HTMLDocumentParser.cpp 117 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658165916601661166216631664166516661667166816691670167116721673167416751676167716781679168016811682168316841685168616871688168916901691169216931694169516961697169816991700170117021703170417051706170717081709171017111712171317141715171617171718171917201721172217231724172517261727172817291730173117321733173417351736173717381739174017411742174317441745174617471748174917501751175217531754175517561757175817591760176117621763176417651766176717681769177017711772177317741775177617771778177917801781178217831784178517861787178817891790179117921793179417951796179717981799180018011802180318041805180618071808180918101811181218131814181518161817181818191820182118221823182418251826182718281829183018311832183318341835183618371838183918401841184218431844184518461847184818491850185118521853185418551856185718581859186018611862186318641865186618671868186918701871187218731874187518761877187818791880188118821883188418851886188718881889189018911892189318941895189618971898189919001901190219031904190519061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192919301931193219331934193519361937193819391940194119421943194419451946194719481949195019511952195319541955195619571958195919601961196219631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994199519961997199819992000200120022003200420052006200720082009201020112012201320142015201620172018201920202021202220232024202520262027202820292030203120322033203420352036203720382039204020412042204320442045204620472048204920502051205220532054205520562057205820592060206120622063206420652066206720682069207020712072207320742075207620772078207920802081208220832084208520862087208820892090209120922093209420952096209720982099210021012102210321042105210621072108210921102111211221132114211521162117211821192120212121222123212421252126212721282129213021312132213321342135213621372138213921402141214221432144214521462147214821492150215121522153215421552156215721582159216021612162216321642165216621672168216921702171217221732174217521762177217821792180218121822183218421852186218721882189219021912192219321942195219621972198219922002201220222032204220522062207220822092210221122122213221422152216221722182219222022212222222322242225222622272228222922302231223222332234223522362237223822392240224122422243224422452246224722482249225022512252225322542255225622572258225922602261226222632264226522662267226822692270227122722273227422752276227722782279228022812282228322842285228622872288228922902291229222932294229522962297229822992300230123022303230423052306230723082309231023112312231323142315231623172318231923202321232223232324232523262327232823292330233123322333233423352336233723382339234023412342234323442345234623472348234923502351235223532354235523562357235823592360236123622363236423652366236723682369237023712372237323742375237623772378237923802381238223832384238523862387238823892390239123922393239423952396239723982399240024012402240324042405240624072408240924102411241224132414241524162417241824192420242124222423242424252426242724282429243024312432243324342435243624372438243924402441244224432444244524462447244824492450245124522453245424552456245724582459246024612462246324642465246624672468246924702471247224732474247524762477247824792480248124822483248424852486248724882489249024912492249324942495249624972498249925002501250225032504250525062507250825092510251125122513251425152516251725182519252025212522252325242525252625272528252925302531253225332534253525362537253825392540254125422543254425452546254725482549255025512552255325542555255625572558255925602561256225632564256525662567256825692570257125722573257425752576257725782579258025812582258325842585258625872588258925902591259225932594259525962597259825992600260126022603260426052606260726082609261026112612261326142615261626172618261926202621262226232624262526262627262826292630263126322633263426352636263726382639264026412642264326442645264626472648264926502651265226532654265526562657265826592660266126622663266426652666266726682669267026712672267326742675267626772678267926802681268226832684268526862687268826892690269126922693269426952696269726982699270027012702270327042705270627072708270927102711271227132714271527162717271827192720272127222723272427252726272727282729273027312732273327342735273627372738273927402741274227432744274527462747274827492750275127522753275427552756275727582759276027612762276327642765276627672768276927702771277227732774277527762777277827792780278127822783278427852786278727882789279027912792279327942795279627972798279928002801280228032804280528062807280828092810281128122813281428152816281728182819282028212822282328242825282628272828282928302831283228332834283528362837283828392840284128422843284428452846284728482849285028512852285328542855285628572858285928602861286228632864286528662867286828692870287128722873287428752876287728782879288028812882288328842885288628872888288928902891289228932894289528962897289828992900290129022903290429052906290729082909291029112912291329142915291629172918291929202921292229232924292529262927292829292930293129322933293429352936293729382939294029412942294329442945294629472948294929502951295229532954295529562957295829592960296129622963296429652966296729682969297029712972297329742975297629772978297929802981298229832984298529862987298829892990299129922993299429952996299729982999300030013002300330043005300630073008300930103011301230133014301530163017301830193020302130223023302430253026302730283029303030313032303330343035303630373038303930403041304230433044304530463047304830493050305130523053
  1. /*
  2. * Copyright (c) 2020, Andreas Kling <kling@serenityos.org>
  3. *
  4. * SPDX-License-Identifier: BSD-2-Clause
  5. */
  6. #include <AK/Debug.h>
  7. #include <AK/SourceLocation.h>
  8. #include <AK/Utf32View.h>
  9. #include <LibTextCodec/Decoder.h>
  10. #include <LibWeb/DOM/Comment.h>
  11. #include <LibWeb/DOM/Document.h>
  12. #include <LibWeb/DOM/DocumentType.h>
  13. #include <LibWeb/DOM/ElementFactory.h>
  14. #include <LibWeb/DOM/Event.h>
  15. #include <LibWeb/DOM/Text.h>
  16. #include <LibWeb/DOM/Window.h>
  17. #include <LibWeb/HTML/EventNames.h>
  18. #include <LibWeb/HTML/HTMLFormElement.h>
  19. #include <LibWeb/HTML/HTMLHeadElement.h>
  20. #include <LibWeb/HTML/HTMLScriptElement.h>
  21. #include <LibWeb/HTML/HTMLTableElement.h>
  22. #include <LibWeb/HTML/HTMLTemplateElement.h>
  23. #include <LibWeb/HTML/Parser/HTMLDocumentParser.h>
  24. #include <LibWeb/HTML/Parser/HTMLEncodingDetection.h>
  25. #include <LibWeb/HTML/Parser/HTMLToken.h>
  26. #include <LibWeb/Namespace.h>
  27. #include <LibWeb/SVG/TagNames.h>
  28. namespace Web::HTML {
  29. static inline void log_parse_error(const SourceLocation& location = SourceLocation::current())
  30. {
  31. dbgln("Parse error! {}", location);
  32. }
  33. static Vector<FlyString> s_quirks_public_ids = {
  34. "+//Silmaril//dtd html Pro v0r11 19970101//",
  35. "-//AS//DTD HTML 3.0 asWedit + extensions//",
  36. "-//AdvaSoft Ltd//DTD HTML 3.0 asWedit + extensions//",
  37. "-//IETF//DTD HTML 2.0 Level 1//",
  38. "-//IETF//DTD HTML 2.0 Level 2//",
  39. "-//IETF//DTD HTML 2.0 Strict Level 1//",
  40. "-//IETF//DTD HTML 2.0 Strict Level 2//",
  41. "-//IETF//DTD HTML 2.0 Strict//",
  42. "-//IETF//DTD HTML 2.0//",
  43. "-//IETF//DTD HTML 2.1E//",
  44. "-//IETF//DTD HTML 3.0//",
  45. "-//IETF//DTD HTML 3.2 Final//",
  46. "-//IETF//DTD HTML 3.2//",
  47. "-//IETF//DTD HTML 3//",
  48. "-//IETF//DTD HTML Level 0//",
  49. "-//IETF//DTD HTML Level 1//",
  50. "-//IETF//DTD HTML Level 2//",
  51. "-//IETF//DTD HTML Level 3//",
  52. "-//IETF//DTD HTML Strict Level 0//",
  53. "-//IETF//DTD HTML Strict Level 1//",
  54. "-//IETF//DTD HTML Strict Level 2//",
  55. "-//IETF//DTD HTML Strict Level 3//",
  56. "-//IETF//DTD HTML Strict//",
  57. "-//IETF//DTD HTML//",
  58. "-//Metrius//DTD Metrius Presentational//",
  59. "-//Microsoft//DTD Internet Explorer 2.0 HTML Strict//",
  60. "-//Microsoft//DTD Internet Explorer 2.0 HTML//",
  61. "-//Microsoft//DTD Internet Explorer 2.0 Tables//",
  62. "-//Microsoft//DTD Internet Explorer 3.0 HTML Strict//",
  63. "-//Microsoft//DTD Internet Explorer 3.0 HTML//",
  64. "-//Microsoft//DTD Internet Explorer 3.0 Tables//",
  65. "-//Netscape Comm. Corp.//DTD HTML//",
  66. "-//Netscape Comm. Corp.//DTD Strict HTML//",
  67. "-//O'Reilly and Associates//DTD HTML 2.0//",
  68. "-//O'Reilly and Associates//DTD HTML Extended 1.0//",
  69. "-//O'Reilly and Associates//DTD HTML Extended Relaxed 1.0//",
  70. "-//SQ//DTD HTML 2.0 HoTMetaL + extensions//",
  71. "-//SoftQuad Software//DTD HoTMetaL PRO 6.0::19990601::extensions to HTML 4.0//",
  72. "-//SoftQuad//DTD HoTMetaL PRO 4.0::19971010::extensions to HTML 4.0//",
  73. "-//Spyglass//DTD HTML 2.0 Extended//",
  74. "-//Sun Microsystems Corp.//DTD HotJava HTML//",
  75. "-//Sun Microsystems Corp.//DTD HotJava Strict HTML//",
  76. "-//W3C//DTD HTML 3 1995-03-24//",
  77. "-//W3C//DTD HTML 3.2 Draft//",
  78. "-//W3C//DTD HTML 3.2 Final//",
  79. "-//W3C//DTD HTML 3.2//",
  80. "-//W3C//DTD HTML 3.2S Draft//",
  81. "-//W3C//DTD HTML 4.0 Frameset//",
  82. "-//W3C//DTD HTML 4.0 Transitional//",
  83. "-//W3C//DTD HTML Experimental 19960712//",
  84. "-//W3C//DTD HTML Experimental 970421//",
  85. "-//W3C//DTD W3 HTML//",
  86. "-//W3O//DTD W3 HTML 3.0//",
  87. "-//WebTechs//DTD Mozilla HTML 2.0//",
  88. "-//WebTechs//DTD Mozilla HTML//"
  89. };
  90. RefPtr<DOM::Document> parse_html_document(const StringView& data, const URL& url, const String& encoding)
  91. {
  92. auto document = DOM::Document::create(url);
  93. HTMLDocumentParser parser(document, data, encoding);
  94. parser.run(url);
  95. return document;
  96. }
  97. HTMLDocumentParser::HTMLDocumentParser(DOM::Document& document, const StringView& input, const String& encoding)
  98. : m_tokenizer(input, encoding)
  99. , m_document(document)
  100. {
  101. m_document->set_should_invalidate_styles_on_attribute_changes(false);
  102. auto standardized_encoding = TextCodec::get_standardized_encoding(encoding);
  103. VERIFY(standardized_encoding.has_value());
  104. m_document->set_encoding(standardized_encoding.value());
  105. }
  106. HTMLDocumentParser::~HTMLDocumentParser()
  107. {
  108. m_document->set_should_invalidate_styles_on_attribute_changes(true);
  109. }
  110. void HTMLDocumentParser::run(const URL& url)
  111. {
  112. m_document->set_url(url);
  113. m_document->set_source(m_tokenizer.source());
  114. for (;;) {
  115. auto optional_token = m_tokenizer.next_token();
  116. if (!optional_token.has_value())
  117. break;
  118. auto& token = optional_token.value();
  119. dbgln_if(PARSER_DEBUG, "[{}] {}", insertion_mode_name(), token.to_string());
  120. // FIXME: If the adjusted current node is a MathML text integration point and the token is a start tag whose tag name is neither "mglyph" nor "malignmark"
  121. // FIXME: If the adjusted current node is a MathML text integration point and the token is a character token
  122. // FIXME: If the adjusted current node is a MathML annotation-xml element and the token is a start tag whose tag name is "svg"
  123. // FIXME: If the adjusted current node is an HTML integration point and the token is a start tag
  124. // FIXME: If the adjusted current node is an HTML integration point and the token is a character token
  125. if (m_stack_of_open_elements.is_empty()
  126. || adjusted_current_node().namespace_() == Namespace::HTML
  127. || token.is_end_of_file()) {
  128. process_using_the_rules_for(m_insertion_mode, token);
  129. } else {
  130. process_using_the_rules_for_foreign_content(token);
  131. }
  132. if (m_stop_parsing) {
  133. dbgln_if(PARSER_DEBUG, "Stop parsing{}! :^)", m_parsing_fragment ? " fragment" : "");
  134. break;
  135. }
  136. }
  137. flush_character_insertions();
  138. // "The end"
  139. m_document->set_ready_state("interactive");
  140. auto scripts_to_execute_when_parsing_has_finished = m_document->take_scripts_to_execute_when_parsing_has_finished({});
  141. for (auto& script : scripts_to_execute_when_parsing_has_finished) {
  142. // FIXME: Spin the event loop until the script is ready to be parser executed and there's no style sheets blocking scripts.
  143. script.execute_script();
  144. }
  145. auto content_loaded_event = DOM::Event::create(HTML::EventNames::DOMContentLoaded);
  146. content_loaded_event->set_bubbles(true);
  147. m_document->dispatch_event(content_loaded_event);
  148. // FIXME: The document parser shouldn't execute these, it should just spin the event loop until the list becomes empty.
  149. // FIXME: Once the set has been added, also spin the event loop until the set becomes empty.
  150. auto scripts_to_execute_as_soon_as_possible = m_document->take_scripts_to_execute_as_soon_as_possible({});
  151. for (auto& script : scripts_to_execute_as_soon_as_possible) {
  152. script.execute_script();
  153. }
  154. // FIXME: Spin the event loop until there is nothing that delays the load event in the Document.
  155. m_document->set_ready_state("complete");
  156. m_document->window().dispatch_event(DOM::Event::create(HTML::EventNames::load));
  157. m_document->set_ready_for_post_load_tasks(true);
  158. m_document->completely_finish_loading();
  159. }
  160. void HTMLDocumentParser::process_using_the_rules_for(InsertionMode mode, HTMLToken& token)
  161. {
  162. switch (mode) {
  163. case InsertionMode::Initial:
  164. handle_initial(token);
  165. break;
  166. case InsertionMode::BeforeHTML:
  167. handle_before_html(token);
  168. break;
  169. case InsertionMode::BeforeHead:
  170. handle_before_head(token);
  171. break;
  172. case InsertionMode::InHead:
  173. handle_in_head(token);
  174. break;
  175. case InsertionMode::InHeadNoscript:
  176. handle_in_head_noscript(token);
  177. break;
  178. case InsertionMode::AfterHead:
  179. handle_after_head(token);
  180. break;
  181. case InsertionMode::InBody:
  182. handle_in_body(token);
  183. break;
  184. case InsertionMode::AfterBody:
  185. handle_after_body(token);
  186. break;
  187. case InsertionMode::AfterAfterBody:
  188. handle_after_after_body(token);
  189. break;
  190. case InsertionMode::Text:
  191. handle_text(token);
  192. break;
  193. case InsertionMode::InTable:
  194. handle_in_table(token);
  195. break;
  196. case InsertionMode::InTableBody:
  197. handle_in_table_body(token);
  198. break;
  199. case InsertionMode::InRow:
  200. handle_in_row(token);
  201. break;
  202. case InsertionMode::InCell:
  203. handle_in_cell(token);
  204. break;
  205. case InsertionMode::InTableText:
  206. handle_in_table_text(token);
  207. break;
  208. case InsertionMode::InSelectInTable:
  209. handle_in_select_in_table(token);
  210. break;
  211. case InsertionMode::InSelect:
  212. handle_in_select(token);
  213. break;
  214. case InsertionMode::InCaption:
  215. handle_in_caption(token);
  216. break;
  217. case InsertionMode::InColumnGroup:
  218. handle_in_column_group(token);
  219. break;
  220. case InsertionMode::InTemplate:
  221. handle_in_template(token);
  222. break;
  223. case InsertionMode::InFrameset:
  224. handle_in_frameset(token);
  225. break;
  226. case InsertionMode::AfterFrameset:
  227. handle_after_frameset(token);
  228. break;
  229. case InsertionMode::AfterAfterFrameset:
  230. handle_after_after_frameset(token);
  231. break;
  232. default:
  233. VERIFY_NOT_REACHED();
  234. }
  235. }
  236. DOM::QuirksMode HTMLDocumentParser::which_quirks_mode(const HTMLToken& doctype_token) const
  237. {
  238. if (doctype_token.m_doctype.force_quirks)
  239. return DOM::QuirksMode::Yes;
  240. // NOTE: The tokenizer puts the name into lower case for us.
  241. if (doctype_token.m_doctype.name.to_string() != "html")
  242. return DOM::QuirksMode::Yes;
  243. auto public_identifier = doctype_token.m_doctype.public_identifier.to_string();
  244. auto system_identifier = doctype_token.m_doctype.system_identifier.to_string();
  245. if (public_identifier.equals_ignoring_case("-//W3O//DTD W3 HTML Strict 3.0//EN//"))
  246. return DOM::QuirksMode::Yes;
  247. if (public_identifier.equals_ignoring_case("-/W3C/DTD HTML 4.0 Transitional/EN"))
  248. return DOM::QuirksMode::Yes;
  249. if (public_identifier.equals_ignoring_case("HTML"))
  250. return DOM::QuirksMode::Yes;
  251. if (system_identifier.equals_ignoring_case("http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd"))
  252. return DOM::QuirksMode::Yes;
  253. for (auto& public_id : s_quirks_public_ids) {
  254. if (public_identifier.starts_with(public_id, CaseSensitivity::CaseInsensitive))
  255. return DOM::QuirksMode::Yes;
  256. }
  257. if (doctype_token.m_doctype.missing_system_identifier) {
  258. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  259. return DOM::QuirksMode::Yes;
  260. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  261. return DOM::QuirksMode::Yes;
  262. }
  263. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Frameset//", CaseSensitivity::CaseInsensitive))
  264. return DOM::QuirksMode::Limited;
  265. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Transitional//", CaseSensitivity::CaseInsensitive))
  266. return DOM::QuirksMode::Limited;
  267. if (!doctype_token.m_doctype.missing_system_identifier) {
  268. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  269. return DOM::QuirksMode::Limited;
  270. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  271. return DOM::QuirksMode::Limited;
  272. }
  273. return DOM::QuirksMode::No;
  274. }
  275. void HTMLDocumentParser::handle_initial(HTMLToken& token)
  276. {
  277. if (token.is_character() && token.is_parser_whitespace()) {
  278. return;
  279. }
  280. if (token.is_comment()) {
  281. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  282. document().append_child(move(comment));
  283. return;
  284. }
  285. if (token.is_doctype()) {
  286. auto doctype = adopt_ref(*new DOM::DocumentType(document()));
  287. doctype->set_name(token.m_doctype.name.to_string());
  288. doctype->set_public_id(token.m_doctype.public_identifier.to_string());
  289. doctype->set_system_id(token.m_doctype.system_identifier.to_string());
  290. document().append_child(move(doctype));
  291. document().set_quirks_mode(which_quirks_mode(token));
  292. m_insertion_mode = InsertionMode::BeforeHTML;
  293. return;
  294. }
  295. log_parse_error();
  296. document().set_quirks_mode(DOM::QuirksMode::Yes);
  297. m_insertion_mode = InsertionMode::BeforeHTML;
  298. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  299. }
  300. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  301. {
  302. if (token.is_doctype()) {
  303. log_parse_error();
  304. return;
  305. }
  306. if (token.is_comment()) {
  307. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  308. document().append_child(move(comment));
  309. return;
  310. }
  311. if (token.is_character() && token.is_parser_whitespace()) {
  312. return;
  313. }
  314. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  315. auto element = create_element_for(token, Namespace::HTML);
  316. document().append_child(element);
  317. m_stack_of_open_elements.push(move(element));
  318. m_insertion_mode = InsertionMode::BeforeHead;
  319. return;
  320. }
  321. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  322. goto AnythingElse;
  323. }
  324. if (token.is_end_tag()) {
  325. log_parse_error();
  326. return;
  327. }
  328. AnythingElse:
  329. auto element = create_element(document(), HTML::TagNames::html, Namespace::HTML);
  330. document().append_child(element);
  331. m_stack_of_open_elements.push(element);
  332. // FIXME: If the Document is being loaded as part of navigation of a browsing context, then: run the application cache selection algorithm with no manifest, passing it the Document object.
  333. m_insertion_mode = InsertionMode::BeforeHead;
  334. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  335. return;
  336. }
  337. DOM::Element& HTMLDocumentParser::current_node()
  338. {
  339. return m_stack_of_open_elements.current_node();
  340. }
  341. DOM::Element& HTMLDocumentParser::adjusted_current_node()
  342. {
  343. if (m_parsing_fragment && m_stack_of_open_elements.elements().size() == 1)
  344. return *m_context_element;
  345. return current_node();
  346. }
  347. DOM::Element& HTMLDocumentParser::node_before_current_node()
  348. {
  349. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  350. }
  351. HTMLDocumentParser::AdjustedInsertionLocation HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  352. {
  353. auto& target = current_node();
  354. HTMLDocumentParser::AdjustedInsertionLocation adjusted_insertion_location;
  355. if (m_foster_parenting && target.local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  356. auto last_template = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::template_);
  357. auto last_table = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::table);
  358. if (last_template.element && (!last_table.element || last_template.index > last_table.index)) {
  359. // This returns the template content, so no need to check the parent is a template.
  360. return { verify_cast<HTMLTemplateElement>(last_template.element)->content(), nullptr };
  361. }
  362. if (!last_table.element) {
  363. VERIFY(m_parsing_fragment);
  364. // Guaranteed not to be a template element (it will be the html element),
  365. // so no need to check the parent is a template.
  366. return { m_stack_of_open_elements.elements().first(), nullptr };
  367. }
  368. if (last_table.element->parent_node())
  369. adjusted_insertion_location = { last_table.element->parent_node(), last_table.element };
  370. else
  371. adjusted_insertion_location = { m_stack_of_open_elements.element_before(*last_table.element), nullptr };
  372. } else {
  373. adjusted_insertion_location = { target, nullptr };
  374. }
  375. if (is<HTMLTemplateElement>(*adjusted_insertion_location.parent))
  376. return { verify_cast<HTMLTemplateElement>(*adjusted_insertion_location.parent).content(), nullptr };
  377. return adjusted_insertion_location;
  378. }
  379. NonnullRefPtr<DOM::Element> HTMLDocumentParser::create_element_for(const HTMLToken& token, const FlyString& namespace_)
  380. {
  381. auto element = create_element(document(), token.tag_name(), namespace_);
  382. for (auto& attribute : token.m_tag.attributes) {
  383. element->set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  384. }
  385. return element;
  386. }
  387. // https://html.spec.whatwg.org/multipage/parsing.html#insert-a-foreign-element
  388. NonnullRefPtr<DOM::Element> HTMLDocumentParser::insert_foreign_element(const HTMLToken& token, const FlyString& namespace_)
  389. {
  390. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  391. // FIXME: Pass in adjusted_insertion_location.parent as the intended parent.
  392. auto element = create_element_for(token, namespace_);
  393. auto pre_insertion_validity = adjusted_insertion_location.parent->ensure_pre_insertion_validity(element, adjusted_insertion_location.insert_before_sibling);
  394. // NOTE: If it's not possible to insert the element at the adjusted insertion location, the element is simply dropped.
  395. if (!pre_insertion_validity.is_exception()) {
  396. if (!m_parsing_fragment) {
  397. // FIXME: push a new element queue onto element's relevant agent's custom element reactions stack.
  398. }
  399. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling);
  400. if (!m_parsing_fragment) {
  401. // FIXME: pop the element queue from element's relevant agent's custom element reactions stack, and invoke custom element reactions in that queue.
  402. }
  403. }
  404. m_stack_of_open_elements.push(element);
  405. return element;
  406. }
  407. NonnullRefPtr<DOM::Element> HTMLDocumentParser::insert_html_element(const HTMLToken& token)
  408. {
  409. return insert_foreign_element(token, Namespace::HTML);
  410. }
  411. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  412. {
  413. if (token.is_character() && token.is_parser_whitespace()) {
  414. return;
  415. }
  416. if (token.is_comment()) {
  417. insert_comment(token);
  418. return;
  419. }
  420. if (token.is_doctype()) {
  421. log_parse_error();
  422. return;
  423. }
  424. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  425. process_using_the_rules_for(InsertionMode::InBody, token);
  426. return;
  427. }
  428. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  429. auto element = insert_html_element(token);
  430. m_head_element = verify_cast<HTMLHeadElement>(*element);
  431. m_insertion_mode = InsertionMode::InHead;
  432. return;
  433. }
  434. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  435. goto AnythingElse;
  436. }
  437. if (token.is_end_tag()) {
  438. log_parse_error();
  439. return;
  440. }
  441. AnythingElse:
  442. m_head_element = verify_cast<HTMLHeadElement>(*insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::head)));
  443. m_insertion_mode = InsertionMode::InHead;
  444. process_using_the_rules_for(InsertionMode::InHead, token);
  445. return;
  446. }
  447. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  448. {
  449. auto data = token.m_comment_or_character.data.to_string();
  450. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  451. adjusted_insertion_location.parent->insert_before(adopt_ref(*new DOM::Comment(document(), data)), adjusted_insertion_location.insert_before_sibling);
  452. }
  453. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  454. {
  455. if (token.is_parser_whitespace()) {
  456. insert_character(token.code_point());
  457. return;
  458. }
  459. if (token.is_comment()) {
  460. insert_comment(token);
  461. return;
  462. }
  463. if (token.is_doctype()) {
  464. log_parse_error();
  465. return;
  466. }
  467. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  468. process_using_the_rules_for(InsertionMode::InBody, token);
  469. return;
  470. }
  471. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  472. insert_html_element(token);
  473. m_stack_of_open_elements.pop();
  474. token.acknowledge_self_closing_flag_if_set();
  475. return;
  476. }
  477. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  478. auto element = insert_html_element(token);
  479. m_stack_of_open_elements.pop();
  480. token.acknowledge_self_closing_flag_if_set();
  481. return;
  482. }
  483. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  484. insert_html_element(token);
  485. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  486. m_original_insertion_mode = m_insertion_mode;
  487. m_insertion_mode = InsertionMode::Text;
  488. return;
  489. }
  490. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled) || token.tag_name() == HTML::TagNames::noframes || token.tag_name() == HTML::TagNames::style)) {
  491. parse_generic_raw_text_element(token);
  492. return;
  493. }
  494. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noscript && !m_scripting_enabled) {
  495. insert_html_element(token);
  496. m_insertion_mode = InsertionMode::InHeadNoscript;
  497. return;
  498. }
  499. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  500. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  501. auto element = create_element_for(token, Namespace::HTML);
  502. auto& script_element = verify_cast<HTMLScriptElement>(*element);
  503. script_element.set_parser_document({}, document());
  504. script_element.set_non_blocking({}, false);
  505. if (m_parsing_fragment) {
  506. script_element.set_already_started({}, true);
  507. }
  508. if (m_invoked_via_document_write) {
  509. TODO();
  510. }
  511. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling, false);
  512. m_stack_of_open_elements.push(element);
  513. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  514. m_original_insertion_mode = m_insertion_mode;
  515. m_insertion_mode = InsertionMode::Text;
  516. return;
  517. }
  518. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  519. m_stack_of_open_elements.pop();
  520. m_insertion_mode = InsertionMode::AfterHead;
  521. return;
  522. }
  523. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  524. goto AnythingElse;
  525. }
  526. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  527. insert_html_element(token);
  528. m_list_of_active_formatting_elements.add_marker();
  529. m_frameset_ok = false;
  530. m_insertion_mode = InsertionMode::InTemplate;
  531. m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  532. return;
  533. }
  534. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  535. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  536. log_parse_error();
  537. return;
  538. }
  539. generate_all_implied_end_tags_thoroughly();
  540. if (current_node().local_name() != HTML::TagNames::template_)
  541. log_parse_error();
  542. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  543. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  544. m_stack_of_template_insertion_modes.take_last();
  545. reset_the_insertion_mode_appropriately();
  546. return;
  547. }
  548. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  549. log_parse_error();
  550. return;
  551. }
  552. AnythingElse:
  553. m_stack_of_open_elements.pop();
  554. m_insertion_mode = InsertionMode::AfterHead;
  555. process_using_the_rules_for(m_insertion_mode, token);
  556. }
  557. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken& token)
  558. {
  559. if (token.is_doctype()) {
  560. log_parse_error();
  561. return;
  562. }
  563. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  564. process_using_the_rules_for(InsertionMode::InBody, token);
  565. return;
  566. }
  567. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::noscript) {
  568. m_stack_of_open_elements.pop();
  569. m_insertion_mode = InsertionMode::InHead;
  570. return;
  571. }
  572. if (token.is_parser_whitespace() || token.is_comment() || (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::style))) {
  573. process_using_the_rules_for(InsertionMode::InHead, token);
  574. return;
  575. }
  576. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  577. goto AnythingElse;
  578. }
  579. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::noscript)) {
  580. log_parse_error();
  581. return;
  582. }
  583. AnythingElse:
  584. log_parse_error();
  585. m_stack_of_open_elements.pop();
  586. m_insertion_mode = InsertionMode::InHead;
  587. process_using_the_rules_for(m_insertion_mode, token);
  588. }
  589. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  590. {
  591. insert_html_element(token);
  592. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  593. m_original_insertion_mode = m_insertion_mode;
  594. m_insertion_mode = InsertionMode::Text;
  595. }
  596. DOM::Text* HTMLDocumentParser::find_character_insertion_node()
  597. {
  598. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  599. if (adjusted_insertion_location.insert_before_sibling) {
  600. TODO();
  601. }
  602. if (adjusted_insertion_location.parent->is_document())
  603. return nullptr;
  604. if (adjusted_insertion_location.parent->last_child() && adjusted_insertion_location.parent->last_child()->is_text())
  605. return verify_cast<DOM::Text>(adjusted_insertion_location.parent->last_child());
  606. auto new_text_node = adopt_ref(*new DOM::Text(document(), ""));
  607. adjusted_insertion_location.parent->append_child(new_text_node);
  608. return new_text_node;
  609. }
  610. void HTMLDocumentParser::flush_character_insertions()
  611. {
  612. if (m_character_insertion_builder.is_empty())
  613. return;
  614. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  615. m_character_insertion_node->parent()->children_changed();
  616. m_character_insertion_builder.clear();
  617. }
  618. void HTMLDocumentParser::insert_character(u32 data)
  619. {
  620. auto node = find_character_insertion_node();
  621. if (node == m_character_insertion_node) {
  622. m_character_insertion_builder.append(Utf32View { &data, 1 });
  623. return;
  624. }
  625. if (!m_character_insertion_node) {
  626. m_character_insertion_node = node;
  627. m_character_insertion_builder.append(Utf32View { &data, 1 });
  628. return;
  629. }
  630. flush_character_insertions();
  631. m_character_insertion_node = node;
  632. m_character_insertion_builder.append(Utf32View { &data, 1 });
  633. }
  634. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  635. {
  636. if (token.is_character() && token.is_parser_whitespace()) {
  637. insert_character(token.code_point());
  638. return;
  639. }
  640. if (token.is_comment()) {
  641. insert_comment(token);
  642. return;
  643. }
  644. if (token.is_doctype()) {
  645. log_parse_error();
  646. return;
  647. }
  648. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  649. process_using_the_rules_for(InsertionMode::InBody, token);
  650. return;
  651. }
  652. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  653. insert_html_element(token);
  654. m_frameset_ok = false;
  655. m_insertion_mode = InsertionMode::InBody;
  656. return;
  657. }
  658. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  659. insert_html_element(token);
  660. m_insertion_mode = InsertionMode::InFrameset;
  661. return;
  662. }
  663. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  664. log_parse_error();
  665. m_stack_of_open_elements.push(*m_head_element);
  666. process_using_the_rules_for(InsertionMode::InHead, token);
  667. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  668. return entry.ptr() == m_head_element;
  669. });
  670. return;
  671. }
  672. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  673. process_using_the_rules_for(InsertionMode::InHead, token);
  674. return;
  675. }
  676. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  677. goto AnythingElse;
  678. }
  679. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  680. log_parse_error();
  681. return;
  682. }
  683. AnythingElse:
  684. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::body));
  685. m_insertion_mode = InsertionMode::InBody;
  686. process_using_the_rules_for(m_insertion_mode, token);
  687. }
  688. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  689. {
  690. while (current_node().local_name() != exception && current_node().local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc))
  691. m_stack_of_open_elements.pop();
  692. }
  693. void HTMLDocumentParser::generate_all_implied_end_tags_thoroughly()
  694. {
  695. while (current_node().local_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::colgroup, HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))
  696. m_stack_of_open_elements.pop();
  697. }
  698. void HTMLDocumentParser::close_a_p_element()
  699. {
  700. generate_implied_end_tags(HTML::TagNames::p);
  701. if (current_node().local_name() != HTML::TagNames::p) {
  702. log_parse_error();
  703. }
  704. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  705. }
  706. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  707. {
  708. if (token.is_character() && token.is_parser_whitespace()) {
  709. process_using_the_rules_for(InsertionMode::InBody, token);
  710. return;
  711. }
  712. if (token.is_comment()) {
  713. auto data = token.m_comment_or_character.data.to_string();
  714. auto& insertion_location = m_stack_of_open_elements.first();
  715. insertion_location.append_child(adopt_ref(*new DOM::Comment(document(), data)));
  716. return;
  717. }
  718. if (token.is_doctype()) {
  719. log_parse_error();
  720. return;
  721. }
  722. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  723. process_using_the_rules_for(InsertionMode::InBody, token);
  724. return;
  725. }
  726. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  727. if (m_parsing_fragment) {
  728. log_parse_error();
  729. return;
  730. }
  731. m_insertion_mode = InsertionMode::AfterAfterBody;
  732. return;
  733. }
  734. if (token.is_end_of_file()) {
  735. stop_parsing();
  736. return;
  737. }
  738. log_parse_error();
  739. m_insertion_mode = InsertionMode::InBody;
  740. process_using_the_rules_for(InsertionMode::InBody, token);
  741. }
  742. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  743. {
  744. if (token.is_comment()) {
  745. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  746. document().append_child(move(comment));
  747. return;
  748. }
  749. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  750. process_using_the_rules_for(InsertionMode::InBody, token);
  751. return;
  752. }
  753. if (token.is_end_of_file()) {
  754. stop_parsing();
  755. return;
  756. }
  757. log_parse_error();
  758. m_insertion_mode = InsertionMode::InBody;
  759. process_using_the_rules_for(m_insertion_mode, token);
  760. }
  761. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  762. {
  763. // FIXME: This needs to care about "markers"
  764. if (m_list_of_active_formatting_elements.is_empty())
  765. return;
  766. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  767. return;
  768. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  769. return;
  770. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  771. RefPtr<DOM::Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  772. VERIFY(entry);
  773. Rewind:
  774. if (index == 0) {
  775. goto Create;
  776. }
  777. --index;
  778. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  779. VERIFY(entry);
  780. if (!m_stack_of_open_elements.contains(*entry))
  781. goto Rewind;
  782. Advance:
  783. ++index;
  784. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  785. VERIFY(entry);
  786. Create:
  787. // FIXME: Hold on to the real token!
  788. auto new_element = insert_html_element(HTMLToken::make_start_tag(entry->local_name()));
  789. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  790. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  791. goto Advance;
  792. }
  793. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  794. {
  795. auto subject = token.tag_name();
  796. // If the current node is an HTML element whose tag name is subject,
  797. // and the current node is not in the list of active formatting elements,
  798. // then pop the current node off the stack of open elements, and return.
  799. if (current_node().local_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  800. m_stack_of_open_elements.pop();
  801. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  802. }
  803. size_t outer_loop_counter = 0;
  804. //OuterLoop:
  805. if (outer_loop_counter >= 8)
  806. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  807. ++outer_loop_counter;
  808. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  809. if (!formatting_element)
  810. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  811. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  812. log_parse_error();
  813. m_list_of_active_formatting_elements.remove(*formatting_element);
  814. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  815. }
  816. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  817. log_parse_error();
  818. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  819. }
  820. if (formatting_element != &current_node()) {
  821. log_parse_error();
  822. }
  823. RefPtr<DOM::Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  824. if (!furthest_block) {
  825. while (&current_node() != formatting_element)
  826. m_stack_of_open_elements.pop();
  827. m_stack_of_open_elements.pop();
  828. m_list_of_active_formatting_elements.remove(*formatting_element);
  829. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  830. }
  831. // FIXME: Implement the rest of the AAA :^)
  832. TODO();
  833. }
  834. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name, const FlyString& namespace_)
  835. {
  836. if (namespace_ == Namespace::HTML) {
  837. return tag_name.is_one_of(
  838. HTML::TagNames::address,
  839. HTML::TagNames::applet,
  840. HTML::TagNames::area,
  841. HTML::TagNames::article,
  842. HTML::TagNames::aside,
  843. HTML::TagNames::base,
  844. HTML::TagNames::basefont,
  845. HTML::TagNames::bgsound,
  846. HTML::TagNames::blockquote,
  847. HTML::TagNames::body,
  848. HTML::TagNames::br,
  849. HTML::TagNames::button,
  850. HTML::TagNames::caption,
  851. HTML::TagNames::center,
  852. HTML::TagNames::col,
  853. HTML::TagNames::colgroup,
  854. HTML::TagNames::dd,
  855. HTML::TagNames::details,
  856. HTML::TagNames::dir,
  857. HTML::TagNames::div,
  858. HTML::TagNames::dl,
  859. HTML::TagNames::dt,
  860. HTML::TagNames::embed,
  861. HTML::TagNames::fieldset,
  862. HTML::TagNames::figcaption,
  863. HTML::TagNames::figure,
  864. HTML::TagNames::footer,
  865. HTML::TagNames::form,
  866. HTML::TagNames::frame,
  867. HTML::TagNames::frameset,
  868. HTML::TagNames::h1,
  869. HTML::TagNames::h2,
  870. HTML::TagNames::h3,
  871. HTML::TagNames::h4,
  872. HTML::TagNames::h5,
  873. HTML::TagNames::h6,
  874. HTML::TagNames::head,
  875. HTML::TagNames::header,
  876. HTML::TagNames::hgroup,
  877. HTML::TagNames::hr,
  878. HTML::TagNames::html,
  879. HTML::TagNames::iframe,
  880. HTML::TagNames::img,
  881. HTML::TagNames::input,
  882. HTML::TagNames::keygen,
  883. HTML::TagNames::li,
  884. HTML::TagNames::link,
  885. HTML::TagNames::listing,
  886. HTML::TagNames::main,
  887. HTML::TagNames::marquee,
  888. HTML::TagNames::menu,
  889. HTML::TagNames::meta,
  890. HTML::TagNames::nav,
  891. HTML::TagNames::noembed,
  892. HTML::TagNames::noframes,
  893. HTML::TagNames::noscript,
  894. HTML::TagNames::object,
  895. HTML::TagNames::ol,
  896. HTML::TagNames::p,
  897. HTML::TagNames::param,
  898. HTML::TagNames::plaintext,
  899. HTML::TagNames::pre,
  900. HTML::TagNames::script,
  901. HTML::TagNames::section,
  902. HTML::TagNames::select,
  903. HTML::TagNames::source,
  904. HTML::TagNames::style,
  905. HTML::TagNames::summary,
  906. HTML::TagNames::table,
  907. HTML::TagNames::tbody,
  908. HTML::TagNames::td,
  909. HTML::TagNames::template_,
  910. HTML::TagNames::textarea,
  911. HTML::TagNames::tfoot,
  912. HTML::TagNames::th,
  913. HTML::TagNames::thead,
  914. HTML::TagNames::title,
  915. HTML::TagNames::tr,
  916. HTML::TagNames::track,
  917. HTML::TagNames::ul,
  918. HTML::TagNames::wbr,
  919. HTML::TagNames::xmp);
  920. } else if (namespace_ == Namespace::SVG) {
  921. return tag_name.is_one_of(
  922. SVG::TagNames::desc,
  923. SVG::TagNames::foreignObject,
  924. SVG::TagNames::title);
  925. } else if (namespace_ == Namespace::MathML) {
  926. TODO();
  927. }
  928. return false;
  929. }
  930. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  931. {
  932. if (token.is_character()) {
  933. if (token.code_point() == 0) {
  934. log_parse_error();
  935. return;
  936. }
  937. if (token.is_parser_whitespace()) {
  938. reconstruct_the_active_formatting_elements();
  939. insert_character(token.code_point());
  940. return;
  941. }
  942. reconstruct_the_active_formatting_elements();
  943. insert_character(token.code_point());
  944. m_frameset_ok = false;
  945. return;
  946. }
  947. if (token.is_comment()) {
  948. insert_comment(token);
  949. return;
  950. }
  951. if (token.is_doctype()) {
  952. log_parse_error();
  953. return;
  954. }
  955. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  956. log_parse_error();
  957. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  958. return;
  959. for (auto& attribute : token.m_tag.attributes) {
  960. if (current_node().has_attribute(attribute.local_name_builder.string_view()))
  961. continue;
  962. current_node().set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  963. }
  964. return;
  965. }
  966. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  967. process_using_the_rules_for(InsertionMode::InHead, token);
  968. return;
  969. }
  970. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  971. process_using_the_rules_for(InsertionMode::InHead, token);
  972. return;
  973. }
  974. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  975. log_parse_error();
  976. if (m_stack_of_open_elements.elements().size() == 1
  977. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body
  978. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  979. VERIFY(m_parsing_fragment);
  980. return;
  981. }
  982. m_frameset_ok = false;
  983. auto& body_element = m_stack_of_open_elements.elements().at(1);
  984. for (auto& attribute : token.m_tag.attributes) {
  985. if (body_element.has_attribute(attribute.local_name_builder.string_view()))
  986. continue;
  987. body_element.set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  988. }
  989. return;
  990. }
  991. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  992. log_parse_error();
  993. if (m_stack_of_open_elements.elements().size() == 1
  994. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body) {
  995. VERIFY(m_parsing_fragment);
  996. return;
  997. }
  998. if (!m_frameset_ok)
  999. return;
  1000. TODO();
  1001. }
  1002. if (token.is_end_of_file()) {
  1003. if (!m_stack_of_template_insertion_modes.is_empty()) {
  1004. process_using_the_rules_for(InsertionMode::InTemplate, token);
  1005. return;
  1006. }
  1007. for (auto& node : m_stack_of_open_elements.elements()) {
  1008. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1009. log_parse_error();
  1010. break;
  1011. }
  1012. }
  1013. stop_parsing();
  1014. return;
  1015. }
  1016. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  1017. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1018. log_parse_error();
  1019. return;
  1020. }
  1021. for (auto& node : m_stack_of_open_elements.elements()) {
  1022. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1023. log_parse_error();
  1024. break;
  1025. }
  1026. }
  1027. m_insertion_mode = InsertionMode::AfterBody;
  1028. return;
  1029. }
  1030. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  1031. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1032. log_parse_error();
  1033. return;
  1034. }
  1035. for (auto& node : m_stack_of_open_elements.elements()) {
  1036. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1037. log_parse_error();
  1038. break;
  1039. }
  1040. }
  1041. m_insertion_mode = InsertionMode::AfterBody;
  1042. process_using_the_rules_for(m_insertion_mode, token);
  1043. return;
  1044. }
  1045. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::article, HTML::TagNames::aside, HTML::TagNames::blockquote, HTML::TagNames::center, HTML::TagNames::details, HTML::TagNames::dialog, HTML::TagNames::dir, HTML::TagNames::div, HTML::TagNames::dl, HTML::TagNames::fieldset, HTML::TagNames::figcaption, HTML::TagNames::figure, HTML::TagNames::footer, HTML::TagNames::header, HTML::TagNames::hgroup, HTML::TagNames::main, HTML::TagNames::menu, HTML::TagNames::nav, HTML::TagNames::ol, HTML::TagNames::p, HTML::TagNames::section, HTML::TagNames::summary, HTML::TagNames::ul)) {
  1046. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1047. close_a_p_element();
  1048. insert_html_element(token);
  1049. return;
  1050. }
  1051. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  1052. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1053. close_a_p_element();
  1054. if (current_node().local_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  1055. log_parse_error();
  1056. m_stack_of_open_elements.pop();
  1057. }
  1058. insert_html_element(token);
  1059. return;
  1060. }
  1061. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  1062. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1063. close_a_p_element();
  1064. insert_html_element(token);
  1065. m_frameset_ok = false;
  1066. // If the next token is a U+000A LINE FEED (LF) character token,
  1067. // then ignore that token and move on to the next one.
  1068. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1069. auto next_token = m_tokenizer.next_token();
  1070. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1071. // Ignore it.
  1072. } else {
  1073. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1074. }
  1075. return;
  1076. }
  1077. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1078. if (m_form_element && !m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1079. log_parse_error();
  1080. return;
  1081. }
  1082. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1083. close_a_p_element();
  1084. auto element = insert_html_element(token);
  1085. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  1086. m_form_element = verify_cast<HTMLFormElement>(*element);
  1087. return;
  1088. }
  1089. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  1090. m_frameset_ok = false;
  1091. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1092. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1093. if (node->local_name() == HTML::TagNames::li) {
  1094. generate_implied_end_tags(HTML::TagNames::li);
  1095. if (current_node().local_name() != HTML::TagNames::li) {
  1096. log_parse_error();
  1097. }
  1098. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1099. break;
  1100. }
  1101. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1102. break;
  1103. }
  1104. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1105. close_a_p_element();
  1106. insert_html_element(token);
  1107. return;
  1108. }
  1109. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1110. m_frameset_ok = false;
  1111. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1112. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1113. if (node->local_name() == HTML::TagNames::dd) {
  1114. generate_implied_end_tags(HTML::TagNames::dd);
  1115. if (current_node().local_name() != HTML::TagNames::dd) {
  1116. log_parse_error();
  1117. }
  1118. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  1119. break;
  1120. }
  1121. if (node->local_name() == HTML::TagNames::dt) {
  1122. generate_implied_end_tags(HTML::TagNames::dt);
  1123. if (current_node().local_name() != HTML::TagNames::dt) {
  1124. log_parse_error();
  1125. }
  1126. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  1127. break;
  1128. }
  1129. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1130. break;
  1131. }
  1132. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1133. close_a_p_element();
  1134. insert_html_element(token);
  1135. return;
  1136. }
  1137. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  1138. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1139. close_a_p_element();
  1140. insert_html_element(token);
  1141. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  1142. return;
  1143. }
  1144. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  1145. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  1146. log_parse_error();
  1147. generate_implied_end_tags();
  1148. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  1149. }
  1150. reconstruct_the_active_formatting_elements();
  1151. insert_html_element(token);
  1152. m_frameset_ok = false;
  1153. return;
  1154. }
  1155. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::article, HTML::TagNames::aside, HTML::TagNames::blockquote, HTML::TagNames::button, HTML::TagNames::center, HTML::TagNames::details, HTML::TagNames::dialog, HTML::TagNames::dir, HTML::TagNames::div, HTML::TagNames::dl, HTML::TagNames::fieldset, HTML::TagNames::figcaption, HTML::TagNames::figure, HTML::TagNames::footer, HTML::TagNames::header, HTML::TagNames::hgroup, HTML::TagNames::listing, HTML::TagNames::main, HTML::TagNames::menu, HTML::TagNames::nav, HTML::TagNames::ol, HTML::TagNames::pre, HTML::TagNames::section, HTML::TagNames::summary, HTML::TagNames::ul)) {
  1156. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1157. log_parse_error();
  1158. return;
  1159. }
  1160. generate_implied_end_tags();
  1161. if (current_node().local_name() != token.tag_name()) {
  1162. log_parse_error();
  1163. }
  1164. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1165. return;
  1166. }
  1167. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  1168. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1169. auto node = m_form_element;
  1170. m_form_element = nullptr;
  1171. if (!node || !m_stack_of_open_elements.has_in_scope(*node)) {
  1172. log_parse_error();
  1173. return;
  1174. }
  1175. generate_implied_end_tags();
  1176. if (&current_node() != node) {
  1177. log_parse_error();
  1178. }
  1179. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  1180. } else {
  1181. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  1182. log_parse_error();
  1183. return;
  1184. }
  1185. generate_implied_end_tags();
  1186. if (current_node().local_name() != HTML::TagNames::form) {
  1187. log_parse_error();
  1188. }
  1189. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  1190. }
  1191. return;
  1192. }
  1193. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  1194. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1195. log_parse_error();
  1196. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::p));
  1197. }
  1198. close_a_p_element();
  1199. return;
  1200. }
  1201. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  1202. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  1203. log_parse_error();
  1204. return;
  1205. }
  1206. generate_implied_end_tags(HTML::TagNames::li);
  1207. if (current_node().local_name() != HTML::TagNames::li) {
  1208. log_parse_error();
  1209. dbgln("Expected <li> current node, but had <{}>", current_node().local_name());
  1210. }
  1211. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1212. return;
  1213. }
  1214. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1215. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1216. log_parse_error();
  1217. return;
  1218. }
  1219. generate_implied_end_tags(token.tag_name());
  1220. if (current_node().local_name() != token.tag_name()) {
  1221. log_parse_error();
  1222. }
  1223. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1224. return;
  1225. }
  1226. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  1227. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  1228. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  1229. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  1230. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  1231. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  1232. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  1233. log_parse_error();
  1234. return;
  1235. }
  1236. generate_implied_end_tags();
  1237. if (current_node().local_name() != token.tag_name()) {
  1238. log_parse_error();
  1239. }
  1240. for (;;) {
  1241. auto popped_element = m_stack_of_open_elements.pop();
  1242. if (popped_element->local_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6))
  1243. break;
  1244. }
  1245. return;
  1246. }
  1247. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1248. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1249. log_parse_error();
  1250. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1251. goto AnyOtherEndTag;
  1252. m_list_of_active_formatting_elements.remove(*element);
  1253. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1254. return entry.ptr() == element;
  1255. });
  1256. }
  1257. reconstruct_the_active_formatting_elements();
  1258. auto element = insert_html_element(token);
  1259. m_list_of_active_formatting_elements.add(*element);
  1260. return;
  1261. }
  1262. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::b, HTML::TagNames::big, HTML::TagNames::code, HTML::TagNames::em, HTML::TagNames::font, HTML::TagNames::i, HTML::TagNames::s, HTML::TagNames::small, HTML::TagNames::strike, HTML::TagNames::strong, HTML::TagNames::tt, HTML::TagNames::u)) {
  1263. reconstruct_the_active_formatting_elements();
  1264. auto element = insert_html_element(token);
  1265. m_list_of_active_formatting_elements.add(*element);
  1266. return;
  1267. }
  1268. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1269. reconstruct_the_active_formatting_elements();
  1270. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1271. log_parse_error();
  1272. run_the_adoption_agency_algorithm(token);
  1273. reconstruct_the_active_formatting_elements();
  1274. }
  1275. auto element = insert_html_element(token);
  1276. m_list_of_active_formatting_elements.add(*element);
  1277. return;
  1278. }
  1279. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::a, HTML::TagNames::b, HTML::TagNames::big, HTML::TagNames::code, HTML::TagNames::em, HTML::TagNames::font, HTML::TagNames::i, HTML::TagNames::nobr, HTML::TagNames::s, HTML::TagNames::small, HTML::TagNames::strike, HTML::TagNames::strong, HTML::TagNames::tt, HTML::TagNames::u)) {
  1280. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1281. goto AnyOtherEndTag;
  1282. return;
  1283. }
  1284. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1285. reconstruct_the_active_formatting_elements();
  1286. insert_html_element(token);
  1287. m_list_of_active_formatting_elements.add_marker();
  1288. m_frameset_ok = false;
  1289. return;
  1290. }
  1291. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1292. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1293. log_parse_error();
  1294. return;
  1295. }
  1296. generate_implied_end_tags();
  1297. if (current_node().local_name() != token.tag_name()) {
  1298. log_parse_error();
  1299. }
  1300. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1301. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1302. return;
  1303. }
  1304. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1305. if (!document().in_quirks_mode()) {
  1306. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1307. close_a_p_element();
  1308. }
  1309. insert_html_element(token);
  1310. m_frameset_ok = false;
  1311. m_insertion_mode = InsertionMode::InTable;
  1312. return;
  1313. }
  1314. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1315. token.drop_attributes();
  1316. goto BRStartTag;
  1317. }
  1318. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::area, HTML::TagNames::br, HTML::TagNames::embed, HTML::TagNames::img, HTML::TagNames::keygen, HTML::TagNames::wbr)) {
  1319. BRStartTag:
  1320. reconstruct_the_active_formatting_elements();
  1321. insert_html_element(token);
  1322. m_stack_of_open_elements.pop();
  1323. token.acknowledge_self_closing_flag_if_set();
  1324. m_frameset_ok = false;
  1325. return;
  1326. }
  1327. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1328. reconstruct_the_active_formatting_elements();
  1329. insert_html_element(token);
  1330. m_stack_of_open_elements.pop();
  1331. token.acknowledge_self_closing_flag_if_set();
  1332. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1333. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1334. m_frameset_ok = false;
  1335. }
  1336. return;
  1337. }
  1338. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1339. insert_html_element(token);
  1340. m_stack_of_open_elements.pop();
  1341. token.acknowledge_self_closing_flag_if_set();
  1342. return;
  1343. }
  1344. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1345. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1346. close_a_p_element();
  1347. insert_html_element(token);
  1348. m_stack_of_open_elements.pop();
  1349. token.acknowledge_self_closing_flag_if_set();
  1350. m_frameset_ok = false;
  1351. return;
  1352. }
  1353. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::image) {
  1354. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1355. log_parse_error();
  1356. token.m_tag.tag_name.clear();
  1357. token.m_tag.tag_name.append(HTML::TagNames::img);
  1358. process_using_the_rules_for(m_insertion_mode, token);
  1359. return;
  1360. }
  1361. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1362. insert_html_element(token);
  1363. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1364. // If the next token is a U+000A LINE FEED (LF) character token,
  1365. // then ignore that token and move on to the next one.
  1366. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1367. auto next_token = m_tokenizer.next_token();
  1368. m_original_insertion_mode = m_insertion_mode;
  1369. m_frameset_ok = false;
  1370. m_insertion_mode = InsertionMode::Text;
  1371. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1372. // Ignore it.
  1373. } else {
  1374. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1375. }
  1376. return;
  1377. }
  1378. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1379. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1380. close_a_p_element();
  1381. }
  1382. reconstruct_the_active_formatting_elements();
  1383. m_frameset_ok = false;
  1384. parse_generic_raw_text_element(token);
  1385. return;
  1386. }
  1387. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1388. m_frameset_ok = false;
  1389. parse_generic_raw_text_element(token);
  1390. return;
  1391. }
  1392. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1393. parse_generic_raw_text_element(token);
  1394. return;
  1395. }
  1396. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1397. reconstruct_the_active_formatting_elements();
  1398. insert_html_element(token);
  1399. m_frameset_ok = false;
  1400. switch (m_insertion_mode) {
  1401. case InsertionMode::InTable:
  1402. case InsertionMode::InCaption:
  1403. case InsertionMode::InTableBody:
  1404. case InsertionMode::InRow:
  1405. case InsertionMode::InCell:
  1406. m_insertion_mode = InsertionMode::InSelectInTable;
  1407. break;
  1408. default:
  1409. m_insertion_mode = InsertionMode::InSelect;
  1410. break;
  1411. }
  1412. return;
  1413. }
  1414. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1415. if (current_node().local_name() == HTML::TagNames::option)
  1416. m_stack_of_open_elements.pop();
  1417. reconstruct_the_active_formatting_elements();
  1418. insert_html_element(token);
  1419. return;
  1420. }
  1421. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1422. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1423. generate_implied_end_tags();
  1424. if (current_node().local_name() != HTML::TagNames::ruby)
  1425. log_parse_error();
  1426. insert_html_element(token);
  1427. return;
  1428. }
  1429. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1430. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1431. generate_implied_end_tags(HTML::TagNames::rtc);
  1432. if (current_node().local_name() != HTML::TagNames::rtc || current_node().local_name() != HTML::TagNames::ruby)
  1433. log_parse_error();
  1434. insert_html_element(token);
  1435. return;
  1436. }
  1437. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::math) {
  1438. reconstruct_the_active_formatting_elements();
  1439. adjust_mathml_attributes(token);
  1440. adjust_foreign_attributes(token);
  1441. insert_foreign_element(token, Namespace::MathML);
  1442. if (token.is_self_closing()) {
  1443. m_stack_of_open_elements.pop();
  1444. token.acknowledge_self_closing_flag_if_set();
  1445. }
  1446. return;
  1447. }
  1448. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::svg) {
  1449. reconstruct_the_active_formatting_elements();
  1450. adjust_svg_attributes(token);
  1451. adjust_foreign_attributes(token);
  1452. insert_foreign_element(token, Namespace::SVG);
  1453. if (token.is_self_closing()) {
  1454. m_stack_of_open_elements.pop();
  1455. token.acknowledge_self_closing_flag_if_set();
  1456. }
  1457. return;
  1458. }
  1459. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::frame, HTML::TagNames::head, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))) {
  1460. log_parse_error();
  1461. return;
  1462. }
  1463. // Any other start tag
  1464. if (token.is_start_tag()) {
  1465. reconstruct_the_active_formatting_elements();
  1466. insert_html_element(token);
  1467. return;
  1468. }
  1469. if (token.is_end_tag()) {
  1470. AnyOtherEndTag:
  1471. RefPtr<DOM::Element> node;
  1472. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1473. node = m_stack_of_open_elements.elements()[i];
  1474. if (node->local_name() == token.tag_name()) {
  1475. generate_implied_end_tags(token.tag_name());
  1476. if (node != current_node()) {
  1477. log_parse_error();
  1478. }
  1479. while (&current_node() != node) {
  1480. m_stack_of_open_elements.pop();
  1481. }
  1482. m_stack_of_open_elements.pop();
  1483. break;
  1484. }
  1485. if (is_special_tag(node->local_name(), node->namespace_())) {
  1486. log_parse_error();
  1487. return;
  1488. }
  1489. }
  1490. return;
  1491. }
  1492. }
  1493. void HTMLDocumentParser::adjust_mathml_attributes(HTMLToken& token)
  1494. {
  1495. token.adjust_attribute_name("definitionurl", "definitionURL");
  1496. }
  1497. void HTMLDocumentParser::adjust_svg_tag_names(HTMLToken& token)
  1498. {
  1499. token.adjust_tag_name("altglyph", "altGlyph");
  1500. token.adjust_tag_name("altglyphdef", "altGlyphDef");
  1501. token.adjust_tag_name("altglyphitem", "altGlyphItem");
  1502. token.adjust_tag_name("animatecolor", "animateColor");
  1503. token.adjust_tag_name("animatemotion", "animateMotion");
  1504. token.adjust_tag_name("animatetransform", "animateTransform");
  1505. token.adjust_tag_name("clippath", "clipPath");
  1506. token.adjust_tag_name("feblend", "feBlend");
  1507. token.adjust_tag_name("fecolormatrix", "feColorMatrix");
  1508. token.adjust_tag_name("fecomponenttransfer", "feComponentTransfer");
  1509. token.adjust_tag_name("fecomposite", "feComposite");
  1510. token.adjust_tag_name("feconvolvematrix", "feConvolveMatrix");
  1511. token.adjust_tag_name("fediffuselighting", "feDiffuseLighting");
  1512. token.adjust_tag_name("fedisplacementmap", "feDisplacementMap");
  1513. token.adjust_tag_name("fedistantlight", "feDistantLight");
  1514. token.adjust_tag_name("fedropshadow", "feDropShadow");
  1515. token.adjust_tag_name("feflood", "feFlood");
  1516. token.adjust_tag_name("fefunca", "feFuncA");
  1517. token.adjust_tag_name("fefuncb", "feFuncB");
  1518. token.adjust_tag_name("fefuncg", "feFuncG");
  1519. token.adjust_tag_name("fefuncr", "feFuncR");
  1520. token.adjust_tag_name("fegaussianblur", "feGaussianBlur");
  1521. token.adjust_tag_name("feimage", "feImage");
  1522. token.adjust_tag_name("femerge", "feMerge");
  1523. token.adjust_tag_name("femergenode", "feMergeNode");
  1524. token.adjust_tag_name("femorphology", "feMorphology");
  1525. token.adjust_tag_name("feoffset", "feOffset");
  1526. token.adjust_tag_name("fepointlight", "fePointLight");
  1527. token.adjust_tag_name("fespecularlighting", "feSpecularLighting");
  1528. token.adjust_tag_name("fespotlight", "feSpotlight");
  1529. token.adjust_tag_name("glyphref", "glyphRef");
  1530. token.adjust_tag_name("lineargradient", "linearGradient");
  1531. token.adjust_tag_name("radialgradient", "radialGradient");
  1532. token.adjust_tag_name("textpath", "textPath");
  1533. }
  1534. void HTMLDocumentParser::adjust_svg_attributes(HTMLToken& token)
  1535. {
  1536. token.adjust_attribute_name("attributename", "attributeName");
  1537. token.adjust_attribute_name("attributetype", "attributeType");
  1538. token.adjust_attribute_name("basefrequency", "baseFrequency");
  1539. token.adjust_attribute_name("baseprofile", "baseProfile");
  1540. token.adjust_attribute_name("calcmode", "calcMode");
  1541. token.adjust_attribute_name("clippathunits", "clipPathUnits");
  1542. token.adjust_attribute_name("diffuseconstant", "diffuseConstant");
  1543. token.adjust_attribute_name("edgemode", "edgeMode");
  1544. token.adjust_attribute_name("filterunits", "filterUnits");
  1545. token.adjust_attribute_name("glyphref", "glyphRef");
  1546. token.adjust_attribute_name("gradienttransform", "gradientTransform");
  1547. token.adjust_attribute_name("gradientunits", "gradientUnits");
  1548. token.adjust_attribute_name("kernelmatrix", "kernelMatrix");
  1549. token.adjust_attribute_name("kernelunitlength", "kernelUnitLength");
  1550. token.adjust_attribute_name("keypoints", "keyPoints");
  1551. token.adjust_attribute_name("keysplines", "keySplines");
  1552. token.adjust_attribute_name("keytimes", "keyTimes");
  1553. token.adjust_attribute_name("lengthadjust", "lengthAdjust");
  1554. token.adjust_attribute_name("limitingconeangle", "limitingConeAngle");
  1555. token.adjust_attribute_name("markerheight", "markerHeight");
  1556. token.adjust_attribute_name("markerunits", "markerUnits");
  1557. token.adjust_attribute_name("markerwidth", "markerWidth");
  1558. token.adjust_attribute_name("maskcontentunits", "maskContentUnits");
  1559. token.adjust_attribute_name("maskunits", "maskUnits");
  1560. token.adjust_attribute_name("numoctaves", "numOctaves");
  1561. token.adjust_attribute_name("pathlength", "pathLength");
  1562. token.adjust_attribute_name("patterncontentunits", "patternContentUnits");
  1563. token.adjust_attribute_name("patterntransform", "patternTransform");
  1564. token.adjust_attribute_name("patternunits", "patternUnits");
  1565. token.adjust_attribute_name("pointsatx", "pointsAtX");
  1566. token.adjust_attribute_name("pointsaty", "pointsAtY");
  1567. token.adjust_attribute_name("pointsatz", "pointsAtZ");
  1568. token.adjust_attribute_name("preservealpha", "preserveAlpha");
  1569. token.adjust_attribute_name("preserveaspectratio", "preserveAspectRatio");
  1570. token.adjust_attribute_name("primitiveunits", "primitiveUnits");
  1571. token.adjust_attribute_name("refx", "refX");
  1572. token.adjust_attribute_name("refy", "refY");
  1573. token.adjust_attribute_name("repeatcount", "repeatCount");
  1574. token.adjust_attribute_name("repeatdur", "repeatDur");
  1575. token.adjust_attribute_name("requiredextensions", "requiredExtensions");
  1576. token.adjust_attribute_name("requiredfeatures", "requiredFeatures");
  1577. token.adjust_attribute_name("specularconstant", "specularConstant");
  1578. token.adjust_attribute_name("specularexponent", "specularExponent");
  1579. token.adjust_attribute_name("spreadmethod", "spreadMethod");
  1580. token.adjust_attribute_name("startoffset", "startOffset");
  1581. token.adjust_attribute_name("stddeviation", "stdDeviation");
  1582. token.adjust_attribute_name("stitchtiles", "stitchTiles");
  1583. token.adjust_attribute_name("surfacescale", "surfaceScale");
  1584. token.adjust_attribute_name("systemlanguage", "systemLanguage");
  1585. token.adjust_attribute_name("tablevalues", "tableValues");
  1586. token.adjust_attribute_name("targetx", "targetX");
  1587. token.adjust_attribute_name("targety", "targetY");
  1588. token.adjust_attribute_name("textlength", "textLength");
  1589. token.adjust_attribute_name("viewbox", "viewBox");
  1590. token.adjust_attribute_name("viewtarget", "viewTarget");
  1591. token.adjust_attribute_name("xchannelselector", "xChannelSelector");
  1592. token.adjust_attribute_name("ychannelselector", "yChannelSelector");
  1593. token.adjust_attribute_name("zoomandpan", "zoomAndPan");
  1594. }
  1595. void HTMLDocumentParser::adjust_foreign_attributes(HTMLToken& token)
  1596. {
  1597. token.adjust_foreign_attribute("xlink:actuate", "xlink", "actuate", Namespace::XLink);
  1598. token.adjust_foreign_attribute("xlink:arcrole", "xlink", "arcrole", Namespace::XLink);
  1599. token.adjust_foreign_attribute("xlink:href", "xlink", "href", Namespace::XLink);
  1600. token.adjust_foreign_attribute("xlink:role", "xlink", "role", Namespace::XLink);
  1601. token.adjust_foreign_attribute("xlink:show", "xlink", "show", Namespace::XLink);
  1602. token.adjust_foreign_attribute("xlink:title", "xlink", "title", Namespace::XLink);
  1603. token.adjust_foreign_attribute("xlink:type", "xlink", "type", Namespace::XLink);
  1604. token.adjust_foreign_attribute("xml:lang", "xml", "lang", Namespace::XML);
  1605. token.adjust_foreign_attribute("xml:space", "xml", "space", Namespace::XML);
  1606. token.adjust_foreign_attribute("xmlns", "", "xmlns", Namespace::XMLNS);
  1607. token.adjust_foreign_attribute("xmlns:xlink", "xmlns", "xlink", Namespace::XMLNS);
  1608. }
  1609. void HTMLDocumentParser::increment_script_nesting_level()
  1610. {
  1611. ++m_script_nesting_level;
  1612. }
  1613. void HTMLDocumentParser::decrement_script_nesting_level()
  1614. {
  1615. VERIFY(m_script_nesting_level);
  1616. --m_script_nesting_level;
  1617. }
  1618. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1619. {
  1620. if (token.is_character()) {
  1621. insert_character(token.code_point());
  1622. return;
  1623. }
  1624. if (token.is_end_of_file()) {
  1625. log_parse_error();
  1626. if (current_node().local_name() == HTML::TagNames::script)
  1627. verify_cast<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1628. m_stack_of_open_elements.pop();
  1629. m_insertion_mode = m_original_insertion_mode;
  1630. process_using_the_rules_for(m_insertion_mode, token);
  1631. return;
  1632. }
  1633. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1634. // Make sure the <script> element has up-to-date text content before preparing the script.
  1635. flush_character_insertions();
  1636. NonnullRefPtr<HTMLScriptElement> script = verify_cast<HTMLScriptElement>(current_node());
  1637. m_stack_of_open_elements.pop();
  1638. m_insertion_mode = m_original_insertion_mode;
  1639. // FIXME: Handle tokenizer insertion point stuff here.
  1640. increment_script_nesting_level();
  1641. script->prepare_script({});
  1642. decrement_script_nesting_level();
  1643. if (script_nesting_level() == 0)
  1644. m_parser_pause_flag = false;
  1645. // FIXME: Handle tokenizer insertion point stuff here too.
  1646. while (document().pending_parsing_blocking_script()) {
  1647. if (script_nesting_level() != 0) {
  1648. m_parser_pause_flag = true;
  1649. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1650. // yielding control back to the caller. (Tokenization will resume when
  1651. // the caller returns to the "outer" tree construction stage.)
  1652. TODO();
  1653. } else {
  1654. auto the_script = document().take_pending_parsing_blocking_script({});
  1655. m_tokenizer.set_blocked(true);
  1656. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1657. // or the script's "ready to be parser-executed" flag is not set:
  1658. // spin the event loop until the parser's Document has no style sheet
  1659. // that is blocking scripts and the script's "ready to be parser-executed"
  1660. // flag is set.
  1661. if (the_script->failed_to_load())
  1662. return;
  1663. VERIFY(the_script->is_ready_to_be_parser_executed());
  1664. if (m_aborted)
  1665. return;
  1666. m_tokenizer.set_blocked(false);
  1667. // FIXME: Handle tokenizer insertion point stuff here too.
  1668. VERIFY(script_nesting_level() == 0);
  1669. increment_script_nesting_level();
  1670. the_script->execute_script();
  1671. decrement_script_nesting_level();
  1672. VERIFY(script_nesting_level() == 0);
  1673. m_parser_pause_flag = false;
  1674. // FIXME: Handle tokenizer insertion point stuff here too.
  1675. }
  1676. }
  1677. return;
  1678. }
  1679. if (token.is_end_tag()) {
  1680. m_stack_of_open_elements.pop();
  1681. m_insertion_mode = m_original_insertion_mode;
  1682. return;
  1683. }
  1684. TODO();
  1685. }
  1686. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1687. {
  1688. while (!current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1689. m_stack_of_open_elements.pop();
  1690. if (current_node().local_name() == HTML::TagNames::html)
  1691. VERIFY(m_parsing_fragment);
  1692. }
  1693. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1694. {
  1695. while (!current_node().local_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1696. m_stack_of_open_elements.pop();
  1697. if (current_node().local_name() == HTML::TagNames::html)
  1698. VERIFY(m_parsing_fragment);
  1699. }
  1700. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1701. {
  1702. while (!current_node().local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1703. m_stack_of_open_elements.pop();
  1704. if (current_node().local_name() == HTML::TagNames::html)
  1705. VERIFY(m_parsing_fragment);
  1706. }
  1707. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1708. {
  1709. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1710. clear_the_stack_back_to_a_table_row_context();
  1711. insert_html_element(token);
  1712. m_insertion_mode = InsertionMode::InCell;
  1713. m_list_of_active_formatting_elements.add_marker();
  1714. return;
  1715. }
  1716. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1717. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1718. log_parse_error();
  1719. return;
  1720. }
  1721. clear_the_stack_back_to_a_table_row_context();
  1722. m_stack_of_open_elements.pop();
  1723. m_insertion_mode = InsertionMode::InTableBody;
  1724. return;
  1725. }
  1726. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr))
  1727. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1728. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1729. log_parse_error();
  1730. return;
  1731. }
  1732. clear_the_stack_back_to_a_table_row_context();
  1733. m_stack_of_open_elements.pop();
  1734. m_insertion_mode = InsertionMode::InTableBody;
  1735. process_using_the_rules_for(m_insertion_mode, token);
  1736. return;
  1737. }
  1738. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1739. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1740. log_parse_error();
  1741. return;
  1742. }
  1743. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1744. return;
  1745. }
  1746. clear_the_stack_back_to_a_table_row_context();
  1747. m_stack_of_open_elements.pop();
  1748. m_insertion_mode = InsertionMode::InTableBody;
  1749. process_using_the_rules_for(m_insertion_mode, token);
  1750. return;
  1751. }
  1752. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::td, HTML::TagNames::th)) {
  1753. log_parse_error();
  1754. return;
  1755. }
  1756. process_using_the_rules_for(InsertionMode::InTable, token);
  1757. }
  1758. void HTMLDocumentParser::close_the_cell()
  1759. {
  1760. generate_implied_end_tags();
  1761. if (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1762. log_parse_error();
  1763. }
  1764. while (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1765. m_stack_of_open_elements.pop();
  1766. m_stack_of_open_elements.pop();
  1767. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1768. m_insertion_mode = InsertionMode::InRow;
  1769. }
  1770. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1771. {
  1772. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1773. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1774. log_parse_error();
  1775. return;
  1776. }
  1777. generate_implied_end_tags();
  1778. if (current_node().local_name() != token.tag_name()) {
  1779. log_parse_error();
  1780. }
  1781. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1782. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1783. m_insertion_mode = InsertionMode::InRow;
  1784. return;
  1785. }
  1786. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1787. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::td) && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::th)) {
  1788. VERIFY(m_parsing_fragment);
  1789. log_parse_error();
  1790. return;
  1791. }
  1792. close_the_cell();
  1793. process_using_the_rules_for(m_insertion_mode, token);
  1794. return;
  1795. }
  1796. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html)) {
  1797. log_parse_error();
  1798. return;
  1799. }
  1800. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1801. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1802. log_parse_error();
  1803. return;
  1804. }
  1805. close_the_cell();
  1806. // Reprocess the token.
  1807. process_using_the_rules_for(m_insertion_mode, token);
  1808. return;
  1809. }
  1810. process_using_the_rules_for(InsertionMode::InBody, token);
  1811. }
  1812. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1813. {
  1814. if (token.is_character()) {
  1815. if (token.code_point() == 0) {
  1816. log_parse_error();
  1817. return;
  1818. }
  1819. m_pending_table_character_tokens.append(token);
  1820. return;
  1821. }
  1822. for (auto& pending_token : m_pending_table_character_tokens) {
  1823. VERIFY(pending_token.is_character());
  1824. if (!pending_token.is_parser_whitespace()) {
  1825. // If any of the tokens in the pending table character tokens list
  1826. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1827. // reprocess the character tokens in the pending table character tokens list using
  1828. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1829. log_parse_error();
  1830. m_foster_parenting = true;
  1831. process_using_the_rules_for(InsertionMode::InBody, token);
  1832. m_foster_parenting = false;
  1833. return;
  1834. }
  1835. }
  1836. for (auto& pending_token : m_pending_table_character_tokens) {
  1837. insert_character(pending_token.code_point());
  1838. }
  1839. m_insertion_mode = m_original_insertion_mode;
  1840. process_using_the_rules_for(m_insertion_mode, token);
  1841. }
  1842. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1843. {
  1844. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1845. clear_the_stack_back_to_a_table_body_context();
  1846. insert_html_element(token);
  1847. m_insertion_mode = InsertionMode::InRow;
  1848. return;
  1849. }
  1850. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1851. log_parse_error();
  1852. clear_the_stack_back_to_a_table_body_context();
  1853. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tr));
  1854. m_insertion_mode = InsertionMode::InRow;
  1855. process_using_the_rules_for(m_insertion_mode, token);
  1856. return;
  1857. }
  1858. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1859. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1860. log_parse_error();
  1861. return;
  1862. }
  1863. clear_the_stack_back_to_a_table_body_context();
  1864. m_stack_of_open_elements.pop();
  1865. m_insertion_mode = InsertionMode::InTable;
  1866. return;
  1867. }
  1868. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead))
  1869. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1870. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tbody)
  1871. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::thead)
  1872. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tfoot)) {
  1873. log_parse_error();
  1874. return;
  1875. }
  1876. clear_the_stack_back_to_a_table_body_context();
  1877. m_stack_of_open_elements.pop();
  1878. m_insertion_mode = InsertionMode::InTable;
  1879. process_using_the_rules_for(InsertionMode::InTable, token);
  1880. return;
  1881. }
  1882. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1883. log_parse_error();
  1884. return;
  1885. }
  1886. process_using_the_rules_for(InsertionMode::InTable, token);
  1887. }
  1888. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1889. {
  1890. if (token.is_character() && current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1891. m_pending_table_character_tokens.clear();
  1892. m_original_insertion_mode = m_insertion_mode;
  1893. m_insertion_mode = InsertionMode::InTableText;
  1894. process_using_the_rules_for(InsertionMode::InTableText, token);
  1895. return;
  1896. }
  1897. if (token.is_comment()) {
  1898. insert_comment(token);
  1899. return;
  1900. }
  1901. if (token.is_doctype()) {
  1902. log_parse_error();
  1903. return;
  1904. }
  1905. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1906. clear_the_stack_back_to_a_table_context();
  1907. m_list_of_active_formatting_elements.add_marker();
  1908. insert_html_element(token);
  1909. m_insertion_mode = InsertionMode::InCaption;
  1910. return;
  1911. }
  1912. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1913. clear_the_stack_back_to_a_table_context();
  1914. insert_html_element(token);
  1915. m_insertion_mode = InsertionMode::InColumnGroup;
  1916. return;
  1917. }
  1918. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1919. clear_the_stack_back_to_a_table_context();
  1920. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::colgroup));
  1921. m_insertion_mode = InsertionMode::InColumnGroup;
  1922. process_using_the_rules_for(m_insertion_mode, token);
  1923. return;
  1924. }
  1925. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1926. clear_the_stack_back_to_a_table_context();
  1927. insert_html_element(token);
  1928. m_insertion_mode = InsertionMode::InTableBody;
  1929. return;
  1930. }
  1931. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1932. clear_the_stack_back_to_a_table_context();
  1933. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tbody));
  1934. m_insertion_mode = InsertionMode::InTableBody;
  1935. process_using_the_rules_for(m_insertion_mode, token);
  1936. return;
  1937. }
  1938. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1939. log_parse_error();
  1940. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table))
  1941. return;
  1942. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1943. reset_the_insertion_mode_appropriately();
  1944. process_using_the_rules_for(m_insertion_mode, token);
  1945. return;
  1946. }
  1947. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::table) {
  1948. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1949. log_parse_error();
  1950. return;
  1951. }
  1952. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1953. reset_the_insertion_mode_appropriately();
  1954. return;
  1955. }
  1956. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1957. log_parse_error();
  1958. return;
  1959. }
  1960. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::style, HTML::TagNames::script, HTML::TagNames::template_))
  1961. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_)) {
  1962. process_using_the_rules_for(InsertionMode::InHead, token);
  1963. return;
  1964. }
  1965. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1966. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1967. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1968. goto AnythingElse;
  1969. }
  1970. log_parse_error();
  1971. insert_html_element(token);
  1972. // FIXME: Is this the correct interpretation of "Pop that input element off the stack of open elements."?
  1973. // Because this wording is the first time it's seen in the spec.
  1974. // Other times it's worded as: "Immediately pop the current node off the stack of open elements."
  1975. m_stack_of_open_elements.pop();
  1976. token.acknowledge_self_closing_flag_if_set();
  1977. return;
  1978. }
  1979. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1980. log_parse_error();
  1981. if (m_form_element || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1982. return;
  1983. }
  1984. m_form_element = verify_cast<HTMLFormElement>(*insert_html_element(token));
  1985. // FIXME: See previous FIXME, as this is the same situation but for form.
  1986. m_stack_of_open_elements.pop();
  1987. return;
  1988. }
  1989. if (token.is_end_of_file()) {
  1990. process_using_the_rules_for(InsertionMode::InBody, token);
  1991. return;
  1992. }
  1993. AnythingElse:
  1994. log_parse_error();
  1995. m_foster_parenting = true;
  1996. process_using_the_rules_for(InsertionMode::InBody, token);
  1997. m_foster_parenting = false;
  1998. }
  1999. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  2000. {
  2001. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::td, HTML::TagNames::th)) {
  2002. log_parse_error();
  2003. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2004. reset_the_insertion_mode_appropriately();
  2005. process_using_the_rules_for(m_insertion_mode, token);
  2006. return;
  2007. }
  2008. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::td, HTML::TagNames::th)) {
  2009. log_parse_error();
  2010. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name()))
  2011. return;
  2012. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2013. reset_the_insertion_mode_appropriately();
  2014. process_using_the_rules_for(m_insertion_mode, token);
  2015. return;
  2016. }
  2017. process_using_the_rules_for(InsertionMode::InSelect, token);
  2018. }
  2019. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  2020. {
  2021. if (token.is_character()) {
  2022. if (token.code_point() == 0) {
  2023. log_parse_error();
  2024. return;
  2025. }
  2026. insert_character(token.code_point());
  2027. return;
  2028. }
  2029. if (token.is_comment()) {
  2030. insert_comment(token);
  2031. return;
  2032. }
  2033. if (token.is_doctype()) {
  2034. log_parse_error();
  2035. return;
  2036. }
  2037. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2038. process_using_the_rules_for(InsertionMode::InBody, token);
  2039. return;
  2040. }
  2041. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  2042. if (current_node().local_name() == HTML::TagNames::option) {
  2043. m_stack_of_open_elements.pop();
  2044. }
  2045. insert_html_element(token);
  2046. return;
  2047. }
  2048. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2049. if (current_node().local_name() == HTML::TagNames::option) {
  2050. m_stack_of_open_elements.pop();
  2051. }
  2052. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2053. m_stack_of_open_elements.pop();
  2054. }
  2055. insert_html_element(token);
  2056. return;
  2057. }
  2058. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2059. if (current_node().local_name() == HTML::TagNames::option && node_before_current_node().local_name() == HTML::TagNames::optgroup)
  2060. m_stack_of_open_elements.pop();
  2061. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2062. m_stack_of_open_elements.pop();
  2063. } else {
  2064. log_parse_error();
  2065. return;
  2066. }
  2067. return;
  2068. }
  2069. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  2070. if (current_node().local_name() == HTML::TagNames::option) {
  2071. m_stack_of_open_elements.pop();
  2072. } else {
  2073. log_parse_error();
  2074. return;
  2075. }
  2076. return;
  2077. }
  2078. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  2079. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2080. VERIFY(m_parsing_fragment);
  2081. log_parse_error();
  2082. return;
  2083. }
  2084. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2085. reset_the_insertion_mode_appropriately();
  2086. return;
  2087. }
  2088. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  2089. log_parse_error();
  2090. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2091. VERIFY(m_parsing_fragment);
  2092. return;
  2093. }
  2094. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2095. reset_the_insertion_mode_appropriately();
  2096. return;
  2097. }
  2098. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  2099. log_parse_error();
  2100. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2101. VERIFY(m_parsing_fragment);
  2102. return;
  2103. }
  2104. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2105. reset_the_insertion_mode_appropriately();
  2106. process_using_the_rules_for(m_insertion_mode, token);
  2107. return;
  2108. }
  2109. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  2110. process_using_the_rules_for(InsertionMode::InHead, token);
  2111. return;
  2112. }
  2113. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2114. process_using_the_rules_for(InsertionMode::InHead, token);
  2115. return;
  2116. }
  2117. if (token.is_end_of_file()) {
  2118. process_using_the_rules_for(InsertionMode::InBody, token);
  2119. return;
  2120. }
  2121. log_parse_error();
  2122. }
  2123. void HTMLDocumentParser::handle_in_caption(HTMLToken& token)
  2124. {
  2125. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::caption) {
  2126. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2127. VERIFY(m_parsing_fragment);
  2128. log_parse_error();
  2129. return;
  2130. }
  2131. generate_implied_end_tags();
  2132. if (current_node().local_name() != HTML::TagNames::caption)
  2133. log_parse_error();
  2134. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2135. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2136. m_insertion_mode = InsertionMode::InTable;
  2137. return;
  2138. }
  2139. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))
  2140. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  2141. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2142. VERIFY(m_parsing_fragment);
  2143. log_parse_error();
  2144. return;
  2145. }
  2146. generate_implied_end_tags();
  2147. if (current_node().local_name() != HTML::TagNames::caption)
  2148. log_parse_error();
  2149. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2150. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2151. m_insertion_mode = InsertionMode::InTable;
  2152. process_using_the_rules_for(m_insertion_mode, token);
  2153. return;
  2154. }
  2155. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  2156. log_parse_error();
  2157. return;
  2158. }
  2159. process_using_the_rules_for(InsertionMode::InBody, token);
  2160. }
  2161. void HTMLDocumentParser::handle_in_column_group(HTMLToken& token)
  2162. {
  2163. if (token.is_character() && token.is_parser_whitespace()) {
  2164. insert_character(token.code_point());
  2165. return;
  2166. }
  2167. if (token.is_comment()) {
  2168. insert_comment(token);
  2169. return;
  2170. }
  2171. if (token.is_doctype()) {
  2172. log_parse_error();
  2173. return;
  2174. }
  2175. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2176. process_using_the_rules_for(InsertionMode::InBody, token);
  2177. return;
  2178. }
  2179. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2180. insert_html_element(token);
  2181. m_stack_of_open_elements.pop();
  2182. token.acknowledge_self_closing_flag_if_set();
  2183. return;
  2184. }
  2185. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  2186. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2187. log_parse_error();
  2188. return;
  2189. }
  2190. m_stack_of_open_elements.pop();
  2191. m_insertion_mode = InsertionMode::InTable;
  2192. return;
  2193. }
  2194. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::col) {
  2195. log_parse_error();
  2196. return;
  2197. }
  2198. if ((token.is_start_tag() || token.is_end_tag()) && token.tag_name() == HTML::TagNames::template_) {
  2199. process_using_the_rules_for(InsertionMode::InHead, token);
  2200. return;
  2201. }
  2202. if (token.is_end_of_file()) {
  2203. process_using_the_rules_for(InsertionMode::InBody, token);
  2204. return;
  2205. }
  2206. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2207. log_parse_error();
  2208. return;
  2209. }
  2210. m_stack_of_open_elements.pop();
  2211. m_insertion_mode = InsertionMode::InTable;
  2212. process_using_the_rules_for(m_insertion_mode, token);
  2213. }
  2214. void HTMLDocumentParser::handle_in_template(HTMLToken& token)
  2215. {
  2216. if (token.is_character() || token.is_comment() || token.is_doctype()) {
  2217. process_using_the_rules_for(InsertionMode::InBody, token);
  2218. return;
  2219. }
  2220. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  2221. process_using_the_rules_for(InsertionMode::InHead, token);
  2222. return;
  2223. }
  2224. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2225. process_using_the_rules_for(InsertionMode::InHead, token);
  2226. return;
  2227. }
  2228. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  2229. m_stack_of_template_insertion_modes.take_last();
  2230. m_stack_of_template_insertion_modes.append(InsertionMode::InTable);
  2231. m_insertion_mode = InsertionMode::InTable;
  2232. process_using_the_rules_for(m_insertion_mode, token);
  2233. return;
  2234. }
  2235. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2236. m_stack_of_template_insertion_modes.take_last();
  2237. m_stack_of_template_insertion_modes.append(InsertionMode::InColumnGroup);
  2238. m_insertion_mode = InsertionMode::InColumnGroup;
  2239. process_using_the_rules_for(m_insertion_mode, token);
  2240. return;
  2241. }
  2242. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  2243. m_stack_of_template_insertion_modes.take_last();
  2244. m_stack_of_template_insertion_modes.append(InsertionMode::InTableBody);
  2245. m_insertion_mode = InsertionMode::InTableBody;
  2246. process_using_the_rules_for(m_insertion_mode, token);
  2247. return;
  2248. }
  2249. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2250. m_stack_of_template_insertion_modes.take_last();
  2251. m_stack_of_template_insertion_modes.append(InsertionMode::InRow);
  2252. m_insertion_mode = InsertionMode::InRow;
  2253. process_using_the_rules_for(m_insertion_mode, token);
  2254. return;
  2255. }
  2256. if (token.is_start_tag()) {
  2257. m_stack_of_template_insertion_modes.take_last();
  2258. m_stack_of_template_insertion_modes.append(InsertionMode::InBody);
  2259. m_insertion_mode = InsertionMode::InBody;
  2260. process_using_the_rules_for(m_insertion_mode, token);
  2261. return;
  2262. }
  2263. if (token.is_end_tag()) {
  2264. log_parse_error();
  2265. return;
  2266. }
  2267. if (token.is_end_of_file()) {
  2268. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  2269. VERIFY(m_parsing_fragment);
  2270. stop_parsing();
  2271. return;
  2272. }
  2273. log_parse_error();
  2274. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  2275. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2276. m_stack_of_template_insertion_modes.take_last();
  2277. reset_the_insertion_mode_appropriately();
  2278. process_using_the_rules_for(m_insertion_mode, token);
  2279. }
  2280. }
  2281. void HTMLDocumentParser::handle_in_frameset(HTMLToken& token)
  2282. {
  2283. if (token.is_character() && token.is_parser_whitespace()) {
  2284. insert_character(token.code_point());
  2285. return;
  2286. }
  2287. if (token.is_comment()) {
  2288. insert_comment(token);
  2289. return;
  2290. }
  2291. if (token.is_doctype()) {
  2292. log_parse_error();
  2293. return;
  2294. }
  2295. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2296. process_using_the_rules_for(InsertionMode::InBody, token);
  2297. return;
  2298. }
  2299. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2300. insert_html_element(token);
  2301. return;
  2302. }
  2303. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2304. // FIXME: If the current node is the root html element, then this is a parse error; ignore the token. (fragment case)
  2305. m_stack_of_open_elements.pop();
  2306. if (!m_parsing_fragment && current_node().local_name() != HTML::TagNames::frameset) {
  2307. m_insertion_mode = InsertionMode::AfterFrameset;
  2308. }
  2309. return;
  2310. }
  2311. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frame) {
  2312. insert_html_element(token);
  2313. m_stack_of_open_elements.pop();
  2314. token.acknowledge_self_closing_flag_if_set();
  2315. return;
  2316. }
  2317. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2318. process_using_the_rules_for(InsertionMode::InHead, token);
  2319. return;
  2320. }
  2321. if (token.is_end_of_file()) {
  2322. //FIXME: If the current node is not the root html element, then this is a parse error.
  2323. stop_parsing();
  2324. return;
  2325. }
  2326. log_parse_error();
  2327. }
  2328. void HTMLDocumentParser::handle_after_frameset(HTMLToken& token)
  2329. {
  2330. if (token.is_character() && token.is_parser_whitespace()) {
  2331. insert_character(token.code_point());
  2332. return;
  2333. }
  2334. if (token.is_comment()) {
  2335. insert_comment(token);
  2336. return;
  2337. }
  2338. if (token.is_doctype()) {
  2339. log_parse_error();
  2340. return;
  2341. }
  2342. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2343. process_using_the_rules_for(InsertionMode::InBody, token);
  2344. return;
  2345. }
  2346. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  2347. m_insertion_mode = InsertionMode::AfterAfterFrameset;
  2348. return;
  2349. }
  2350. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2351. process_using_the_rules_for(InsertionMode::InHead, token);
  2352. return;
  2353. }
  2354. if (token.is_end_of_file()) {
  2355. stop_parsing();
  2356. return;
  2357. }
  2358. log_parse_error();
  2359. }
  2360. void HTMLDocumentParser::handle_after_after_frameset(HTMLToken& token)
  2361. {
  2362. if (token.is_comment()) {
  2363. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  2364. document().append_child(move(comment));
  2365. return;
  2366. }
  2367. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  2368. process_using_the_rules_for(InsertionMode::InBody, token);
  2369. return;
  2370. }
  2371. if (token.is_end_of_file()) {
  2372. stop_parsing();
  2373. return;
  2374. }
  2375. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2376. process_using_the_rules_for(InsertionMode::InHead, token);
  2377. return;
  2378. }
  2379. log_parse_error();
  2380. }
  2381. void HTMLDocumentParser::process_using_the_rules_for_foreign_content(HTMLToken& token)
  2382. {
  2383. if (token.is_character()) {
  2384. if (token.code_point() == 0) {
  2385. log_parse_error();
  2386. insert_character(0xFFFD);
  2387. return;
  2388. }
  2389. if (token.is_parser_whitespace()) {
  2390. insert_character(token.code_point());
  2391. return;
  2392. }
  2393. insert_character(token.code_point());
  2394. m_frameset_ok = false;
  2395. return;
  2396. }
  2397. if (token.is_comment()) {
  2398. insert_comment(token);
  2399. return;
  2400. }
  2401. if (token.is_doctype()) {
  2402. log_parse_error();
  2403. return;
  2404. }
  2405. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::b, HTML::TagNames::big, HTML::TagNames::blockquote, HTML::TagNames::body, HTML::TagNames::br, HTML::TagNames::center, HTML::TagNames::code, HTML::TagNames::dd, HTML::TagNames::div, HTML::TagNames::dl, HTML::TagNames::dt, HTML::TagNames::em, HTML::TagNames::embed, HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6, HTML::TagNames::head, HTML::TagNames::hr, HTML::TagNames::i, HTML::TagNames::img, HTML::TagNames::li, HTML::TagNames::listing, HTML::TagNames::menu, HTML::TagNames::meta, HTML::TagNames::nobr, HTML::TagNames::ol, HTML::TagNames::p, HTML::TagNames::pre, HTML::TagNames::ruby, HTML::TagNames::s, HTML::TagNames::small, HTML::TagNames::span, HTML::TagNames::strong, HTML::TagNames::strike, HTML::TagNames::sub, HTML::TagNames::sup, HTML::TagNames::table, HTML::TagNames::tt, HTML::TagNames::u, HTML::TagNames::ul, HTML::TagNames::var))
  2406. || (token.is_start_tag() && token.tag_name() == HTML::TagNames::font && (token.has_attribute(HTML::AttributeNames::color) || token.has_attribute(HTML::AttributeNames::face) || token.has_attribute(HTML::AttributeNames::size)))) {
  2407. log_parse_error();
  2408. if (m_parsing_fragment) {
  2409. goto AnyOtherStartTag;
  2410. }
  2411. TODO();
  2412. }
  2413. if (token.is_start_tag()) {
  2414. AnyOtherStartTag:
  2415. if (adjusted_current_node().namespace_() == Namespace::MathML) {
  2416. adjust_mathml_attributes(token);
  2417. } else if (adjusted_current_node().namespace_() == Namespace::SVG) {
  2418. adjust_svg_tag_names(token);
  2419. adjust_svg_attributes(token);
  2420. }
  2421. adjust_foreign_attributes(token);
  2422. insert_foreign_element(token, adjusted_current_node().namespace_());
  2423. if (token.is_self_closing()) {
  2424. if (token.tag_name() == SVG::TagNames::script && current_node().namespace_() == Namespace::SVG) {
  2425. token.acknowledge_self_closing_flag_if_set();
  2426. goto ScriptEndTag;
  2427. }
  2428. m_stack_of_open_elements.pop();
  2429. token.acknowledge_self_closing_flag_if_set();
  2430. }
  2431. return;
  2432. }
  2433. if (token.is_end_tag() && current_node().namespace_() == Namespace::SVG && current_node().tag_name() == SVG::TagNames::script) {
  2434. ScriptEndTag:
  2435. m_stack_of_open_elements.pop();
  2436. TODO();
  2437. }
  2438. if (token.is_end_tag()) {
  2439. RefPtr<DOM::Element> node = current_node();
  2440. // FIXME: Not sure if this is the correct to_lowercase, as the specification says "to ASCII lowercase"
  2441. if (node->tag_name().to_lowercase() != token.tag_name())
  2442. log_parse_error();
  2443. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2444. if (node == m_stack_of_open_elements.first()) {
  2445. VERIFY(m_parsing_fragment);
  2446. return;
  2447. }
  2448. // FIXME: See the above FIXME
  2449. if (node->tag_name().to_lowercase() == token.tag_name()) {
  2450. while (current_node() != node)
  2451. m_stack_of_open_elements.pop();
  2452. m_stack_of_open_elements.pop();
  2453. return;
  2454. }
  2455. node = m_stack_of_open_elements.elements().at(i - 1);
  2456. if (node->namespace_() != Namespace::HTML)
  2457. continue;
  2458. process_using_the_rules_for(m_insertion_mode, token);
  2459. return;
  2460. }
  2461. }
  2462. VERIFY_NOT_REACHED();
  2463. }
  2464. // https://html.spec.whatwg.org/multipage/parsing.html#reset-the-insertion-mode-appropriately
  2465. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  2466. {
  2467. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2468. bool last = i == 0;
  2469. // NOTE: When parsing fragments, we substitute the context element for the root of the stack of open elements.
  2470. RefPtr<DOM::Element> node;
  2471. if (last && m_parsing_fragment) {
  2472. node = m_context_element;
  2473. } else {
  2474. node = m_stack_of_open_elements.elements().at(i);
  2475. }
  2476. if (node->local_name() == HTML::TagNames::select) {
  2477. if (!last) {
  2478. for (ssize_t j = i; j > 0; --j) {
  2479. auto& ancestor = m_stack_of_open_elements.elements().at(j - 1);
  2480. if (is<HTMLTemplateElement>(ancestor))
  2481. break;
  2482. if (is<HTMLTableElement>(ancestor)) {
  2483. m_insertion_mode = InsertionMode::InSelectInTable;
  2484. return;
  2485. }
  2486. }
  2487. }
  2488. m_insertion_mode = InsertionMode::InSelect;
  2489. return;
  2490. }
  2491. if (!last && node->local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2492. m_insertion_mode = InsertionMode::InCell;
  2493. return;
  2494. }
  2495. if (node->local_name() == HTML::TagNames::tr) {
  2496. m_insertion_mode = InsertionMode::InRow;
  2497. return;
  2498. }
  2499. if (node->local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  2500. m_insertion_mode = InsertionMode::InTableBody;
  2501. return;
  2502. }
  2503. if (node->local_name() == HTML::TagNames::caption) {
  2504. m_insertion_mode = InsertionMode::InCaption;
  2505. return;
  2506. }
  2507. if (node->local_name() == HTML::TagNames::colgroup) {
  2508. m_insertion_mode = InsertionMode::InColumnGroup;
  2509. return;
  2510. }
  2511. if (node->local_name() == HTML::TagNames::table) {
  2512. m_insertion_mode = InsertionMode::InTable;
  2513. return;
  2514. }
  2515. if (node->local_name() == HTML::TagNames::template_) {
  2516. m_insertion_mode = m_stack_of_template_insertion_modes.last();
  2517. return;
  2518. }
  2519. if (!last && node->local_name() == HTML::TagNames::head) {
  2520. m_insertion_mode = InsertionMode::InHead;
  2521. return;
  2522. }
  2523. if (node->local_name() == HTML::TagNames::body) {
  2524. m_insertion_mode = InsertionMode::InBody;
  2525. return;
  2526. }
  2527. if (node->local_name() == HTML::TagNames::frameset) {
  2528. VERIFY(m_parsing_fragment);
  2529. m_insertion_mode = InsertionMode::InFrameset;
  2530. return;
  2531. }
  2532. if (node->local_name() == HTML::TagNames::html) {
  2533. if (!m_head_element) {
  2534. VERIFY(m_parsing_fragment);
  2535. m_insertion_mode = InsertionMode::BeforeHead;
  2536. return;
  2537. }
  2538. m_insertion_mode = InsertionMode::AfterHead;
  2539. return;
  2540. }
  2541. }
  2542. VERIFY(m_parsing_fragment);
  2543. m_insertion_mode = InsertionMode::InBody;
  2544. }
  2545. const char* HTMLDocumentParser::insertion_mode_name() const
  2546. {
  2547. switch (m_insertion_mode) {
  2548. #define __ENUMERATE_INSERTION_MODE(mode) \
  2549. case InsertionMode::mode: \
  2550. return #mode;
  2551. ENUMERATE_INSERTION_MODES
  2552. #undef __ENUMERATE_INSERTION_MODE
  2553. }
  2554. VERIFY_NOT_REACHED();
  2555. }
  2556. DOM::Document& HTMLDocumentParser::document()
  2557. {
  2558. return *m_document;
  2559. }
  2560. NonnullRefPtrVector<DOM::Node> HTMLDocumentParser::parse_html_fragment(DOM::Element& context_element, const StringView& markup)
  2561. {
  2562. auto temp_document = DOM::Document::create();
  2563. HTMLDocumentParser parser(*temp_document, markup, "utf-8");
  2564. parser.m_context_element = context_element;
  2565. parser.m_parsing_fragment = true;
  2566. parser.document().set_quirks_mode(context_element.document().mode());
  2567. if (context_element.local_name().is_one_of(HTML::TagNames::title, HTML::TagNames::textarea)) {
  2568. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  2569. } else if (context_element.local_name().is_one_of(HTML::TagNames::style, HTML::TagNames::xmp, HTML::TagNames::iframe, HTML::TagNames::noembed, HTML::TagNames::noframes)) {
  2570. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2571. } else if (context_element.local_name().is_one_of(HTML::TagNames::script)) {
  2572. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  2573. } else if (context_element.local_name().is_one_of(HTML::TagNames::noscript)) {
  2574. if (context_element.document().is_scripting_enabled())
  2575. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2576. } else if (context_element.local_name().is_one_of(HTML::TagNames::plaintext)) {
  2577. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  2578. }
  2579. auto root = create_element(context_element.document(), HTML::TagNames::html, Namespace::HTML);
  2580. parser.document().append_child(root);
  2581. parser.m_stack_of_open_elements.push(root);
  2582. if (context_element.local_name() == HTML::TagNames::template_) {
  2583. parser.m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  2584. }
  2585. // FIXME: Create a start tag token whose name is the local name of context and whose attributes are the attributes of context.
  2586. parser.reset_the_insertion_mode_appropriately();
  2587. for (auto* form_candidate = &context_element; form_candidate; form_candidate = form_candidate->parent_element()) {
  2588. if (is<HTMLFormElement>(*form_candidate)) {
  2589. parser.m_form_element = verify_cast<HTMLFormElement>(*form_candidate);
  2590. break;
  2591. }
  2592. }
  2593. parser.run(context_element.document().url());
  2594. NonnullRefPtrVector<DOM::Node> children;
  2595. while (RefPtr<DOM::Node> child = root->first_child()) {
  2596. root->remove_child(*child);
  2597. context_element.document().adopt_node(*child);
  2598. children.append(*child);
  2599. }
  2600. return children;
  2601. }
  2602. NonnullOwnPtr<HTMLDocumentParser> HTMLDocumentParser::create_with_uncertain_encoding(DOM::Document& document, const ByteBuffer& input)
  2603. {
  2604. if (document.has_encoding())
  2605. return make<HTMLDocumentParser>(document, input, document.encoding().value());
  2606. auto encoding = run_encoding_sniffing_algorithm(input);
  2607. dbgln("The encoding sniffing algorithm returned encoding '{}'", encoding);
  2608. return make<HTMLDocumentParser>(document, input, encoding);
  2609. }
  2610. }