Re% @ s d d l m Z m Z m Z d d l m Z d d l Z d d l m Z d d l m Z d d l m
Z
d d l m Z d d l m
Z
d d
l m Z d d l m Z m Z d d l m Z d d
l m Z d d l m Z m Z d d l m Z m Z m Z d d l m Z m Z d d l m Z d d d d d Z d d d d d d Z d d Z Gd d d e Z d d Z! d d d d! d" Z" Gd# d$ d$ e# Z$ d S)% )absolute_importdivisionunicode_literals)with_metaclassN ) inputstream) tokenizer) treebuilders)Marker)utils) constants)spaceCharactersasciiUpper2Lower)specialElements)headingElements)
cdataElementsrcdataElements)
tokenTypesReparseException
namespaces) htmlIntegrationPointElements"mathmlTextIntegrationPointElements)adjustForeignAttributesetreeTc C s4 t j | } t | d | } | j | d | S)z.Parse a string or file-like object into a treenamespaceHTMLElementsencoding)r getTreeBuilder
HTMLParserparse)doc treebuilderr r tbp r# ;/tmp/pip-ztkk0jow-build/pip/_vendor/html5lib/html5parser.pyr s r divc C s: t j | } t | d | } | j | d | d | S)Nr containerr )r r r
parseFragment)r r& r r r r! r" r# r# r$ r' s r' c s G f d d d t } | S)Nc s" e Z d Z f d d Z d S)z-method_decorator_metaclass..Decoratedc s^ xE | j D]7 \ } } t | t j r: | } n | | | .Decorated.__new__N)__name__
__module__ __qualname__r- r# )r4 r# r$ Decorated' s r8 )r, )r4 r8 r# )r4 r$ method_decorator_metaclass& s r9 c @ s e Z d Z d Z d e j d d d d d Z d d d d d d d Z d
d Z d d
Z d d Z
d d Z d d Z d d d d d Z
d d d d d d Z d i d d Z d d Z d d Z d d Z d! d" Z d# d$ Z d% d& Z d' d( Z d S))r zZHTML parser. Generates a tree structure from a stream of (possibly
malformed) HTMLNFTc sz | _ | d k r' t j d } n | | _ | _ g _ t f d d t | j D _ d S)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)
tokenizer - a class that provides a stream of tokens to the treebuilder.
This may be replaced for e.g. a sanitizer which converts some tags to
text
Nr c s. g | ]$ \ } } | | j f q Sr# )tree).0namecls)selfr# r$
M s z'HTMLParser.__init__..)
strictr r r: tokenizer_classerrorsdict getPhasesr( phases)r> r: r r@ r debugr# )r> r$ __init__6 s zHTMLParser.__init__r% c K s | | _ | | _ | j | d | d | d | d | | | _ | j x2 y | j PWqL t k
ry | j YqL XqL Wd S)Nr parseMeta
useChardetparser)
innerHTMLModer& rA r resetmainLoopr )r> stream innerHTMLr& r rH rI kwargsr# r# r$ _parseP s
zHTMLParser._parsec C s | j j d | _ g | _ g | _ d | _ | j r | j j | _ | j t
k rp | j j | j _
nH | j t k r | j j | j _
n$ | j d k r | j j | j _
n | j d | _ | j j | j n d | _ | j d | _ d | _ d | _ d | _ d S)NFz no quirks plaintext
beforeHtmlinitialT)r: rL
firstStartTagrB log
compatModerK r& lowerrO r r rcdataStatestater rawtextStateplaintextStaterE phaseinsertHtmlElementresetInsertionMode lastPhasebeforeRCDataPhase
framesetOK)r> r# r# r$ rL b s*
zHTMLParser.resetc C se | j d k rK | j t d k rK d | j k oJ | j d j t d k S| j | j f t k Sd S)Nzannotation-xmlmathmlr text/htmlapplication/xhtml+xml)rd re )r< namespacer
attributes translater r )r> elementr# r# r$ isHTMLIntegrationPoint s
z!HTMLParser.isHTMLIntegrationPointc C s | j | j f t k S)N)rf r< r )r> ri r# r# r$ isMathMLTextIntegrationPoint s z'HTMLParser.isMathMLTextIntegrationPointc C s* t d } t d } t d } t d } t d } t d } t d } x| j D]y} | } x+| d k r| j j r | j j d n d }
|
r |
j n d } |
r |
j n d } | d }
|
| k r | j | d
| j d i d } qb t | j j d k s| | j j k s| j
|
rl|
| k rZ| d
t d d g k s|
| | f k s| t d k r| d k r| d
d k s| j
|
r|
| | | f k r| j } n
| j d } |
| k r| j | } qb |
| k r| j | } qb |
| k r/| j | } qb |
| k rM| j | } qb |
| k rk| j | } qb |
| k rb | j | } qb qb W|
| k rS | d rS | d rS | j d i | d
d
6 qS qS Wd } g } xG | r%| j | j | j j } | r| j | k s"t qqWd S)N
CharactersSpaceCharactersStartTagEndTagCommentDoctype
ParseErrorr r, datadatavarsr r< mglyph
malignmarkrc zannotation-xmlsvginForeignContent selfClosingselfClosingAcknowledgedz&non-void-element-with-trailing-solidusT)r normalizedTokensr: openElementsrf r<
parseErrorgetlendefaultNamespacerk frozensetr rj r] rE processCharactersprocessSpaceCharactersprocessStartTag
processEndTagprocessCommentprocessDoctypeappend
processEOFAssertionError)r> CharactersTokenSpaceCharactersToken
StartTagToken EndTagToken CommentToken DoctypeTokenParseErrorTokentoken new_token currentNodecurrentNodeNamespacecurrentNodeNamer, r] reprocessrE r# r# r$ rM sj
"
zHTMLParser.mainLoopc c s& x | j D] } | j | Vq
Wd S)N)r normalizeToken)r> r r# r# r$ r| s zHTMLParser.normalizedTokensc
C s2 | j | d d d | d | d | | j j S)af 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)
rO Fr rH rI )rQ r: getDocument)r> rN r rH rI r# r# r$ r s
zHTMLParser.parsec C s) | j | d d | d | | j j S)a 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)
Tr& r )rQ r: getFragment)r> rN r& r rH rI r# r# r$ r' s zHTMLParser.parseFragmentzXXX-undefined-errorc C s; | j j | j j j | | f | j r7 t n d S)N)rB r r rN positionr@ rr )r> errorcodert r# r# r$ r~ s % zHTMLParser.parseErrorc C s< | d t d k r8 t | d d d d | d r r# r# r$ r s $zHTMLParser.normalizeTokenc C sb i d d 6} xN | j D]@ \ } } | | d k r | d | | d | <| d | =q q Wd S)N
definitionURL
definitionurlrs )r( )r> r replacementskvr# r# r$ adjustMathMLAttributes s
z!HTMLParser.adjustMathMLAttributesc C s i> d d 6d d 6d d 6d d 6d d
6d d 6d
d 6d d 6d d 6d d 6d d 6d d 6d d 6d d 6d d 6d d 6d! d" 6d# d$ 6d% d& 6d' d( 6d) d* 6d+ d, 6d- d. 6d/ d0 6d1 d2 6d3 d4 6d5 d6 6d7 d8 6d9 d: 6d; d< 6d= d> 6d? d@ 6dA dB 6dC dD 6dE dF 6dG dH 6dI dJ 6dK dL 6dM dN 6dO dP 6dQ dR 6dS dT 6dU dV 6dW dX 6dY dZ 6d[ d\ 6d] d^ 6d_ d` 6da db 6dc dd 6de df 6dg dh 6di dj 6dk dl 6dm dn 6do dp 6dq dr 6ds dt 6du dv 6dw dx 6dy dz 6d{ d| 6} xX t | d} j D]@ } | | k r| | } | d} | | d} | <| d} | =qqWd S)~Nr2
attributename
attributeType
attributetype
baseFrequency
basefrequency baseProfile baseprofilecalcModecalcmode
clipPathUnits
clippathunitscontentScriptTypecontentscripttypecontentStyleTypecontentstyletypediffuseConstantdiffuseconstantedgeModeedgemodeexternalResourcesRequiredexternalresourcesrequired filterRes filterres filterUnits filterunitsglyphRefglyphrefgradientTransformgradienttransform
gradientUnits
gradientunits kernelMatrix kernelmatrixkernelUnitLengthkernelunitlength keyPoints keypoints
keySplines
keysplineskeyTimeskeytimes lengthAdjust lengthadjustlimitingConeAnglelimitingconeangle markerHeight markerheight markerUnits markerunits markerWidth markerwidthmaskContentUnitsmaskcontentunits maskUnits maskunits
numOctaves
numoctaves
pathLength
pathlengthpatternContentUnitspatterncontentunitspatternTransformpatterntransform patternUnits patternunits pointsAtX pointsatx pointsAtY pointsaty pointsAtZ pointsatz
preserveAlpha
preservealphapreserveAspectRatiopreserveaspectratioprimitiveUnitsprimitiveunitsrefXrefxrefYrefy repeatCount repeatcount repeatDur repeatdurrequiredExtensionsrequiredextensionsrequiredFeaturesrequiredfeaturesspecularConstantspecularconstantspecularExponentspecularexponent spreadMethod spreadmethod startOffset startoffset stdDeviation stddeviation stitchTiles stitchtiles surfaceScale surfacescalesystemLanguagesystemlanguage tableValues tablevaluestargetXtargetxtargetYtargety
textLength
textlengthviewBoxviewbox
viewTarget
viewtargetxChannelSelectorxchannelselectoryChannelSelectorychannelselector
zoomAndPan
zoomandpanrs )listkeys)r> r r originalNamesvgNamer# r# r$ adjustSVGAttributes s
z HTMLParser.adjustSVGAttributesc C s_ t } xR | d j D]@ } | | k r | | } | d | | d | <| d | =q q Wd S)Nrs )adjustForeignAttributesMapr )r> r r r! foreignNamer# r# r$ r P s
z"HTMLParser.adjustForeignAttributesc C s | j j d S)N)rJ r] )r> r r# r# r$ reparseTokenNormalY s z HTMLParser.reparseTokenNormalc C s] d } i d d 6d d 6d d 6d d 6d d
6d d 6d d 6d
d 6d d 6d d 6d d 6d d 6d d 6d d 6} x | j j d d d D] } | j } d } | | j j d k r | j s t d } | j } n | d k r | j s t n | r| j | j j k rq n | | k r5| j | | } Pq | r | j d } Pq q W| | _ d S) NFinSelectselectinCelltdthinRowtr inTableBodytbodytheadtfoot inCaptioncaption
inColumnGroupcolgroupinTabletableinBodyheadbody
inFramesetframeset
beforeHeadhtmlr r Tr{ )zselectzcolgroupzheadzhtml) r: r} r< rO r rf r rE r] )r> lastnewModesnodenodeName new_phaser# r# r$ r_ \ sD
z HTMLParser.resetInsertionModec C su | d k s t | j j | | d k rC | j j | j _ n | j j | j _ | j | _ | j d | _ d S)zYGeneric RCDATA/RAWTEXT Parsing algorithm
contentType - RCDATA or RAWTEXT
RAWTEXTRCDATAtextN)zRAWTEXTzRCDATA)
r r:
insertElementr r[ rZ rY r]
originalPhaserE )r> r contentTyper# r# r$ parseRCDataRawtext s z HTMLParser.parseRCDataRawtext)r5 r6 r7 __doc__r
HTMLTokenizerrG rQ rL rj rk rM r| r r' r~ r r r# r r& r_ rJ r# r# r# r$ r 2 s* "
@ G ,r c s7 d d } d d } Gd d d t | | | Gd d d } Gd d
d
} G f d d d } G f d
d d } G f d d d } G f d d d } G f d d d } G f d d d }
G f d d d } G f d d d } G f d d d }
G f d d d } G f d d d } G f d! d" d" } G f d# d$ d$ } G f d% d& d& } G f d' d( d( } G f d) d* d* } G f d+ d, d, } G f d- d. d. } G f d/ d0 d0 } G f d1 d2 d2 } i | d3 6| d4 6| d5 6| d6 6| d7 6| d8 6| d9 6|
d: 6| d; 6| d< 6|
d= 6| d> 6| d? 6| d@ 6| dA 6| dB 6| dC 6| dD 6| dE 6| dF 6| dG 6| dH 6S)INc s; t d d t j j D f d d } | S)z4Logger that records which phase processes each tokenc s s! | ] \ } } | | f Vq d S)Nr# )r; keyvaluer# r# r$ s z)getPhases..log..c s j j d r t | d k r | d } y i | d d 6} Wn Yn X| d t j k ry | d | d argsrP r info)r4
type_namesr# r$ wrapped s $
z'getPhases..log..wrapped)rC r r r( )r4 rW r# )r4 rV r$ rV s zgetPhases..logc S s | r t | St Sd S)N)r9 r, )
use_metaclassmetaclass_funcr# r# r$ getMetaclass s
z getPhases..getMetaclassc @ s e Z d Z d Z d d Z d d Z d d Z d d Z d
d Z d d
Z d d Z
d d Z d d Z d S)zgetPhases..PhasezNBase class for helper object that implements each phase of processing
c S s | | _ | | _ d S)N)rJ r: )r> rJ r: r# r# r$ rG 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 j d d S)Nr r{ )r:
insertCommentr} )r> r r# r# r$ r s z'getPhases..Phase.processCommentc S s | j j d d S)Nzunexpected-doctype)rJ 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 s | j j r0 | d d k r0 | j j d n xQ | d j D]? \ } } | | j j d j k rA | | j j d j | z
non-html-rootrs r F)rJ rU r~ r( r: r} rg )r> r attrrN 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)
r5 r6 r7 rK rG r r r r r r r` r r# r# r# r$ Phase s
rb c @ sp e Z d Z d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d S)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 su | d } | d } | d } | d } | d k sX | d k sX | d k rk | d k rk | j j d n | d k r d } n | j j | | d k r | j t } n | s| d d k s| j dJ s| dK k s| j dL r | d k s| r"| j dD k r"dE | j _ n9 | j dM sL| j dN r[| d k r[dH | j _ n | 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 )7rh ri rj rk rl rm rn ro rp rq rr rs rt ru rv rw rx ry rz 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 zhtml)r r )r r )r r ) rJ r~ r:
insertDoctyperh r rQ rX rW rE r] )r> r r< rd re rf r# r# r$ r s
z.getPhases..InitialPhase.processDoctypec S s& d | j _ | j j d | j _ d S)Nr rS )rJ rW rE r] )r> r# r# r$ anythingElseJ s z,getPhases..InitialPhase.anythingElsec S s | j j d | j | S)Nz expected-doctype-but-got-chars)rJ r~ r )r> r r# r# r$ r N s
z1getPhases..InitialPhase.processCharactersc S s, | j j d i | d d 6 | j | S)Nz"expected-doctype-but-got-start-tagr< )rJ r~ r )r> r r# r# r$ r S s
z/getPhases..InitialPhase.processStartTagc S s, | j j d i | d d 6 | j | S)Nz expected-doctype-but-got-end-tagr< )rJ r~ r )r> r r# r# r$ r Y s
z-getPhases..InitialPhase.processEndTagc S s | j j d | j d S)Nz expected-doctype-but-got-eofT)rJ r~ r )r> r# r# r$ r _ s
z*getPhases..InitialPhase.processEOFN) r5 r6 r7 r r r r r r r r r# r# r# r$ InitialPhase s ` r c @ sd e Z d Z d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d S)z"getPhases..BeforeHtmlPhasec S s3 | j j t d d | j j d | j _ d S)Nr> rn r= )r:
insertRootimpliedTagTokenrJ rE r] )r> r# r# r$ r^ f s z4getPhases..BeforeHtmlPhase.insertHtmlElementc S s | j d S)NT)r^ )r> r# r# r$ r k s
z-getPhases..BeforeHtmlPhase.processEOFc S s | j j | | j j d S)N)r: r\ rc )r> r r# r# r$ r o s z1getPhases..BeforeHtmlPhase.processCommentc S s d S)Nr# )r> r r# r# r$ r r s z9getPhases..BeforeHtmlPhase.processSpaceCharactersc S s | j | S)N)r^ )r> r r# r# r$ r u s
z4getPhases..BeforeHtmlPhase.processCharactersc S s- | d d k r d | j _ n | j | S)Nr< r> T)rJ rU r^ )r> r r# r# r$ r y s
z2getPhases..BeforeHtmlPhase.processStartTagc S sC | d d k r1 | j j d i | d d 6 n | j | Sd S)Nr< r9 r: r> brz unexpected-end-tag-before-html)zheadzbodyzhtmlzbr)rJ r~ r^ )r> r r# r# r$ r s
z0getPhases..BeforeHtmlPhase.processEndTagN)
r5 r6 r7 r^ r r r r r r r# r# r# r$ BeforeHtmlPhased s r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d S)z"getPhases..BeforeHeadPhasec s} j | | | t j d | j f d | j f g | _ | j | j _ t j d | j f g | _ | j
| j _ d S)Nr> r9 r: r )zheadzbodyzhtmlzbr) rG r MethodDispatcherr` startTagHeadr^
startTagOtherdefaultendTagImplyHeadra endTagOther)r> rJ r: )rb r# r$ rG s z+getPhases..BeforeHeadPhase.__init__c S s | j t d d d S)Nr9 rn 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 t d d | S)Nr9 rn )r r )r> r r# r# r$ r s z4getPhases..BeforeHeadPhase.processCharactersc S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r` s z/getPhases..BeforeHeadPhase.startTagHtmlc S s@ | j j | | j j d | j _ | j j d | j _ d S)Nr inHeadr{ )r: rG r} headPointerrJ rE r] )r> r r# r# r$ r s z/getPhases..BeforeHeadPhase.startTagHeadc S s | j t d d | S)Nr9 rn )r r )r> r r# r# r$ r s z0getPhases..BeforeHeadPhase.startTagOtherc S s | j t d d | S)Nr9 rn )r r )r> r r# r# r$ r s z2getPhases..BeforeHeadPhase.endTagImplyHeadc S s" | j j d i | d d 6 d S)Nzend-tag-after-implied-rootr< )rJ r~ )r> r r# r# r$ r s z.getPhases..BeforeHeadPhase.endTagOtherN) r5 r6 r7 rG r r r r` r r r r r# )rb r# r$ BeforeHeadPhase s r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d d Z
d d Z d d Z d d Z d d Z d S) z getPhases..InHeadPhasec
s j | | | t j d | j f d | j f d | j f d | j f d | j f d | j f d
| j f g | _
| j | j
_ t j d
| j
f d | j f g | _ | j | j _ d S)Nr> titlenoscriptnoframesstylescriptbasebasefontbgsoundcommandlinkr. r9 r r: )znoscriptznoframeszstyle)zbasezbasefontzbgsoundzcommandzlink)zbrzhtmlzbody)rG r r r`
startTagTitle startTagNoScriptNoFramesStylestartTagScriptstartTagBaseLinkCommand startTagMetar r^ r r
endTagHeadendTagHtmlBodyBrra r )r> rJ r: )rb r# r$ rG s z'getPhases..InHeadPhase.__init__c S s | j d S)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 j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r` s z+getPhases..InHeadPhase.startTagHtmlc S s | j j d d S)Nz!two-heads-are-not-better-than-one)rJ r~ )r> r r# r# r$ r s z+getPhases..InHeadPhase.startTagHeadc S s. | j j | | j j j d | d r r# r# r$ r s z6getPhases..InHeadPhase.startTagBaseLinkCommandc S s | j j | | j j j d | d <| d } | j j j j d d k r d | k ry | j j j j | d q d | k r d | k r | d j d k r t
j | d j d
} t
j
| } | j } | j j j j | q n d S) NTrz rs r tentativecharsetcontentz
http-equivz content-typezutf-8)r: rG r} r rJ r rN charEncodingchangeEncodingrX r
EncodingBytesencodeContentAttrParserr )r> r rg rs rJ codecr# r# r$ r s
z+getPhases..InHeadPhase.startTagMetac S s | j j | d d S)NrE )rJ rJ )r> r r# r# r$ r s z,getPhases..InHeadPhase.startTagTitlec S s | j j | d d S)NrD )rJ rJ )r> r r# r# r$ r s z.InHeadPhase.startTagNoScriptNoFramesStylec S sT | j j | | j j j | j j _ | j j | j _ | j j d | j _ d S)NrF ) r: rG rJ r scriptDataStaterZ r] rH rE )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 sQ | j j j j } | j d k s7 t d | j | j j d | j _ d S)Nr9 zExpected head got %s afterHead)rJ r: r} r r< r rE r] )r> r rA 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 j d i | d d 6 d S)Nzunexpected-end-tagr< )rJ r~ )r> r r# r# r$ r s z*getPhases..InHeadPhase.endTagOtherc S s | j t d d S)Nr9 )r r )r> r# r# r$ r s z+getPhases..InHeadPhase.anythingElseN)r5 r6 r7 rG r r r` r r r r r r r r r r r r# )rb r# r$ InHeadPhase s r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d d Z
d d Z d S)z!getPhases..AfterHeadPhasec
s j | | | t j d | j f d | j f d | j f d | j f d
| j f g | _ | j | j _
t j d | j f g | _ | j
| j _
d S)Nr> r: r< r r r r r. r r r r r9 r ) zbasezbasefontzbgsoundzlinkzmetaznoframeszscriptzstyleztitle)zbodyzhtmlzbr)rG r r r` startTagBodystartTagFramesetstartTagFromHeadr r^ r r r ra r )r> rJ r: )rb r# r$ rG s z*getPhases..AfterHeadPhase.__init__c S s | j d S)NT)r )r> r# r# r$ r * s
z,getPhases..AfterHeadPhase.processEOFc S s | j | S)N)r )r> r r# r# r$ r . s
z3getPhases..AfterHeadPhase.processCharactersc S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r` 2 s z.getPhases..AfterHeadPhase.startTagHtmlc S s6 d | j _ | j j | | j j d | j _ d S)NFr8 )rJ rb r: rG rE r] )r> r r# r# r$ r 5 s z.getPhases..AfterHeadPhase.startTagBodyc S s* | j j | | j j d | j _ d S)Nr; )r: rG rJ rE r] )r> r r# r# r$ r : s z2getPhases..AfterHeadPhase.startTagFramesetc S s | j j d i | d d 6 | j j j | j j | j j d j | xG | j j d d d D], } | j d k rh | j j j | Pqh qh Wd S)Nz#unexpected-start-tag-out-of-my-headr< r r r9 r{ )
rJ r~ r: r} r r rE r r< remove)r> r rA r# r# r$ r > s z2getPhases..AfterHeadPhase.startTagFromHeadc S s" | j j d i | d d 6 d S)Nzunexpected-start-tagr< )rJ r~ )r> r r# r# r$ r H s z.getPhases..AfterHeadPhase.startTagHeadc S s | j | S)N)r )r> r r# r# r$ r K s
z/getPhases..AfterHeadPhase.startTagOtherc S s | j | S)N)r )r> r r# r# r$ r O s
z2getPhases..AfterHeadPhase.endTagHtmlBodyBrc S s" | j j d i | d d 6 d S)Nzunexpected-end-tagr< )rJ r~ )r> r r# r# r$ r S s z-getPhases..AfterHeadPhase.endTagOtherc S s? | j j t d d | j j d | j _ d | j _ d S)Nr: rn r8 T)r: rG r rJ rE r] rb )r> r# r# r$ r V s z.getPhases..AfterHeadPhase.anythingElseN)r5 r6 r7 rG r r r` r r r r r r r r r# )rb r# r$ AfterHeadPhase s
r c sn e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d d Z
d d Z d d Z d d Z d d Z d d Z d! d" Z d# d$ Z d% d& Z d' d( Z d) d* Z d+ d, Z d- d. Z d/ d0 Z d1 d2 Z d3 d4 Z d5 d6 Z d7 d8 Z d9 d: Z d; d< Z d= d> Z! d? d@ Z" dA dB Z# dC dD Z$ dE dF Z% dG dH Z& dI dJ Z' dK dL Z( dM dN Z) dO dP Z* dQ dR Z+ dS dT Z, dU dV Z- dW dX Z. dY dZ Z/ d[ d\ Z0 d] d^ Z1 d_ d` Z2 da db Z3 dc dd Z4 de S)fz getPhases..InBodyPhasec + s] j | | | | j | _ t j d | j f dd | j f d | j f d
| j f de | j f t
| j f df | j f d' | j
f dg | j f d+ | j f d, | j f dh | j f d9 | j f d: | j f di | j f d> | j f d? | j f dj | j f dk | j f dI | j f dJ | j f dK | j f dL | j f dM | j f dN | j f dl | j f dQ | j f dm | j! f dn | j" f dV | j# f dW | j$ f do | j% f g | _& | j' | j& _( t j d | j) f d | j* f dp | j+ f d' | j, f d! | j- f dq | j. f t
| j/ f dr | j0 f ds | j1 f dA | j2 f g
| _3 | j4 | j3 _( d S)tNr> r r r r r r. r r r r r: r< addressarticleaside
blockquotecenterdetailsdirr% dlfieldset
figcaptionfigurefooterheaderhgroupmainmenunavolr" sectionsummaryulprelistingformlidddtrR abbigcodeemfontissmallstrikestrongttunobrbuttonappletmarqueeobjectxmpr7 arear embedimgkeygenwbrparamsourcetrackinputhrimageisindextextareaiframenoembedr r( rprtoptionoptgroupmathrw r3 colr5 framer9 r/ r* r1 r+ r0 r- dialog)
zbasezbasefontzbgsoundzcommandzlinkzmetanoframeszscriptzstyleztitle)addressarticleaside
blockquotecenterdetailsr$ dirdivdlfieldset
figcaptionfigurefooterheaderhgroupmainmenunavolr" sectionsummaryul)prelisting)lidddt) r bigcodeemfontr r smallstrikestrongttr )appletmarqueeobject)zareazbrzembedzimgzkeygenzwbr)zparamzsourceztrack)znoembedr znoscript)zrpzrt)zoptionzoptgroup) zcaptionzcolzcolgroupzframezheadztbodyztdztfootzthztheadztr)r r r! r" zbuttonr# r$ zdialogr% r& r' r( r) r* r+ r, r- r6 r. r/ r0 r1 r5 r2 r3 r4 )r8 r9 r7 )r r r: r; r< r= r znobrr r> r? r@ rA r )rB rC rD )5rG r processSpaceCharactersNonPrer r r` startTagProcessInHeadr r startTagClosePr startTagHeadingstartTagPreListing startTagFormstartTagListItemstartTagPlaintext startTagAstartTagFormatting startTagNobrstartTagButtonstartTagAppletMarqueeObject startTagXmp
startTagTablestartTagVoidFormattingstartTagParamSource
startTagInput
startTagHr
startTagImagestartTagIsIndexstartTagTextareastartTagIFramestartTagRawtextstartTagSelect startTagRpRt startTagOpt startTagMath startTagSvgstartTagMisplacedr^ r r
endTagBody
endTagHtml endTagBlock
endTagFormendTagPendTagListItem
endTagHeadingendTagFormattingendTagAppletMarqueeObjectendTagBrra r )r> rJ r: )rb r# r$ rG ^ s| z'getPhases..InBodyPhase.__init__c S s | j | j k s$ | j | j k r( d St | j t | j k rJ d St | j j } t | j j } x- t | | D] \ } } | | k r d Sq Wd S)NFT)r< rf r rg sortedr( zip)r> node1node2 attributes1 attributes2attr1attr2r# r# r$ isMatchingFormattingElement s $ z:getPhases..InBodyPhase.isMatchingFormattingElementc S s | j j | | j j d } g } xS | j j d d d D]8 } | t k rV Pq@ | j | | r@ | j | q@ q@ Wt | d k s t t | d k r | j j j | d n | j j j | d S)Nr r{ r{ r{ )
r: rG r} activeFormattingElementsr
ru r r r r )r> r ri matchingElementsrA r# r# r$ addFormattingElement s z3getPhases..InBodyPhase.addFormattingElementc
S sW t d } xD | j j d d d D]) } | j | k r& | j j d Pq& q& Wd S)Nr r r r" r/ r* r1 r+ r0 r- r: r> r z expected-closing-tag-but-got-eof) zddzdtzlir" ztbodyztdztfootzthztheadztrzbodyzhtmlr{ )r r: r} r< rJ r~ )r> allowed_elementsrA r# r# r$ r s z)getPhases..InBodyPhase.processEOFc S s | d } | j | _ | j d rh | j j d j d k rh | j j d j rh | d d } n | r | j j | j j | n d S)
Nrs
r r r r r{ )zprezlistingztextarear{ ) rE r rQ r: r} r<
hasContent#reconstructActiveFormattingElementsr] )r> r rs r# r# r$ !processSpaceCharactersDropNewline s
z@getPhases..InBodyPhase.processSpaceCharactersDropNewlinec S sq | d d k r d S| j j | j j | d | j j rm t d d | d D rm d | j _ n d S)Nrs c S s g | ] } | t k q Sr# )r
)r; charr# r# r$ r? s zDgetPhases..InBodyPhase.processCharacters..F)r: r} r] rJ rb 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
z5getPhases..InBodyPhase.processSpaceCharactersc S s | j j d j | S)Nr )rJ rE r )r> r r# r# r$ rF s z4getPhases..InBodyPhase.startTagProcessInHeadc S s | j j d i d d 6 t | j j d k sK | j j d j d k r` | j j s t n` d | j _ xQ | d j D]? \ } } | | j j d j
k r} | | j j d j
| r r_ rN r# r# r$ r s z+getPhases..InBodyPhase.startTagBodyc S s | j j d i d d 6 t | j j d k sK | j j d j d k r` | j j s t n | j j so n | j j d j r | j j d j j
| j j d n x- | j j d j d k r | j j j q W| j j | | j j
d | j _ d S) Nzunexpected-start-tagr< r< r r: r> r; r{ )rJ r~ r r: r} r< rO r rb parent removeChildr rG rE r] )r> r r# r# r$ r s 1 ' z/getPhases..InBodyPhase.startTagFramesetc S sB | j j d d d r. | j t d n | j j | d S)Nr" variantr )r: elementInScoperg r rG )r> r r# r# r$ rG s z-getPhases..InBodyPhase.startTagClosePc S sZ | j j d d d r. | j t d n | j j | d | j _ | j | _ d S)Nr" r r F) r: r rg r rG rJ rb r~ r )r> r r# r# r$ rI s
z1getPhases..InBodyPhase.startTagPreListingc S s | j j r) | j j d i d d 6 nT | j j d d d rW | j t d n | j j | | j j d | j _ d S) Nzunexpected-start-tagr r< r" r r r r{ ) r: formPointerrJ r~ r rg r rG r} )r> r r# r# r$ rJ s z+getPhases..InBodyPhase.startTagFormc S s d | j _ i d g d 6d d g d 6d d g d 6} | | d } xn t | j j D]Z } | j | k r | j j j t | j d Pn | j t
k rW | j d k rW PqW qW W| j j d d
d r | j j j t d d n | j j | d S)
NFr r r r< ro r r% r" r r )zaddresszdivr" )
rJ rb reversedr: r} r< r] r r nameTupler r rG )r> r stopNamesMap stopNamesrA r# r# r$ rK s"
z/getPhases..InBodyPhase.startTagListItemc S sZ | j j d d d r. | j t d n | j j | | j j j | j j _ d S)Nr" r r ) r: r rg r rG rJ r r\ rZ )r> r r# r# r$ rL 1 s z0getPhases..InBodyPhase.startTagPlaintextc S s | j j d d d r. | j t d n | j j d j t k rx | j j d i | d d 6 | j j j n | j j
| d S)Nr" r r r zunexpected-start-tagr< r{ ) r: r rg r r} r< r rJ r~ r rG )r> r r# r# r$ rH 7 s z.getPhases..InBodyPhase.startTagHeadingc S s | j j d } | r | j j d i d d 6d d 6 | j t d | | j j k rt | j j j | n | | j j k r | j j j | q n | j j | j
| d S)Nr z$unexpected-start-tag-implies-end-tag startNameendName) r: !elementInActiveFormattingElementsrJ r~ rj r r} r rw r} ry )r> r afeAElementr# r# r$ rM ? s
z(getPhases..InBodyPhase.startTagAc S s | j j | j | d S)N)r: r} ry )r> r r# r# r$ rN L s
z1getPhases..InBodyPhase.startTagFormattingc S st | j j | j j d rc | j j d i d d 6d d 6 | j t d | j j n | j | d S)Nr z$unexpected-start-tag-implies-end-tagr r )r: r} r rJ r~ r r ry )r> r r# r# r$ rO P s
z+getPhases..InBodyPhase.startTagNobrc S sw | j j d rJ | j j d i d d 6d d 6 | j t d | S| j j | j j | d | j _ d S)Nr z$unexpected-start-tag-implies-end-tagr r F) r: r rJ r~ r r r} rG rb )r> r r# r# r$ rP Z s
z-getPhases..InBodyPhase.startTagButtonc S s@ | j j | j j | | j j j t d | j _ d S)NF)r: r} rG rw r r
rJ rb )r> r r# r# r$ rQ e s
z:getPhases..InBodyPhase.startTagAppletMarqueeObjectc S s^ | j j d d d r. | j t d n | j j d | j _ | j j | d d S)Nr" r r FrD )r: r rg r r} rJ rb rJ )r> r r# r# r$ rR k s
z*getPhases..InBodyPhase.startTagXmpc S sy | j j d k rC | j j d d d rC | j t d qC n | j j | d | j _ | j j d | j _ d S)Nr r" r r Fr6 )
rJ rW r: r r r rG rb rE r] )r> r r# r# r$ rS r s z,getPhases..InBodyPhase.startTagTablec S sG | j j | j j | | j j j d | d r r# r# r$ rT z s
z5getPhases..InBodyPhase.startTagVoidFormattingc S sY | j j } | j | d | d k rU | d d j t d k rU | | j _ n d S)Nr, rs hidden)rJ rb rT rh r )r> r rb r# r# r$ rV s
z,getPhases..InBodyPhase.startTagInputc S s. | j j | | j j j d | d r r# r# r$ rU s z2getPhases..InBodyPhase.startTagParamSourcec S sh | j j d d d r. | j t d n | j j | | j j j d | d r r# r# r$ rW s
z)getPhases..InBodyPhase.startTagHrc S sO | j j d i d d 6d d 6 | j t d d d | d d | d d S)
Nz unexpected-start-tag-treated-asr r! r newNamern rg rs ry )rJ r~ r r )r> r r# r# r$ rX s
z,getPhases..InBodyPhase.startTagImagec S s | j j d i d d 6 | j j r* d Si } d | d k rU | d d | d r
form_attrsr rg r# r# r$ rY s6
z.getPhases..InBodyPhase.startTagIsIndexc S sD | j j | | j j j | j j _ | j | _ d | j _ d S)NF) r: rG rJ r rY rZ r~ r rb )r> r r# r# r$ rZ s z/getPhases..InBodyPhase.startTagTextareac S s d | j _ | j | d S)NF)rJ rb r\ )r> r r# r# r$ r[ s z-getPhases..InBodyPhase.startTagIFramec S s | j j | d d S)z8iframe, noembed noframes, noscript(if scripting enabled)rD N)rJ rJ )r> r r# r# r$ r\ s z.getPhases..InBodyPhase.startTagRawtextc S sY | j j d j d k r5 | j j j t d n | j j | j j j | d S)Nr r r{ ) r: r} r< rJ r] r r r} rG )r> r r# r# r$ r_ s
z*getPhases..InBodyPhase.startTagOptc S s | j j | j j | d | j _ | j j | j j d | j j d | j j d | j j d | j j d | j j d f k r | j j d | j _ n | j j d | j _ d S)
NFr6 r2 r4 r. r, r) inSelectInTabler' )r: r} rG rJ rb r] rE )r> r r# r# r$ r] s
z-getPhases..InBodyPhase.startTagSelectc S s_ | j j d rK | j j | j j d j d k rK | j j qK n | j j | d S)Nrubyr r{ )r: r generateImpliedEndTagsr} r< rJ r~ rG )r> r r# r# r$ r^ s
z+getPhases..InBodyPhase.startTagRpRtc S sv | j j | j j | | j j | t d | d <| j j | | d rr | j j j d | d r r# r# r$ r` s
z+getPhases..InBodyPhase.startTagMathc S sv | j j | j j | | j j | t d | d <| j j | | d rr | j j j d | d r r# r# r$ ra s
z*getPhases..InBodyPhase.startTagSvgc S s" | j j d i | d d 6 d S)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)rJ r~ )r> r r# r# r$ rb s z0getPhases..InBodyPhase.startTagMisplacedc S s! | j j | j j | d S)N)r: r} rG )r> r r# r# r$ r s
z,getPhases..InBodyPhase.startTagOtherc S s | j j d d d sa | j t d d | j j d i d d 6 | j t d d n | j j d | j j d j d k r | j j d i d d 6 n | j j j
} x% | j d k r | j j j
} q Wd S)
Nr" r r rn zunexpected-end-tagr< ro r r{ ) r: r rG r rJ r~ rg r r} r< r )r> r rA r# r# r$ rg s z&getPhases..InBodyPhase.endTagPc S s | j j d s# | j j d S| j j d j d k r x^ | j j d d D]C } | j t d k rS | j j d i d d 6| j d 6 PqS qS Wn | j j d | j _ d S)Nr: r r r r r r r" r r r/ r* r1 r+ r0 r- r> z$expected-one-end-tag-but-got-another expectedNamegotName afterBodyr{ )zddzdtzlizoptgroupzoptionr" zrpzrtztbodyztdztfootzthztheadztrzbodyzhtml) r: r rJ r~ r} r< r rE r] )r> r rA r# r# r$ rc s
z)getPhases..InBodyPhase.endTagBodyc S s- | j j d r) | j t d | Sd S)Nr: )r: r rc r )r> r r# r# r$ rd + s z)getPhases..InBodyPhase.endTagHtmlc S s | d d k r | j | _ n | j j | d } | rK | j j n | j j d j | d k r | j j d i | d d 6 n | r | j j j } x, | j | d k r | j j j } q Wn d S)Nr< r r zend-tag-too-earlyr{ )
rE r r: r r r} r< rJ r~ r )r> r inScoperA r# r# r$ re 1 s !z*getPhases..InBodyPhase.endTagBlockc S s | j j } d | j _ | d k s7 | j j | rT | j j d i d d 6 nS | j j | j j d | k r | j j d i d d 6 n | j j j | d S)Nzunexpected-end-tagr r< r zend-tag-too-early-ignoredr{ )r: r r rJ r~ r r} r )r> r rA r# r# r$ rf ? s
z)getPhases..InBodyPhase.endTagFormc S s | d d k r d } n d } | j j | d d | s\ | j j d i | d d 6 n | j j d | d | j j d j | d k r | j j d i | d d 6 n | j j j } x) | j | d k r | j j j } q Wd S)
Nr< r r r zunexpected-end-tagexcluder zend-tag-too-earlyr{ )r: r rJ r~ r r} r< r )r> r r rA r# r# r$ rh L s ! z-getPhases..InBodyPhase.endTagListItemc S s x1 t D]) } | j j | r | j j Pq q W| j j d j | d k rr | j j d i | d d 6 n x^ t D]V } | j j | ry | j j j } x% | j t k r | j j j } q WPqy qy Wd S)Nr r< zend-tag-too-earlyr{ ) r r: r r r} r< rJ r~ r )r> r itemr# r# r$ ri ] s
!
z,getPhases..InBodyPhase.endTagHeadingc S s9 d } x,| d k r4| d 7} | j j | d } | sd | | j j k ru | j j | j ru | j | d S| | j j k r | j j d i | d d 6 | j j j | d S| j j | j s | j j d i | d d 6 d S| | j j d k r*| j j d i | d d 6 n | j j j
| } d } x7 | j j | d D] } | j t k r\| } Pq\q\W| d k r| j j j
} x" | | k r| j j j
} qW| j j j | d S| j j | d } | j j j
| } | } }
d } | j j j
|
} x| d k r?| d 7} | d 8} | j j | }
|
| j j k r| j j j |
q*n |
| k rPn | | k r| j j j
|
d } n |
j }
|
| j j | j j j
|
<|
| j j | j j j
|
<|
}
| j r)| j j | n |
j | |
} q*W| j r\| j j | n | j t d k r| j j \ } } | j | | n
| j | | j }
| j |
| j |
| j j j | | j j j | |
| j j j | | j j j | j j j
| d |
q Wd S)z)The much-feared adoption agency algorithmr r r< Nzadoption-agency-1.2zadoption-agency-4.4zadoption-agency-1.3rv r7 r/ r1 r0 r- r{ )ztableztbodyztfootztheadztr)r: r r} r r< r rJ r~ rw r indexr r r cloneNoder r appendChildr getTableMisnestedNodePosition insertBeforereparentChildreninsert)r> r outerLoopCounterformattingElementafeIndex
furthestBlockri commonAncestorbookmarklastNoderA innerLoopCounterr cloner r r# r# r$ rj l s
!
z/getPhases..InBodyPhase.endTagFormattingc S s | j j | d r& | j j n | j j d j | d k rd | j j d i | d d 6 n | j j | d r | j j j } x) | j | d k r | j j j } q W| j j n d S)Nr< r zend-tag-too-earlyr{ ) r: r r r} r< rJ r~ r clearActiveFormattingElements)r> r ri r# r# r$ rk s !z8getPhases..InBodyPhase.endTagAppletMarqueeObjectc S s[ | j j d i d d 6d d 6 | j j | j j t d d | j j j d S)Nz unexpected-end-tag-treated-asr r! z
br elementr rn )rJ r~ r: r} rG r r} r )r> r r# r# r$ rl s
z'getPhases..InBodyPhase.endTagBrc S s x | j j d d d D] } | j | d k r | j j d | d | j j d j | d k r | j j d i | d d 6 n x | j j j | k r q WPq | j t k r | j j d i | d d 6 Pq q Wd S)Nr r< r zunexpected-end-tagr{ r{ ) r: r} r< r rJ r~ r r r )r> r rA r# r# r$ r " s ! z*getPhases..InBodyPhase.endTagOtherN)5r5 r6 r7 rG ru ry r r~ r r rF r r rG rI rJ rK rL rH rM rN rO rP rQ rR rS rT rV rU rW rX rY rZ r[ r\ r_ r] r^ r` ra rb r rg rc rd re rf rh ri rj rk rl r r# )rb r# r$ InBodyPhase[ sd F
r c s^ e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
S)z getPhases..TextPhasec se j | | | t j g | _ | j | j _ t j d | j f g | _ | j | j _ d S)Nr ) rG r r r^ r r endTagScriptra r )r> rJ r: )rb r# r$ rG 1 s z%getPhases..TextPhase.__init__c S s | j j | d d S)Nrs )r: r] )r> r r# r# r$ r 9 s z.getPhases..TextPhase.processCharactersc S sM | j j d i | j j d j d 6 | j j j | j j | j _ d S)Nz&expected-named-closing-tag-but-got-eofr r< Tr{ )rJ r~ r: r} r< r rH r] )r> r# r# r$ r < s
z'getPhases..TextPhase.processEOFc S s d s t d | d d S)NFz4Tried to process start tag %s in RCDATA/RAWTEXT moder< )r )r> r r# r# r$ r C s z*getPhases..TextPhase.startTagOtherc S s= | j j j } | j d k s' t | j j | j _ d S)Nr )r: r} r r< r rJ rH r] )r> r rA r# r# r$ r F s z)getPhases..TextPhase.endTagScriptc S s& | j j j | j j | j _ d S)N)r: r} r rJ rH r] )r> r r# r# r$ r M s z(getPhases..TextPhase.endTagOtherN) r5 r6 r7 rG r r r r r r# )rb r# r$ TextPhase0 s r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d d Z
d d Z d d Z d d Z d d Z d d Z d! d" Z d# d$ Z d% d& Z d' S)(z getPhases..InTablePhasec
s j | | | t j d | j f d | j f d | j f d | j f d | j f d | j f d | j f d | j
f d | j f d | j f g
| _
| j | j
_ t j d | j f d | j f g | _ | j | j _ d S)Nr> r3 r5 r r/ r1 r0 r* r+ r- r7 r r r r r: )tbodytfootthead)tdthtr)zstylezscript) zbodyzcaptionzcolzcolgroupzhtmlr r r r r r )rG r r r` startTagCaptionstartTagColgroup startTagColstartTagRowGroupstartTagImplyTbodyrS startTagStyleScriptrV rJ r^ r r endTagTable endTagIgnorera r )r> rJ r: )rb r# r$ rG S s$ z(getPhases..InTablePhase.__init__c S s4 x- | j j d j d k r/ | j j j q Wd S)Nr r7 r> r{ )ztablezhtml)r: r} r< r )r> r# r# r$ clearStackToTableContextk s z8getPhases..InTablePhase.clearStackToTableContextc S sB | j j d j d k r, | j j d n | j j s> t d S)Nr r> z eof-in-tabler{ )r: r} r< rJ r~ rO r )r> r# r# r$ r t s z*getPhases..InTablePhase.processEOFc S sH | j j } | j j d | j _ | | j j _ | j j j | d S)N inTableText)rJ r] rE rH r )r> r rH r# r# r$ r { s z6getPhases..InTablePhase.processSpaceCharactersc S sH | j j } | j j d | j _ | | j j _ | j j j | d S)Nr )rJ r] rE rH r )r> r rH r# r# r$ r s z1getPhases..InTablePhase.processCharactersc S s3 d | j _ | j j d j | d | j _ d S)NTr8 F)r: insertFromTablerJ rE r )r> r r# r# r$ r] s z*getPhases..InTablePhase.insertTextc S sG | j | j j j t | j j | | j j d | j _ d S)Nr2 ) r r: rw r r
rG rJ rE r] )r> r r# r# r$ r s
z/getPhases..InTablePhase.startTagCaptionc S s4 | j | j j | | j j d | j _ d S)Nr4 )r r: rG rJ rE r] )r> r r# r# r$ r s
z0getPhases..InTablePhase.startTagColgroupc S s | j t d d | S)Nr5 rn )r r )r> r r# r# r$ r s z+getPhases..InTablePhase.startTagColc S s4 | j | j j | | j j d | j _ d S)Nr. )r r: rG rJ rE r] )r> r r# r# r$ r s
z0getPhases..InTablePhase.startTagRowGroupc S s | j t d d | S)Nr/ rn )r r )r> r r# r# r$ r s z2getPhases..InTablePhase.startTagImplyTbodyc S sN | j j d i d d 6d d 6 | j j j t d | j j sJ | Sd S)Nz$unexpected-start-tag-implies-end-tagr7 r r )rJ r~ r] r r rO )r> r r# r# r$ rS s
z-getPhases..InTablePhase.startTagTablec S s | j j d j | S)Nr )rJ rE r )r> r r# r# r$ r s z3getPhases..InTablePhase.startTagStyleScriptc S sq d | d k r` | d d j t d k r` | j j d | j j | | j j j n
| j | d S)Nr, rs r z unexpected-hidden-input-in-table) rh r rJ r~ r: rG r} r r )r> r r# r# r$ rV s z-getPhases..InTablePhase.startTagInputc S s_ | j j d | j j d k r[ | j j | | j j d | j _ | j j j n d S)Nzunexpected-form-in-tabler r{ )rJ r~ r: r rG r} r )r> r r# r# r$ rJ s
z,getPhases..InTablePhase.startTagFormc S sQ | j j d i | d d 6 d | j _ | j j d j | d | j _ d S)Nz)unexpected-start-tag-implies-table-voodoor< Tr8 F)rJ r~ r: r rE r )r> r r# r# r$ r s z-getPhases..InTablePhase.startTagOtherc S s | j j d d d r | j j | j j d j d k ro | j j d i d d 6| j j d j d 6 n x- | j j d j d k r | j j j qr W| j j j | j j n | j j s t
| j j d S)
Nr7 r r zend-tag-too-early-namedr r r{ r{ r{ ) r: r r r} r< rJ r~ r r_ rO r )r> r r# r# r$ r s
z+getPhases..InTablePhase.endTagTablec S s" | j j d i | d d 6 d S)Nzunexpected-end-tagr< )rJ r~ )r> r r# r# r$ r s z,getPhases..InTablePhase.endTagIgnorec S sQ | j j d i | d d 6 d | j _ | j j d j | d | j _ d S)Nz'unexpected-end-tag-implies-table-voodoor< Tr8 F)rJ r~ r: r rE r )r> r r# r# r$ r s z+getPhases..InTablePhase.endTagOtherN)r5 r6 r7 rG r r r r r] r r r r r rS r rV rJ r r r r r# )rb r# r$ InTablePhaseQ s&
r c sv e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d S)z#getPhases..InTableTextPhasec s) j | | | d | _ g | _ d S)N)rG rH characterTokens)r> rJ r: )rb r# r$ rG s z,getPhases..InTableTextPhase.__init__c S s d j d d | j D } t d d | D rj i t d d 6| d 6} | j j d j | n | r | j j | n g | _ d S) Nrg c S s g | ] } | d q S)rs r# )r; r r# r# r$ r? s zGgetPhases..InTableTextPhase.flushCharacters..c S s g | ] } | t k q Sr# )r
)r; r r# r# r$ r? s rl r, rs r6 )joinr r r rJ rE r] r: )r> rs r r# r# r$ flushCharacters s z3getPhases..InTableTextPhase.flushCharactersc S s | j | j | j _ | S)N)r rH rJ r] )r> r r# r# r$ r s
z2getPhases..InTableTextPhase.processCommentc S s | j | j | j _ d S)NT)r rH rJ r] )r> r# r# r$ r s
z.getPhases..InTableTextPhase.processEOFc S s( | d d k r d 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 rH rJ r] )r> r r# r# r$ r s
z3getPhases..InTableTextPhase.processStartTagc S s | j | j | j _ | S)N)r rH rJ r] )r> r r# r# r$ r
s
z1getPhases..InTableTextPhase.processEndTagN) r5 r6 r7 rG r r r r r r r r# )rb r# r$ InTableTextPhase s r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d S)z!getPhases..InCaptionPhasec
s j | | | t j d | j f d
| j f g | _ | j | j _ t j d | j f d | j f d | j
f g | _ | j | j _ d S)Nr> r3 r r5 r/ r* r1 r+ r0 r- r7 r: ) zcaptioncolcolgrouptbodytdtfootththeadtr)
zbodyr r zhtmlr r r r r r )
rG r r r` startTagTableElementr^ r r
endTagCaptionr r ra r )r> rJ r: )rb r# r$ rG s z*getPhases..InCaptionPhase.__init__c S s | j j d d d S)Nr3 r r7 )r: r )r> r# r# r$ ignoreEndTagCaption# s z5getPhases..InCaptionPhase.ignoreEndTagCaptionc S s | j j d j d S)Nr8 )rJ rE r )r> r# r# r$ r & s z,getPhases..InCaptionPhase.processEOFc S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r ) s z3getPhases..InCaptionPhase.processCharactersc S s@ | j j | j } | j j j t d | s< | Sd S)Nr3 )rJ r~ r r] r r )r> r ignoreEndTagr# r# r$ r , s
z6getPhases..InCaptionPhase.startTagTableElementc S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r 4 s z/getPhases..InCaptionPhase.startTagOtherc S s | j s | j j | j j d j d k rc | j j d i d d 6| j j d j d 6 n x- | j j d j d k r | j j j qf W| j j j | j j | j j d | j _
n | j j s t | j j d S)
Nr r3 z$expected-one-end-tag-but-got-anotherr r r6 r{ r{ r{ )
r r: r r} r< rJ r~ r r rE r] rO r )r> r r# r# r$ r 7 s
z/getPhases..InCaptionPhase.endTagCaptionc S s@ | j j | j } | j j j t d | s< | Sd S)Nr3 )rJ r~ r r] r r )r> r r r# r# r$ r I s
z-getPhases..InCaptionPhase.endTagTablec S s" | j j d i | d d 6 d S)Nzunexpected-end-tagr< )rJ r~ )r> r r# r# r$ r P s z.getPhases..InCaptionPhase.endTagIgnorec S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r S s z-getPhases..InCaptionPhase.endTagOtherN)
r5 r6 r7 rG r r r r r r r r r r# )rb r# r$ InCaptionPhase s r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d S)z%getPhases..InColumnGroupPhasec s j | | | t j d | j f d | j f g | _ | j | j _ t j d | j f d | j f g | _
| j | j
_ d S)Nr> r r5 ) rG r r r` r r^ r r endTagColgroup endTagColra r )r> rJ r: )rb r# r$ rG Y s z.getPhases..InColumnGroupPhase.__init__c S s | j j d j d k S)Nr r> r{ )r: r} r< )r> r# r# r$ ignoreEndTagColgrouph s z:getPhases..InColumnGroupPhase.ignoreEndTagColgroupc S s\ | j j d j d k r/ | j j s+ t d S| j } | j t d | sX d Sd S)Nr r> r5 Tr{ ) r: r} r< rJ rO r r r r )r> r r# r# r$ r k s z0getPhases..InColumnGroupPhase.processEOFc S s- | j } | j t d | s) | Sd S)Nr5 )r r r )r> r r r# r# r$ r u s z7getPhases..InColumnGroupPhase.processCharactersc S s$ | j j | | j j j d S)N)r: rG r} r )r> r r# r# r$ r { s z1getPhases..InColumnGroupPhase.startTagColc S s- | j } | j t d | s) | Sd S)Nr5 )r r r )r> r r r# r# r$ r s z3getPhases..InColumnGroupPhase.startTagOtherc S sX | j r. | j j s t | j j n& | j j j | j j d | j _ d S)Nr6 )
r rJ rO r r~ r: r} r rE r] )r> r r# r# r$ r s
z4getPhases..InColumnGroupPhase.endTagColgroupc S s | j j d i d d 6 d S)Nz
no-end-tagr r< )rJ r~ )r> r r# r# r$ r s z/getPhases..InColumnGroupPhase.endTagColc S s- | j } | j t d | s) | Sd S)Nr5 )r r r )r> r r r# r# r$ r s z1getPhases..InColumnGroupPhase.endTagOtherN) r5 r6 r7 rG r r r r r r r r r# )rb r# r$ InColumnGroupPhaseV s
r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d d Z
d d Z d d Z d S) z#getPhases..InTableBodyPhasec s j | | | t j d | j f d | j f d
| j f d | j f g | _ | j | j _ t j d | j
f d | j f d | j f g | _
| j | j
_ d S)Nr> r- r* r+ r3 r r5 r/ r1 r0 r7 r: )tdth)captioncolcolgrouptbodytfootthead)r r r )zbodyr r r zhtmlr r ztr)rG r r r`
startTagTrstartTagTableCellstartTagTableOtherr^ r r endTagTableRowGroupr r ra r )r> rJ r: )rb r# r$ rG s z,getPhases..InTableBodyPhase.__init__c S sb x- | j j d j d k r/ | j j j q W| j j d j d k r^ | j j s^ t n d S) Nr r/ r1 r0 r> r{ )ztbodyztfootztheadzhtmlr{ )r: r} r< r rJ rO r )r> r# r# r$ clearStackToTableBodyContext s
z@getPhases..InTableBodyPhase.clearStackToTableBodyContextc S s | j j d j d S)Nr6 )rJ rE r )r> r# r# r$ r s z.getPhases..InTableBodyPhase.processEOFc S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r s z:getPhases..InTableBodyPhase.processSpaceCharactersc S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r s z5getPhases..InTableBodyPhase.processCharactersc S s4 | j | j j | | j j d | j _ d S)Nr, )r r: rG rJ rE r] )r> r r# r# r$ r s
z.getPhases..InTableBodyPhase.startTagTrc S s8 | j j d i | d d 6 | j t d d | S)Nz unexpected-cell-in-table-bodyr< r- rn )rJ r~ r r )r> r r# r# r$ r s z5getPhases..InTableBodyPhase.startTagTableCellc S s | j j d d d sH | j j d d d sH | j j d d d rv | j | j t | j j d j | S| j j s t | j j
d S)Nr/ r r7 r0 r1 r r{ ) r: r r r r r} r< rJ rO r r~ )r> r r# r# r$ r s
z6getPhases..InTableBodyPhase.startTagTableOtherc S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r s z1getPhases..InTableBodyPhase.startTagOtherc S sq | j j | d d d rO | j | j j j | j j d | j _ n | j j d i | d d 6 d S)Nr< r r7 r6 z unexpected-end-tag-in-table-body) r: r r r} r rJ rE r] r~ )r> r r# r# r$ r s
z7getPhases..InTableBodyPhase.endTagTableRowGroupc S s | j j d d d sH | j j d d d sH | j j d d d rv | j | j t | j j d j | S| j j s t | j j
d S)Nr/ r r7 r0 r1 r r{ ) r: r r r r r} r< rJ rO r r~ )r> r r# r# r$ r s
z/getPhases..InTableBodyPhase.endTagTablec S s" | j j d i | d d 6 d S)Nz unexpected-end-tag-in-table-bodyr< )rJ r~ )r> r r# r# r$ r s z0getPhases..InTableBodyPhase.endTagIgnorec S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r s z/getPhases..InTableBodyPhase.endTagOtherN)r5 r6 r7 rG r r r r r r r r r r r r r# )rb r# r$ InTableBodyPhase s
r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d d Z
d d Z d d Z d d Z d S) z getPhases..InRowPhasec s j | | | t j d | j f d
| j f d | j f g | _ | j | j _ t j d
| j f d | j
f d | j f d | j f g | _
| j | j
_ d S)Nr> r* r+ r3 r r5 r/ r1 r0 r- r7 r: )tdth)captioncolcolgrouptbodytfoottheadztr)r r r )zbodyr r r zhtmlr r )rG r r r` r r r^ r r endTagTrr r r ra r )r> rJ r: )rb r# r$ rG s z&getPhases..InRowPhase.__init__c S s[ xT | j j d j d k rV | j j d i | j j d j d 6 | j j j q Wd S) Nr r- r> z'unexpected-implied-end-tag-in-table-rowr< r{ )ztrzhtmlr{ )r: r} r< rJ r~ r )r> r# r# r$ clearStackToTableRowContext s z9getPhases..InRowPhase.clearStackToTableRowContextc S s | j j d d d S)Nr- r r7 )r: r )r> r# r# r$ ignoreEndTagTr s z,getPhases..InRowPhase.ignoreEndTagTrc S s | j j d j d S)Nr6 )rJ rE r )r> r# r# r$ r s z(getPhases..InRowPhase.processEOFc S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r s z4getPhases..InRowPhase.processSpaceCharactersc S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r s z/getPhases..InRowPhase.processCharactersc S sG | j | j j | | j j d | j _ | j j j t d S)Nr) ) r r: rG rJ rE r] rw r r
)r> r r# r# r$ r " s
z/getPhases..InRowPhase.startTagTableCellc S s- | j } | j t d | s) | Sd S)Nr- )r r r )r> r r r# r# r$ r ( s z0getPhases..InRowPhase.startTagTableOtherc S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r / s z+getPhases..InRowPhase.startTagOtherc S sb | j s? | j | j j j | j j d | j _ n | j j sQ t | j j
d S)Nr. ) r r r: r} r rJ rE r] rO r r~ )r> r r# r# r$ r 2 s
z&getPhases..InRowPhase.endTagTrc S s- | j } | j t d | s) | Sd S)Nr- )r r r )r> r r r# r# r$ r < s z)getPhases..InRowPhase.endTagTablec S sD | j j | d d d r3 | j t d | S| j j d S)Nr< r r7 r- )r: r r r rJ r~ )r> r r# r# r$ r D s z1getPhases..InRowPhase.endTagTableRowGroupc S s" | j j d i | d d 6 d S)Nz unexpected-end-tag-in-table-rowr< )rJ r~ )r> r r# r# r$ r K s z*getPhases..InRowPhase.endTagIgnorec S s | j j d j | S)Nr6 )rJ rE r )r> r r# r# r$ r O s z)getPhases..InRowPhase.endTagOtherN)r5 r6 r7 rG r r r r r r r r r r r r r r# )rb r# r$
InRowPhase s
r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d S)z getPhases..InCellPhasec s j | | | t j d | j f d
| j f g | _ | j | j _ t j d | j f d | j f d | j
f g | _ | j | j _ d S)Nr> r3 r r5 r/ r* r1 r+ r0 r- r: r7 ) captioncolcolgrouptbodytdtfootththeadtr)r r )zbodyr r r zhtml)ztabler r r r )
rG r r r` r r^ r r endTagTableCellr endTagImplyra r )r> rJ r: )rb r# r$ rG T s z'getPhases..InCellPhase.__init__c S s` | j j d d d r. | j t d n. | j j d d d r\ | j t d n d S)Nr* r r7 r+ )r: r r r )r> r# r# r$ closeCelle s z(getPhases..InCellPhase.closeCellc S s | j j d j d S)Nr8 )rJ rE r )r> r# r# r$ r l s z)getPhases..InCellPhase.processEOFc S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r o s z0getPhases..InCellPhase.processCharactersc S sa | j j d d d s0 | j j d d d r> | j | S| j j sP t | j j d S)Nr* r r7 r+ )r: r r rJ rO r r~ )r> r r# r# r$ r r s
z1getPhases..InCellPhase.startTagTableOtherc S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r | s z,getPhases..InCellPhase.startTagOtherc S s | j j | d d d r | j j | d | j j d j | d k r | j j d i | d d 6 x@ | j j j } | j | d k rn Pqn qn Wn | j j j | j j | j j d | j _
n | j j d i | d d 6 d S) Nr< r r7 r zunexpected-cell-end-tagr, zunexpected-end-tagr{ ) r: r r r} r< rJ r~ r r rE r] )r> r rA r# r# r$ r s
z.getPhases..InCellPhase.endTagTableCellc S s" | j j d i | d d 6 d S)Nzunexpected-end-tagr< )rJ r~ )r> r r# r# r$ r s z+getPhases..InCellPhase.endTagIgnorec S s; | j j | d d d r* | j | S| j j d S)Nr< r r7 )r: r r rJ r~ )r> r r# r# r$ r s
z*getPhases..InCellPhase.endTagImplyc S s | j j d j | S)Nr8 )rJ rE r )r> r r# r# r$ r s z*getPhases..InCellPhase.endTagOtherN)
r5 r6 r7 rG r r r r r r r r r r# )rb r# r$ InCellPhaseR s
r c s e Z d Z f d d Z d d Z d d Z d d Z d d
Z d d Z d
d Z d d Z
d d Z d d Z d d Z
d d Z d d Z d S) z getPhases..InSelectPhasec s j | | | t j d | j f d | j f d | j f d | j f d | j f d | j f g | _ | j
| j _ t j d | j f d | j
f d | j f g | _ | j | j _ d S)
Nr> r r r( r r
r r )zinputzkeygenztextarea)rG r r r` startTagOptionstartTagOptgroupr] rV r r^ r r endTagOptionendTagOptgroup endTagSelectra r )r> rJ r: )rb r# r$ rG s z)getPhases..InSelectPhase.__init__c S sB | j j d j d k r, | j j d n | j j s> t d S)Nr r> z
eof-in-selectr{ )r: r} r< rJ r~ rO r )r> r# r# r$ r s z+getPhases..InSelectPhase.processEOFc S s, | d d k r d S| j j | d d S)Nrs r )r: r] )r> r r# r# r$ r s z2getPhases..InSelectPhase.processCharactersc S s@ | j j d j d k r, | j j j n | j j | d S)Nr r r{ )r: r} r< r rG )r> r r# r# r$ r s z/getPhases..InSelectPhase.startTagOptionc S sl | j j d j d k r, | j j j n | j j d j d k rX | j j j n | j j | d S)Nr r r r{ r{ )r: r} r< r rG )r> r r# r# r$ r s
z1getPhases..InSelectPhase.startTagOptgroupc S s' | j j d | j t d d S)Nzunexpected-select-in-selectr( )rJ r~ r r )r> r r# r# r$ r] s z/getPhases..InSelectPhase.startTagSelectc S sU | j j d | j j d d d r? | j t d | S| j j sQ t d S)Nzunexpected-input-in-selectr( r )rJ r~ r: r r r rO r )r> r r# r# r$ rV s
z.getPhases..InSelectPhase.startTagInputc S s | j j d j | S)Nr )rJ rE r )r> r r# r# r$ r s z/getPhases..InSelectPhase.startTagScriptc S s" | j j d i | d d 6 d S)Nz unexpected-start-tag-in-selectr< )rJ r~ )r> r r# r# r$ r s z.getPhases..InSelectPhase.startTagOtherc S sJ | j j d j d k r, | j j j n | j j d i d d 6 d S)Nr r z unexpected-end-tag-in-selectr< r{ )r: r} r< r rJ r~ )r> r r# r# r$ r s z-getPhases.