0//#-//webtechs//dtd mozilla html 2.0// -//webtechs//dtd mozilla html//$-//w3o//dtd w3 html strict 3.0//en//"-/w3c/dtd html 4.0 transitional/en -//w3c//dtd html 4.01 frameset//$-//w3c//dtd html 4.01 transitional//z:http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtdquirks -//w3c//dtd xhtml 1.0 frameset//$-//w3c//dtd xhtml 1.0 transitional//zlimited quirksrS )7r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r )r r r )r r )r r
)r r ) rI r| r< Z
insertDoctyperj r r rX rW rF r] )r@ r r> r r r r) r) r* r s
z.getPhases..InitialPhase.processDoctypec S s d| j _| j jd | j _d S )Nr rS )rI rW rF r] )r@ r) r) r* anythingElse s z,getPhases..InitialPhase.anythingElsec S s | j jd | j |S )Nz expected-doctype-but-got-chars)rI r| r )r@ r r) r) r* r s z1getPhases..InitialPhase.processCharactersc S s" | j jdd|d i | j |S )Nz"expected-doctype-but-got-start-tagr> )rI r| r )r@ r r) r) r* r s z/getPhases..InitialPhase.processStartTagc S s" | j jdd|d i | j |S )Nz expected-doctype-but-got-end-tagr> )rI r| r )r@ r r) r) r* r s z-getPhases..InitialPhase.processEndTagc S s | j jd | j dS )Nz expected-doctype-but-got-eofT)rI r| r )r@ r) r) r* r % s z*getPhases..InitialPhase.processEOFN) r7 r8 r9 r r r r r r r r r) r) r) r* InitialPhase s _r c @ sD e Zd Zdd Zdd Zdd Zdd Zd d
Zd d Zd
d Z dS )z"getPhases..BeforeHtmlPhasec S s& | j jtdd | jjd | j_d S )Nr ro r )r< Z
insertRootimpliedTagTokenrI rF r] )r@ r) r) r* r^ , s z4getPhases..BeforeHtmlPhase.insertHtmlElementc S s | j dS )NT)r^ )r@ r) r) r* r 1 s z-getPhases..BeforeHtmlPhase.processEOFc S s | j j|| j j d S )N)r< r r )r@ r r) r) r* r 5 s z1getPhases..BeforeHtmlPhase.processCommentc S s d S )Nr) )r@ r r) r) r* r 8 s z9getPhases..BeforeHtmlPhase.processSpaceCharactersc S s | j |S )N)r^ )r@ r r) r) r* r ; s z4getPhases..BeforeHtmlPhase.processCharactersc S s |d dkrd| j _| j |S )Nr> r T)rI rU r^ )r@ r r) r) r* r ? s z2getPhases..BeforeHtmlPhase.processStartTagc S s4 |d dkr$| j jdd|d i n | j |S d S )Nr> r r r brz unexpected-end-tag-before-html)r r r r )rI r| r^ )r@ r r) r) r* r E s
z0getPhases..BeforeHtmlPhase.processEndTagN)
r7 r8 r9 r^ r r r r r r r) r) r) r* BeforeHtmlPhase* s r c sX e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dS )z"getPhases..BeforeHeadPhasec sV j | || tjd| jfd| jfg| _| j| j_tjd| jfg| _ | j
| j _d S )Nr r r r )r r r r ) rH r MethodDispatcherr startTagHeadr
startTagOtherdefaultendTagImplyHeadr endTagOther)r@ rI r< )r r) r* rH N s
z+getPhases..BeforeHeadPhase.__init__c S s | j tdd dS )Nr ro T)r r
)r@ r) r) r* r \ s z-getPhases..BeforeHeadPhase.processEOFc S s d S )Nr) )r@ r r) r) r* r ` s z9getPhases..BeforeHeadPhase.processSpaceCharactersc S s | j tdd |S )Nr ro )r r
)r@ r r) r) r* r c s z4getPhases..BeforeHeadPhase.processCharactersc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r g s z/getPhases..BeforeHeadPhase.startTagHtmlc S s0 | j j| | j jd | j _| jjd | j_d S )Nr inHeadry )r< r r{ headPointerrI rF r] )r@ r r) r) r* r j s z/getPhases..BeforeHeadPhase.startTagHeadc S s | j tdd |S )Nr ro )r r
)r@ r r) r) r* r o s z0getPhases..BeforeHeadPhase.startTagOtherc S s | j tdd |S )Nr ro )r r
)r@ r r) r) r* r s s z2getPhases..BeforeHeadPhase.endTagImplyHeadc S s | j jdd|d i d S )Nzend-tag-after-implied-rootr> )rI r| )r@ r r) r) r* r w s z.getPhases..BeforeHeadPhase.endTagOtherN) r7 r8 r9 rH r r r r r r r r r) )r r) r* BeforeHeadPhaseM s r c s e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd Zdd Zdd Zd d Zd d Zd!S )"z getPhases..InHeadPhasec s j | || tjd| jfd| jfd| jfd| jfd| jfd| jfd | j fd
| j
fg| _ | j | j _
tjd
| jfd| jfg| _| j| j_
d S )Nr titlenoframesstylenoscriptscriptbasebasefontbgsoundcommandlinkr4 r r r )r r )r r r r! r" )r r r )rH r r r
startTagTitlestartTagNoFramesStylestartTagNoscriptstartTagScriptstartTagBaseLinkCommand startTagMetar r r r
endTagHeadendTagHtmlBodyBrr r )r@ rI r< )r r) r* rH | s
z'getPhases..InHeadPhase.__init__c S s | j dS )NT)r )r@ r) r) r* r s z)getPhases..InHeadPhase.processEOFc S s | j |S )N)r )r@ r r) r) r* r s z0getPhases..InHeadPhase.processCharactersc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z+getPhases..InHeadPhase.startTagHtmlc S s | j jd d S )Nz!two-heads-are-not-better-than-one)rI r| )r@ r r) r) r* r s z+getPhases..InHeadPhase.startTagHeadc S s$ | j j| | j jj d|d< d S )NTrx )r< r r{ pop)r@ r r) r) r* r' s z6getPhases..InHeadPhase.startTagBaseLinkCommandc S s | j j| | j jj d|d< |d }| jjjjd dkrd|krZ| jjjj|d nVd|krd|kr|d j d krt
j |d j d
}t
j
|}|j }| jjjj| d S ) NTrx rs r Z tentativecharsetZcontentz
http-equivz content-typezutf-8)r< r r{ r+ rI rL rO rb ZchangeEncodingrX r
Z
EncodingBytesencodeZContentAttrParserr# )r@ r ri rs rI codecr) r) r* r( s
z+getPhases..InHeadPhase.startTagMetac S s | j j|d d S )Nr )rI r )r@ r r) r) r* r# s z,getPhases..InHeadPhase.startTagTitlec S s | j j|d d S )Nr )rI r )r@ r r) r) r* r$ s z4getPhases..InHeadPhase.startTagNoFramesStylec S s8 | j jr| j j|d n | jj| | j jd | j _d S )Nr inHeadNoscript)rI rK r r< r rF r] )r@ r r) r) r* r% s z/getPhases..InHeadPhase.startTagNoscriptc S s< | j j| | jjj| jj_| jj| j_| jjd | j_d S )Nr ) r< r rI rL ZscriptDataStaterZ r] r rF )r@ r r) r) r* r& s z-getPhases..InHeadPhase.startTagScriptc S s | j |S )N)r )r@ r r) r) r* r s z,getPhases..InHeadPhase.startTagOtherc S s: | j jjj }|jdks&td|j | j jd | j _d S )Nr zExpected head got %s afterHead)rI r< r{ r+ r> r rF r] )r@ r r r) r) r* r) s z)getPhases..InHeadPhase.endTagHeadc S s | j |S )N)r )r@ r r) r) r* r* s z/getPhases..InHeadPhase.endTagHtmlBodyBrc S s | j jdd|d i d S )Nzunexpected-end-tagr> )rI r| )r@ r r) r) r* r s z*getPhases..InHeadPhase.endTagOtherc S s | j td d S )Nr )r) r
)r@ r) r) r* r s z+getPhases..InHeadPhase.anythingElseN)r7 r8 r9 rH r r r r r' r( r# r$ r% r& r r) r* r r r) )r r) r* InHeadPhase{ s r1 c sx e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd Zdd ZdS ) z&getPhases..InHeadNoscriptPhasec sf j | || tjd| jfd | jfd | jfg| _| j| j_tjd | j fd
| j
fg| _ | j | j _d S )
Nr r r r" r4 r r r r r )r r r" r4 r r )r r )
rH r r r r' startTagHeadNoscriptr r r endTagNoscriptendTagBrr r )r@ rI r< )r r) r* rH s
z/getPhases..InHeadNoscriptPhase.__init__c S s | j jd | j dS )Nzeof-in-head-noscriptT)rI r| r )r@ r) r) r* r s z1getPhases..InHeadNoscriptPhase.processEOFc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z5getPhases..InHeadNoscriptPhase.processCommentc S s | j jd | j |S )Nzchar-in-head-noscript)rI r| r )r@ r r) r) r* r s z8getPhases..InHeadNoscriptPhase.processCharactersc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z=getPhases..InHeadNoscriptPhase.processSpaceCharactersc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z3getPhases..InHeadNoscriptPhase.startTagHtmlc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r' s z>getPhases..InHeadNoscriptPhase.startTagBaseLinkCommandc S s | j jdd|d i d S )Nzunexpected-start-tagr> )rI r| )r@ r r) r) r* r2 s z;getPhases..InHeadNoscriptPhase.startTagHeadNoscriptc S s" | j jdd|d i | j |S )Nz unexpected-inhead-noscript-tagr> )rI r| r )r@ r r) r) r* r s z4getPhases..InHeadNoscriptPhase.startTagOtherc S s: | j jjj }|jdks&td|j | j jd | j _d S )Nr zExpected noscript got %sr )rI r< r{ r+ r> r rF r] )r@ r r r) r) r* r3 s z5getPhases..InHeadNoscriptPhase.endTagNoscriptc S s" | j jdd|d i | j |S )Nz unexpected-inhead-noscript-tagr> )rI r| r )r@ r r) r) r* r4 s z/getPhases..InHeadNoscriptPhase.endTagBrc S s | j jdd|d i d S )Nzunexpected-end-tagr> )rI r| )r@ r r) r) r* r s z2getPhases..InHeadNoscriptPhase.endTagOtherc S s | j td d S )Nr )r3 r
)r@ r) r) r* r s z3getPhases..InHeadNoscriptPhase.anythingElseN)r7 r8 r9 rH r r r r r r' r2 r r3 r4 r r r) )r r) r* InHeadNoscriptPhase s r5 c sp e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd ZdS )z!getPhases..AfterHeadPhasec
sn j | || tjd| jfd| jfd| jfd| jfd
| jfg| _| j | j_
tjd| j fg| _ | j
| j _
d S )Nr r r r r r r" r4 r r r r r r ) r r r r" r4 r r r r )r r r )rH r r r startTagBodystartTagFramesetstartTagFromHeadr r r r r* r r )r@ rI r< )r r) r* rH # s
z*getPhases..AfterHeadPhase.__init__c S s | j dS )NT)r )r@ r) r) r* r 4 s z,getPhases..AfterHeadPhase.processEOFc S s | j |S )N)r )r@ r r) r) r* r 8 s z3getPhases..AfterHeadPhase.processCharactersc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r < s z.getPhases..AfterHeadPhase.startTagHtmlc S s( d| j _| jj| | j jd | j _d S )NFr )rI r` r< r rF r] )r@ r r) r) r* r6 ? s z.getPhases..AfterHeadPhase.startTagBodyc S s | j j| | jjd | j_d S )Nr )r< r rI rF r] )r@ r r) r) r* r7 D s z2getPhases..AfterHeadPhase.startTagFramesetc S st | j jdd|d i | jjj| jj | j jd j| x4| jjd d d D ] }|jdkrN| jjj | P qNW d S )Nz#unexpected-start-tag-out-of-my-headr> r r r ry )
rI r| r< r{ r r rF r r> remove)r@ r r r) r) r* r8 H s
z2getPhases..AfterHeadPhase.startTagFromHeadc S s | j jdd|d i d S )Nzunexpected-start-tagr> )rI r| )r@ r r) r) r* r R s z.getPhases..AfterHeadPhase.startTagHeadc S s | j |S )N)r )r@ r r) r) r* r U s z/getPhases..AfterHeadPhase.startTagOtherc S s | j |S )N)r )r@ r r) r) r* r* Y s z2getPhases..AfterHeadPhase.endTagHtmlBodyBrc S s | j jdd|d i d S )Nzunexpected-end-tagr> )rI r| )r@ r r) r) r* r ] s z-getPhases..AfterHeadPhase.endTagOtherc S s. | j jtdd | jjd | j_d| j_d S )Nr ro r T)r< r r
rI rF r] r` )r@ r) r) r* r ` s z.getPhases..AfterHeadPhase.anythingElseN)r7 r8 r9 rH r r r r6 r7 r8 r r r* r r r) )r r) r* AfterHeadPhase" s
r: c s e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd Zdd Zdd Zd d Zd d Zd!d" Zd#d$ Zd%d& Zd'd( Zd)d* Zd+d, Zd-d. Zd/d0 Zd1d2 Zd3d4 Z d5d6 Z d7d8 Z d9d: Z d;d< Z d=d> Z!d?d@ Z"dAdB Z#dCdD Z$dEdF Z%dGdH Z&dIdJ Z'dKdL Z(dMdN Z)dOdP Z*dQdR Z+dSdT Z,dUdV Z-dWdX Z.dYdZ Z/d[d\ Z0d]d^ Z1d_d` Z2dadb Z3dcdd Z4dedf Z5dgS )hz getPhases..InBodyPhasec , s j | || | j| _tjd| jfdd| jfd | jfd | jfde| j ft
| j fdf| j fd&| j
fdg| jfd*| jfd+| jfdh| jfd8| jfd9| jfdi| jfd=| jfd>| jfdj| jfdk| jfdH| jfdI| jfdJ| jfdK| j fdL| j fdM| j fdN| j fdl| j fdQ| j!fdm| j"fdn| j#fdV| j$fdW| j%fdo| j&fg!| _'| j(| j'_)tjd | j*fd| j+fdp| j,fd&| j-fd | j.fdq| j/ft
| j0fdr| j1fds| j2fd@| j3fg
| _4| j5| j4_)d S )tNr r r r r! r" r4 r r r r r addressarticleaside
blockquotecenterdetailsdirr+ dlfieldset
figcaptionfigurefooterheaderhgroupmainmenunavolr( sectionsummaryulprelistingformlidddtrR abbigcodeemfontissmallstrikestrongttunobrbuttonappletmarqueeobjectZxmpr arear embedimgkeygenwbrparamsourcetrackinputhrimageisindextextareaZiframer noembedr r rprtoptionoptgroupZmathru r colr framer r r r r r r dialog) r r r r! r" r4 r r r )r; r< r= r> r? r@ rA r+ rB rC rD rE rF rG rH rI rJ rK rL r( rM rN rO )rP rQ )rS rT rU ) rW rX rY rZ r[ r\ r] r^ r_ r` ra rb )re rf rg )rh r ri rj rk rl )rm rn ro )ru r )rv rw )rx ry ) r rz r r{ r r r r r r r )r; r< r= r> rd r? r@ r| rA r+ rB rC rD rE rF rG rH rQ rI rJ rK rL rP rM rN rO )rT rU rS )rV rW rX rY rZ r[ r\ rc r] r^ r_ r` ra rb )re rf rg )6rH processSpaceCharactersNonPrer r r r startTagProcessInHeadr6 r7 startTagClosePr startTagHeadingstartTagPreListing startTagFormstartTagListItemstartTagPlaintext startTagAstartTagFormatting startTagNobrstartTagButtonstartTagAppletMarqueeObject startTagXmp
startTagTablestartTagVoidFormattingstartTagParamSource
startTagInput
startTagHr
startTagImagestartTagIsIndexstartTagTextareastartTagIFramer% startTagRawtextstartTagSelect startTagRpRt startTagOpt startTagMath startTagSvgstartTagMisplacedr r r
endTagBody
endTagHtml endTagBlock
endTagFormendTagPendTagListItem
endTagHeadingendTagFormattingendTagAppletMarqueeObjectr4 r r )r@ rI r< )r r) r* rH h s~
z'getPhases..InBodyPhase.__init__c S s$ |j |j ko"|j|jko"|j|jkS )N)r> rh ri )r@ Znode1Znode2r) r) r* isMatchingFormattingElement s z:getPhases..InBodyPhase.isMatchingFormattingElementc S s | j j| | j jd }g }x<| j jd d d D ]&}|tkr@P q0| j||r0|j| q0W t|dksjtt|dkr| j jj |d | j jj| d S )Nr ry ry ry )
r< r r{ activeFormattingElementsr
r r r~ r r9 )r@ r rk ZmatchingElementsr r) r) r* addFormattingElement s z3getPhases..InBodyPhase.addFormattingElementc
S s@ t d}x2| jjd d d D ] }|j|kr | jjd P q W d S )NrT rU rS r( r r r r r r r r r z expected-closing-tag-but-got-eof) rT rU rS r( r r r r r r r r ry )r r< r{ r> rI r| )r@ Zallowed_elementsr r) r) r* r s
z)getPhases..InBodyPhase.processEOFc S sh |d }| j | _|jdrJ| jjd jdkrJ| jjd j rJ|dd }|rd| jj | jj| d S )
Nrs
r rP rQ rt ry )rP rQ rt ry ) r} r r r< r{ r> Z
hasContent#reconstructActiveFormattingElementsr )r@ r rs r) r) r* !processSpaceCharactersDropNewline s
z@getPhases..InBodyPhase.processSpaceCharactersDropNewlinec S sT |d dkrd S | j j | j j|d | jjrPtdd |d D rPd| j_d S )Nrs c S s g | ] }|t kqS r) )r )r= charr) r) r* rA s zDgetPhases..InBodyPhase.processCharacters..F)r< r r rI r` any)r@ r r) r) r* r s
z0getPhases..InBodyPhase.processCharactersc S s | j j | j j|d d S )Nrs )r< r r )r@ r r) r) r* r} s
z;getPhases..InBodyPhase.processSpaceCharactersNonPrec S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r~ s z4getPhases..InBodyPhase.startTagProcessInHeadc S s | j jdddi t| jjdks4| jjd jdkrB| j jstnFd| j _x<|d j D ],\}}|| jjd j
krX|| jjd j
|< qXW d S )Nzunexpected-start-tagr> r r Frs ) rI r| r~ r< r{ r> rP r r` r. ri )r@ r r r r) r) r* r6 s z+getPhases..InBodyPhase.startTagBodyc S s | j jdddi t| jjdks4| jjd jdkrB| j jstnt| j jsLnj| jjd j rv| jjd j j
| jjd x"| jjd jdkr| jjj qxW | jj | | j j
d | j _d S ) Nzunexpected-start-tagr> r r r r r ry )rI r| r~ r< r{ r> rP r r` parent removeChildr+ r rF r] )r@ r r) r) r* r7 s " z/getPhases..InBodyPhase.startTagFramesetc S s. | j jdddr | jtd | j j| d S )Nr( rd )variant)r< elementInScoper r
r )r@ r r) r) r* r s z-getPhases..InBodyPhase.startTagClosePc S s> | j jdddr | jtd | j j| d| j_| j| _d S )Nr( rd )r F) r< r r r
r rI r` r r )r@ r r) r) r* r s
z1getPhases..InBodyPhase.startTagPreListingc S sZ | j jr | jjdddi n:| j jdddr:| jtd | j j| | j jd | j _d S ) Nzunexpected-start-tagr> rR r( rd )r r ry ) r< formPointerrI r| r r r
r r{ )r@ r r) r) r* r s z+getPhases..InBodyPhase.startTagFormc S s d| j _dgddgddgd}||d }xLt| jjD ]<}|j|kr^| j jjt|jd P |j t
kr8|jd
kr8P q8W | jj d
d d r| j jjtd
d | jj | d S )NFrS rU rT )rS rU rT r> rp r; r+ r( rd )r )r; r+ r( )
rI r` reversedr< r{ r> r] r r
nameTupler r r )r@ r Z stopNamesMapZ stopNamesr r) r) r* r s"
z/getPhases..InBodyPhase.startTagListItemc S s> | j jdddr | jtd | j j| | jjj| jj_d S )Nr( rd )r ) r< r r r
r rI rL r\ rZ )r@ r r) r) r* r 4 s z0getPhases..InBodyPhase.startTagPlaintextc S sb | j jdddr | jtd | j jd jtkrR| jjdd|d i | j jj | j j
| d S )Nr( rd )r r zunexpected-start-tagr> ry ) r< r r r
r{ r> r rI r| r+ r )r@ r r) r) r* r : s z.getPhases..InBodyPhase.startTagHeadingc S s~ | j jd}|rf| jjdddd | jtd || j jkrL| j jj| || j jkrf| j jj| | j j | j
| d S )NrV z$unexpected-start-tag-implies-end-tag) startNameendName) r< !elementInActiveFormattingElementsrI r| r r
r{ r9 r r r )r@ r Z afeAElementr) r) r* r B s
z(getPhases..InBodyPhase.startTagAc S s | j j | j| d S )N)r< r r )r@ r r) r) r* r O s
z1getPhases..InBodyPhase.startTagFormattingc S sP | j j | j jdrB| jjdddd | jtd | j j | j| d S )Nrc z$unexpected-start-tag-implies-end-tag)r r )r< r r rI r| r r
r )r@ r r) r) r* r S s
z+getPhases..InBodyPhase.startTagNobrc S sT | j jdr2| jjdddd | jtd |S | j j | j j| d| j_d S )Nrd z$unexpected-start-tag-implies-end-tag)r r F) r< r rI r| r r
r r r` )r@ r r) r) r* r ] s
z-getPhases..InBodyPhase.startTagButtonc S s0 | j j | j j| | j jjt d| j_d S )NF)r< r r r r r
rI r` )r@ r r) r) r* r h s
z:getPhases..InBodyPhase.startTagAppletMarqueeObjectc S sB | j jdddr | jtd | j j d| j_| jj|d d S )Nr( rd )r Fr )r< r r r
r rI r` r )r@ r r) r) r* r n s
z*getPhases..InBodyPhase.startTagXmpc S sR | j jdkr*| jjdddr*| jtd | jj| d| j _| j jd | j _ d S )Nr r( rd )r Fr )
rI rW r< r r r
r r` rF r] )r@ r r) r) r* r u s z,getPhases..InBodyPhase.startTagTablec S s6 | j j | j j| | j jj d|d< d| j_d S )NTrx F)r< r r r{ r+ rI r` )r@ r r) r) r* r } s
z5getPhases..InBodyPhase.startTagVoidFormattingc S s@ | j j}| j| d|d kr<|d d jtdkr<|| j _d S )Nr2 rs hidden)rI r` r rj r )r@ r r` r) r) r* r s
z,getPhases..InBodyPhase.startTagInputc S s$ | j j| | j jj d|d< d S )NTrx )r< r r{ r+ )r@ r r) r) r* r s z2getPhases..InBodyPhase.startTagParamSourcec S sJ | j jdddr | jtd | j j| | j jj d|d< d| j_d S )Nr( rd )r Trx F) r< r r r
r r{ r+ rI r` )r@ r r) r) r* r s z)getPhases..InBodyPhase.startTagHrc S s6 | j jdddd | jtdd|d |d d d S ) Nz unexpected-start-tag-treated-asrr rj ) originalNamenewNamero rs rw )ri rw )rI r| r r
)r@ r r) r) r* r s
z,getPhases..InBodyPhase.startTagImagec S s | j jdddi | jjr d S i }d|d kr>|d d |d< | jtdd|d | jtd d | jtd
d d |d kr|d d }nd }| jtd
|d |d j }d|kr|d= d |kr|d = d|d< | jtdd||d d | j td
| jtd d | j td d S )Nzdeprecated-tagr> rs actionrs rR ro )ri rq Zlabelpromptz3This is a searchable index. Enter search keywords: rn )r2 rs rp rw )ri rw )
rI r| r< r r r
r r copyr )r@ r Z
form_attrsr ri r) r) r* r s6
z.getPhases..InBodyPhase.startTagIsIndexc S s0 | j j| | jjj| jj_| j| _d| j_d S )NF) r< r rI rL rY rZ r r r` )r@ r r) r) r* r s z/getPhases..InBodyPhase.startTagTextareac S s d| j _| j| d S )NF)rI r` r )r@ r r) r) r* r s z-getPhases..InBodyPhase.startTagIFramec S s" | j jr| j| n
| j| d S )N)rI rK r r )r@ r r) r) r* r% s z/getPhases..InBodyPhase.startTagNoscriptc S s | j j|d dS )z8iframe, noembed noframes, noscript(if scripting enabled)r N)rI r )r@ r r) r) r* r s z.getPhases..InBodyPhase.startTagRawtextc S s@ | j jd jdkr$| jjjtd | j j | jj j| d S )Nr rx ry ) r< r{ r> rI r] r r
r r )r@ r r) r) r* r s
z*getPhases..InBodyPhase.startTagOptc S s | j j | j j| d| j_| jj| jjd | jjd | jjd | jjd | jjd | jjd fkrx| jjd | j_n| jjd | j_d S )
NFr r r r r r inSelectInTabler )r< r r rI r` r] rF )r@ r r) r) r* r s
z-getPhases..InBodyPhase.startTagSelectc S sB | j jdr2| j j | j jd jdkr2| jj | j j| d S )Nrubyr ry )r< r generateImpliedEndTagsr{ r> rI r| r )r@ r r) r) r* r s
z+getPhases..InBodyPhase.startTagRpRtc S sZ | j j | jj| | jj| td |d< | j j| |d rV| j jj d|d< d S )Nrd rh rw Trx ) r< r rI r r r r r{ r+ )r@ r r) r) r* r s
z+getPhases..InBodyPhase.startTagMathc S sZ | j j | jj| | jj| td |d< | j j| |d rV| j jj d|d< d S )Nru rh rw Trx ) r< r rI r r r r r{ r+ )r@ r r) r) r* r s
z*getPhases..InBodyPhase.startTagSvgc S s | j jdd|d i dS )a5 Elements that should be children of other elements that have a
different insertion mode; here they are ignored
"caption", "col", "colgroup", "frame", "frameset", "head",
"option", "optgroup", "tbody", "td", "tfoot", "th", "thead",
"tr", "noscript"
z unexpected-start-tag-ignoredr> N)rI r| )r@ r r) r) r* r s z0getPhases..InBodyPhase.startTagMisplacedc S s | j j | j j| d S )N)r< r r )r@ r r) r) r* r s
z,getPhases..InBodyPhase.startTagOtherc S s | j jdddsD| jtdd | jjdddi | jtdd nX| j jd | j jd j dkrt| jjdddi | j jj
}x|j dkr| j jj
}qW d S )
Nr( rd )r ro zunexpected-end-tagr> rp r ry ) r< r r r
rI r| r r r{ r> r+ )r@ r r r) r) r* r s z&getPhases..InBodyPhase.endTagPc S s | j jds| jj d S | j jd jdkrlx>| j jdd D ]*}|jtdkr>| jjdd|jd P q>W | jjd | j_d S )Nr r rT rU rS ry rx r( rv rw r r r r r r r z$expected-one-end-tag-but-got-another)gotName expectedName afterBodyry )rT rU rS ry rx r( rv rw r r r r r r r r ) r< r rI r| r{ r> r rF r] )r@ r r r) r) r* r ! s
z)getPhases..InBodyPhase.endTagBodyc S s" | j jdr | jtd |S d S )Nr )r< r r r
)r@ r r) r) r* r 3 s z)getPhases..InBodyPhase.endTagHtmlc S s |d dkr| j | _| jj|d }|r2| jj | jjd j|d kr^| jjdd|d i |r| jjj }x |j|d kr| jjj }qpW d S )Nr> rP r zend-tag-too-earlyry )
r} r r< r r r{ r> rI r| r+ )r@ r ZinScoper r) r) r* r 9 s
z*getPhases..InBodyPhase.endTagBlockc S sx | j j}d | j _|d ks&| j j| r:| jjdddi n:| j j | j jd |krf| jjdddi | j jj| d S )Nzunexpected-end-tagr> rR r zend-tag-too-early-ignoredry )r< r r rI r| r r{ r9 )r@ r r r) r) r* r G s
z)getPhases..InBodyPhase.endTagFormc S s |d dkrd}nd }| j j|d |dsB| jjdd|d i nj| j j|d d | j jd j|d kr| jjdd|d i | j jj }x |j|d kr| j jj }qW d S )
Nr> rS list)r zunexpected-end-tag)excluder zend-tag-too-earlyry )r< r rI r| r r{ r> r+ )r@ r r r r) r) r* r T s z-getPhases..InBodyPhase.endTagListItemc S s x$t D ] }| jj|r| jj P qW | jjd j|d krR| jjdd|d i xBt D ]:}| jj|rX| jjj }x|jt kr| jjj }qvW P qXW d S )Nr r> zend-tag-too-earlyry ) r r< r r r{ r> rI r| r+ )r@ r itemr) r) r* r e s
z,getPhases..InBodyPhase.endTagHeadingc S s" d}x|dk r |d7 }| j j|d }| sL|| j jkrZ| j j|j rZ| j| dS || j jkr| jjdd|d i | j jj | dS | j j|js| jjdd|d i dS || j jd kr| jjdd|d i | j jj
|}d}x,| j j|d D ]}|j t kr |}P q W |dkrb| j jj
}x ||krN| j jj
}q4W | j jj | dS | j j|d }| j jj
|}| } }
d} | j jj
|
} x| d k rh| d7 } | d8 } | j j| }
|
| j jkr| j jj |
q|
|krP | |kr
| j jj
|
d }|
j }
|
| j j| j jj
|
< |
| j j| j jj
|
< |
}
| jrV| jj| |
j| |
} qW | jr~| jj| |jtdkr| j j \}}|j| | n
|j| |j }
|j|
|j|
| j jj | | j jj||
| j jj | | j jj| j jj
|d |
qW dS )z)The much-feared adoption agency algorithmr r r> Nzadoption-agency-1.2zadoption-agency-4.4zadoption-agency-1.3r r r r r r ry )r r r r r )r< r r{ r r> r rI r| r r9 indexr r r+ Z cloneNoder r Z appendChildr Z getTableMisnestedNodePosition insertBeforeZreparentChildreninsert)r@ r ZouterLoopCounterZformattingElementZafeIndexZ
furthestBlockrk ZcommonAncestorZbookmarkZlastNoder ZinnerLoopCounterr Zcloner r r) r) r* r t s
z/getPhases..InBodyPhase.endTagFormattingc S s | j j|d r| j j | j jd j|d krF| jjdd|d i | j j|d r| j jj }x |j|d kr| j jj }qdW | j j d S )Nr> r zend-tag-too-earlyry ) r< r r r{ r> rI r| r+ clearActiveFormattingElements)r@ r rk r) r) r* r s
z8getPhases..InBodyPhase.endTagAppletMarqueeObjectc S s@ | j jdddd | jj | jjtdd | jjj d S )Nz unexpected-end-tag-treated-asr z
br element)r r ro )rI r| r< r r r
r{ r+ )r@ r r) r) r* r4 # s
z'getPhases..InBodyPhase.endTagBrc S s x| j jd d d D ]}|j|d kr~| j j|d d | j jd j|d krd| jjdd|d i x| j jj |krxqfW P q|jtkr| jjdd|d i P qW d S )Nr r> )r zunexpected-end-tagry ry ) r< r{ r> r rI r| r+ r r )r@ r r r) r) r* r * s
z*getPhases..InBodyPhase.endTagOtherN)6r7 r8 r9 rH r r r r r r} r~ r6 r7 r r r r r r r r r r r r r r r r r r r r r r% r r r r r r r r r r r r r r r r r r4 r r) )r r) r* InBodyPhasee sh G
$ r c s@ e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
S )z getPhases..TextPhasec sF j | || tjg | _| j| j_tjd| jfg| _| j| j_d S )Nr ) rH r r r r r endTagScriptr r )r@ rI r< )r r) r* rH 9 s
z%getPhases..TextPhase.__init__c S s | j j|d d S )Nrs )r< r )r@ r r) r) r* r A s z.getPhases..TextPhase.processCharactersc S s8 | j jdd| jjd ji | jjj | j j| j _dS )Nz&expected-named-closing-tag-but-got-eofr> r Try )rI r| r< r{ r> r+ r r] )r@ r) r) r* r D s
z'getPhases..TextPhase.processEOFc S s dst d|d d S )NFz4Tried to process start tag %s in RCDATA/RAWTEXT moder> )r )r@ r r) r) r* r K s z*getPhases..TextPhase.startTagOtherc S s* | j jj }|jdkst| jj| j_d S )Nr )r< r{ r+ r> r rI r r] )r@ r r r) r) r* r N s z)getPhases..TextPhase.endTagScriptc S s | j jj | jj| j_d S )N)r< r{ r+ rI r r] )r@ r r) r) r* r U s z(getPhases..TextPhase.endTagOtherN) r7 r8 r9 rH r r r r r r) )r r) r* TextPhase8 s r c s e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd Zdd Zdd Zd d Zd d Zd!d" Zd#d$ Zd%d& Zd'S )(z getPhases..InTablePhasec
s j | || tjd| jfd| jfd| jfd| jfd| jfd| jfd | j fd| j
fd| j fd| j fg
| _
| j| j
_tjd | jfd| jfg| _| j| j_d S )Nr r r rz r r r r r r r r r rp rR r )r r r )r r r )r r ) r r rz r r r r r r r r )rH r r r startTagCaptionstartTagColgroup startTagColstartTagRowGroupstartTagImplyTbodyr startTagStyleScriptr r r r r endTagTable endTagIgnorer r )r@ rI r< )r r) r* rH [ s$
z(getPhases..InTablePhase.__init__c S s( x"| j jd jdkr"| j jj qW d S )Nr r r ry )r r )r< r{ r> r+ )r@ r) r) r* clearStackToTableContexts s z8getPhases..InTablePhase.clearStackToTableContextc S s0 | j jd jdkr | jjd n | jjs,td S )Nr r z eof-in-tablery )r< r{ r> rI r| rP r )r@ r) r) r* r | s z*getPhases..InTablePhase.processEOFc S s4 | j j}| j jd | j _|| j j_| j jj| d S )N inTableText)rI r] rF r r )r@ r r r) r) r* r s
z6getPhases..InTablePhase.processSpaceCharactersc S s4 | j j}| j jd | j _|| j j_| j jj| d S )Nr )rI r] rF r r )r@ r r r) r) r* r s
z1getPhases..InTablePhase.processCharactersc S s& d| j _| jjd j| d| j _d S )NTr F)r< insertFromTablerI rF r )r@ r r) r) r* r s z*getPhases..InTablePhase.insertTextc S s6 | j | jjjt | jj| | jjd | j_d S )Nr ) r r< r r r
r rI rF r] )r@ r r) r) r* r s z/getPhases..InTablePhase.startTagCaptionc S s( | j | jj| | jjd | j_d S )Nr )r r< r rI rF r] )r@ r r) r) r* r s z0getPhases..InTablePhase.startTagColgroupc S s | j tdd |S )Nr ro )r r
)r@ r r) r) r* r s z+getPhases..InTablePhase.startTagColc S s( | j | jj| | jjd | j_d S )Nr )r r< r rI rF r] )r@ r r) r) r* r s z0getPhases..InTablePhase.startTagRowGroupc S s | j tdd |S )Nr ro )r r
)r@ r r) r) r* r s z2getPhases..InTablePhase.startTagImplyTbodyc S s6 | j jdddd | j jjtd | j js2|S d S )Nz$unexpected-start-tag-implies-end-tagr )r r )rI r| r] r r
rP )r@ r r) r) r* r s
z-getPhases..InTablePhase.startTagTablec S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z3getPhases..InTablePhase.startTagStyleScriptc S sV d|d krH|d d j tdkrH| jjd | jj| | jjj n
| j| d S )Nr2 rs r z unexpected-hidden-input-in-table) rj r rI r| r< r r{ r+ r )r@ r r) r) r* r s z-getPhases..InTablePhase.startTagInputc S sD | j jd | jjd kr@| jj| | jjd | j_| jjj d S )Nzunexpected-form-in-tabler ry )rI r| r< r r r{ r+ )r@ r r) r) r* r s
z,getPhases..InTablePhase.startTagFormc S s< | j jdd|d i d| j_| j jd j| d| j_d S )Nz)unexpected-start-tag-implies-table-voodoor> Tr F)rI r| r< r rF r )r@ r r) r) r* r s z-getPhases..InTablePhase.startTagOtherc S s | j jdddr| j j | j jd jdkrJ| jjdd| j jd jd x"| j jd jdkrl| j jj qLW | j jj | jj n| jj st
| jj d S ) Nr )r r zend-tag-too-early-named)r r ry ry ry ) r< r r r{ r> rI r| r+ r_ rP r )r@ r r) r) r* r s
z+getPhases..InTablePhase.endTagTablec S s | j jdd|d i d S )Nzunexpected-end-tagr> )rI r| )r@ r r) r) r* r s z,getPhases..InTablePhase.endTagIgnorec S s< | j jdd|d i d| j_| j jd j| d| j_d S )Nz'unexpected-end-tag-implies-table-voodoor> Tr F)rI r| r< r rF r )r@ r r) r) r* r s z+getPhases..InTablePhase.endTagOtherN)r7 r8 r9 rH r r r r r r r r r r r r r r r r r r r) )r r) r* InTablePhaseY s&
r c sP e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dS )z#getPhases..InTableTextPhasec s j | || d | _g | _d S )N)rH r characterTokens)r@ rI r< )r r) r* rH s z,getPhases..InTableTextPhase.__init__c S sd dj dd | jD }tdd |D rJtd |d}| jjd j| n|rZ| jj| g | _d S )Nr c S s g | ] }|d qS )rs r) )r= r r) r) r* rA s zGgetPhases..InTableTextPhase.flushCharacters..c S s g | ] }|t kqS r) )r )r= r r) r) r* rA s rn )r2 rs r )joinr r r rI rF r r< )r@ rs r r) r) r* flushCharacters s z3getPhases..InTableTextPhase.flushCharactersc S s | j | j| j_|S )N)r r rI r] )r@ r r) r) r* r s
z2getPhases..InTableTextPhase.processCommentc S s | j | j| j_dS )NT)r r rI r] )r@ r) r) r* r s
z.getPhases..InTableTextPhase.processEOFc S s |d dkrd S | j j| d S )Nrs r )r r )r@ r r) r) r* r s z5getPhases..InTableTextPhase.processCharactersc S s | j j| d S )N)r r )r@ r r) r) r* r s z:getPhases..InTableTextPhase.processSpaceCharactersc S s | j | j| j_|S )N)r r rI r] )r@ r r) r) r* r
s
z3getPhases..InTableTextPhase.processStartTagc S s | j | j| j_|S )N)r r rI r] )r@ r r) r) r* r s
z1getPhases..InTableTextPhase.processEndTagN) r7 r8 r9 rH r r r r r r r r) )r r) r* InTableTextPhase s r c s` e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dS )z!getPhases..InCaptionPhasec
sf j | || tjd| jfd
| jfg| _| j| j_tjd| jfd | j fd| j
fg| _ | j | j _d S )Nr r rz r r r r r r r r r ) r rz r r r r r r r )
r rz r r r r r r r r )
rH r r r startTagTableElementr r r
endTagCaptionr r r r )r@ rI r< )r r) r* rH s
z*getPhases..InCaptionPhase.__init__c S s | j jddd S )Nr r )r )r< r )r@ r) r) r* ignoreEndTagCaption+ s z5getPhases..InCaptionPhase.ignoreEndTagCaptionc S s | j jd j d S )Nr )rI rF r )r@ r) r) r* r . s z,getPhases..InCaptionPhase.processEOFc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r 1 s z3getPhases..InCaptionPhase.processCharactersc S s0 | j j | j }| j jjtd |s,|S d S )Nr )rI r| r r] r r
)r@ r ignoreEndTagr) r) r* r 4 s
z6getPhases..InCaptionPhase.startTagTableElementc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r < s z/getPhases..InCaptionPhase.startTagOtherc S s | j s| jj | jjd jdkrB| jjdd| jjd jd x"| jjd jdkrd| jjj qDW | jjj | jj | jj d | j_
n| jj st | jj d S ) Nr r z$expected-one-end-tag-but-got-another)r r r ry ry ry )
r r< r r{ r> rI r| r+ r rF r] rP r )r@ r r) r) r* r ? s
z/getPhases..InCaptionPhase.endTagCaptionc S s0 | j j | j }| j jjtd |s,|S d S )Nr )rI r| r r] r r
)r@ r r r) r) r* r Q s
z-getPhases..InCaptionPhase.endTagTablec S s | j jdd|d i d S )Nzunexpected-end-tagr> )rI r| )r@ r r) r) r* r X s z.getPhases..InCaptionPhase.endTagIgnorec S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r [ s z-getPhases..InCaptionPhase.endTagOtherN)
r7 r8 r9 rH r r r r r r r r r r) )r r) r* InCaptionPhase s r c sX e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dS )z%getPhases..InColumnGroupPhasec s^ j | || tjd| jfd| jfg| _| j| j_tjd| jfd| j fg| _
| j | j
_d S )Nr rz r ) rH r r r r r r r endTagColgroup endTagColr r )r@ rI r< )r r) r* rH a s
z.getPhases..InColumnGroupPhase.__init__c S s | j jd jdkS )Nr r ry )r< r{ r> )r@ r) r) r* ignoreEndTagColgroupp s z:getPhases..InColumnGroupPhase.ignoreEndTagColgroupc S sD | j jd jdkr"| jjs td S | j }| jtd |s@dS d S )Nr r r Try ) r< r{ r> rI rP r r r r
)r@ r r) r) r* r s s z0getPhases..InColumnGroupPhase.processEOFc S s" | j }| jtd |s |S d S )Nr )r r r
)r@ r r r) r) r* r } s z7getPhases..InColumnGroupPhase.processCharactersc S s$ | j j| | j jj d|d< d S )NTrx )r< r r{ r+ )r@ r r) r) r* r s z1getPhases..InColumnGroupPhase.startTagColc S s" | j }| jtd |s |S d S )Nr )r r r
)r@ r r r) r) r* r s z3getPhases..InColumnGroupPhase.startTagOtherc S s@ | j r | jjst| jj n | jjj | jjd | j_ d S )Nr )
r rI rP r r| r< r{ r+ rF r] )r@ r r) r) r* r s
z4getPhases..InColumnGroupPhase.endTagColgroupc S s | j jdddi d S )Nz
no-end-tagr> rz )rI r| )r@ r r) r) r* r s z/getPhases..InColumnGroupPhase.endTagColc S s" | j }| jtd |s |S d S )Nr )r r r
)r@ r r r) r) r* r s z1getPhases..InColumnGroupPhase.endTagOtherN) r7 r8 r9 rH r r r r r r r r r) )r r) r* InColumnGroupPhase^ s
r c sx e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd Zdd ZdS ) z#getPhases..InTableBodyPhasec sv j | || tjd| jfd| jfd
| jfd| jfg| _| j| j_ tjd| j
fd | j fd| j fg| _
| j| j
_ d S )Nr r r r r rz r r r r r r )r r )r rz r r r r )r r r )r r rz r r r r r )rH r r r
startTagTrstartTagTableCellstartTagTableOtherr r r endTagTableRowGroupr r r r )r@ rI r< )r r) r* rH s
z,getPhases..InTableBodyPhase.__init__c S sF x"| j jd jdkr"| j jj qW | j jd jdkrB| jjsBtd S ) Nr r r r r ry )r r r r ry )r< r{ r> r+ rI rP r )r@ r) r) r* clearStackToTableBodyContext s
z@getPhases..InTableBodyPhase.clearStackToTableBodyContextc S s | j jd j d S )Nr )rI rF r )r@ r) r) r* r s z.getPhases..InTableBodyPhase.processEOFc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z:getPhases..InTableBodyPhase.processSpaceCharactersc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z5getPhases..InTableBodyPhase.processCharactersc S s( | j | jj| | jjd | j_d S )Nr )r r< r rI rF r] )r@ r r) r) r* r s z.getPhases..InTableBodyPhase.startTagTrc S s* | j jdd|d i | jtdd |S )Nz unexpected-cell-in-table-bodyr> r ro )rI r| r r
)r@ r r) r) r* r s z5getPhases..InTableBodyPhase.startTagTableCellc S sn | j jddds0| j jddds0| j jdddrT| j | jt| j jd j |S | jjs`t | jj
d S )Nr r )r r r r ry ) r< r r r r
r{ r> rI rP r r| )r@ r r) r) r* r s z6getPhases..InTableBodyPhase.startTagTableOtherc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z1getPhases..InTableBodyPhase.startTagOtherc S sT | j j|d ddr:| j | j jj | jjd | j_n| jjdd|d i d S )Nr> r )r r z unexpected-end-tag-in-table-body) r< r r r{ r+ rI rF r] r| )r@ r r) r) r* r s z7getPhases..InTableBodyPhase.endTagTableRowGroupc S sn | j jddds0| j jddds0| j jdddrT| j | jt| j jd j |S | jjs`t | jj
d S )Nr r )r r r r ry ) r< r r r r
r{ r> rI rP r r| )r@ r r) r) r* r s z/getPhases..InTableBodyPhase.endTagTablec S s | j jdd|d i d S )Nz unexpected-end-tag-in-table-bodyr> )rI r| )r@ r r) r) r* r s z0getPhases..InTableBodyPhase.endTagIgnorec S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z/getPhases..InTableBodyPhase.endTagOtherN)r7 r8 r9 rH r r r r r r r r r r r r r) )r r) r* InTableBodyPhase s
r c s e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd Zdd Zdd Zd S ) z getPhases..InRowPhasec sv j | || tjd| jfd
| jfd| jfg| _| j| j_tjd
| j fd | j
fd| j fd| j fg| _
| j| j
_d S )Nr r r r rz r r r r r r r )r r )r rz r r r r r )r r r )r r rz r r r r )rH r r r r r r r r endTagTrr r r r r )r@ rI r< )r r) r* rH s
z&getPhases..InRowPhase.__init__c S sD x>| j jd jdkr>| jjdd| j jd ji | j jj qW d S ) Nr r r z'unexpected-implied-end-tag-in-table-rowr> ry )r r ry )r< r{ r> rI r| r+ )r@ r) r) r* clearStackToTableRowContext s z9getPhases..InRowPhase.clearStackToTableRowContextc S s | j jddd S )Nr r )r )r< r )r@ r) r) r* ignoreEndTagTr s z,getPhases..InRowPhase.ignoreEndTagTrc S s | j jd j d S )Nr )rI rF r )r@ r) r) r* r " s z(getPhases..InRowPhase.processEOFc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r % s z4getPhases..InRowPhase.processSpaceCharactersc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r ( s z/getPhases..InRowPhase.processCharactersc S s6 | j | jj| | jjd | j_| jjjt d S )Nr ) r r< r rI rF r] r r r
)r@ r r) r) r* r + s z/getPhases..InRowPhase.startTagTableCellc S s" | j }| jtd |s |S d S )Nr )r r r
)r@ r r r) r) r* r 1 s z0getPhases..InRowPhase.startTagTableOtherc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r 8 s z+getPhases..InRowPhase.startTagOtherc S sH | j s.| j | jjj | jjd | j_n| jjs:t | jj
d S )Nr ) r r r< r{ r+ rI rF r] rP r r| )r@ r r) r) r* r ; s z&getPhases..InRowPhase.endTagTrc S s" | j }| jtd |s |S d S )Nr )r r r
)r@ r r r) r) r* r E s z)getPhases..InRowPhase.endTagTablec S s4 | j j|d ddr&| jtd |S | jj d S )Nr> r )r r )r< r r r
rI r| )r@ r r) r) r* r M s z1getPhases..InRowPhase.endTagTableRowGroupc S s | j jdd|d i d S )Nz unexpected-end-tag-in-table-rowr> )rI r| )r@ r r) r) r* r T s z*getPhases..InRowPhase.endTagIgnorec S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r X s z)getPhases..InRowPhase.endTagOtherN)r7 r8 r9 rH r r r r r r r r r r r r r r) )r r) r*
InRowPhase s
r c s` e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dS )z getPhases..InCellPhasec sf j | || tjd| jfd
| jfg| _| j| j_tjd| jfd| j fd| j
fg| _ | j | j _d S )Nr r rz r r r r r r r r r ) r rz r r r r r r r )r r )r r rz r r )r r r r r )
rH r r r r r r r endTagTableCellr endTagImplyr r )r@ rI r< )r r) r* rH ] s
z'getPhases..InCellPhase.__init__c S sB | j jdddr | jtd n | j jdddr>| jtd d S )Nr r )r r )r< r r r
)r@ r) r) r* closeCelln s z(getPhases..InCellPhase.closeCellc S s | j jd j d S )Nr )rI rF r )r@ r) r) r* r u s z)getPhases..InCellPhase.processEOFc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r x s z0getPhases..InCellPhase.processCharactersc S sF | j jddds | j jdddr,| j |S | jjs8t| jj d S )Nr r )r r )r< r r rI rP r r| )r@ r r) r) r* r { s z1getPhases..InCellPhase.startTagTableOtherc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z,getPhases..InCellPhase.startTagOtherc S s | j j|d ddr| j j|d | j jd j|d krt| jjdd|d i x.| j jj }|j|d krRP qRW n | j jj | j j | jj d | j_
n| jjdd|d i d S ) Nr> r )r r zunexpected-cell-end-tagr zunexpected-end-tagry ) r< r r r{ r> rI r| r+ r rF r] )r@ r r r) r) r* r s
z.getPhases..InCellPhase.endTagTableCellc S s | j jdd|d i d S )Nzunexpected-end-tagr> )rI r| )r@ r r) r) r* r s z+getPhases..InCellPhase.endTagIgnorec S s. | j j|d ddr | j |S | jj d S )Nr> r )r )r< r r rI r| )r@ r r) r) r* r s z*getPhases..InCellPhase.endTagImplyc S s | j jd j|S )Nr )rI rF r )r@ r r) r) r* r s z*getPhases..InCellPhase.endTagOtherN)
r7 r8 r9 rH r r r r r r r r r r) )r r) r* InCellPhase[ s
r c sx e Zd Z fddZdd Zdd Zdd Zd d
Zd d Zd
d Z dd Z
dd Z dd Z dd Z
dd Zdd ZdS ) z getPhases..InSelectPhasec s j | || tjd| jfd| jfd| jfd| jfd | jfd| jfg| _ | j
| j _ tjd| j fd| j
fd| jfg| _| j| j_ d S )
Nr rx ry r rp rk rt r )rp rk rt )rH r r r startTagOptionstartTagOptgroupr r r& r r r endTagOptionendTagOptgroup endTagSelectr r )r@ rI r< )r r) r* rH s
z)getPhases..InSelectPhase.__init__c S s0 | j jd jdkr | jjd n | jjs,td S )Nr r z
eof-in-selectry )r< r{ r> rI r| rP r )r@ r) r) r* r s z+getPhases..InSelectPhase.processEOFc S s$ |d dkrd S | j j|d d S )Nrs r )r< r )r@ r r) r) r* r s z2getPhases..InSelectPhase.processCharactersc S s. | j jd jdkr | j jj | j j| d S )Nr rx ry )r< r{ r> r+ r )r@ r r) r) r* r s z/getPhases.