HTMLDocumentParser.cpp 117 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658165916601661166216631664166516661667166816691670167116721673167416751676167716781679168016811682168316841685168616871688168916901691169216931694169516961697169816991700170117021703170417051706170717081709171017111712171317141715171617171718171917201721172217231724172517261727172817291730173117321733173417351736173717381739174017411742174317441745174617471748174917501751175217531754175517561757175817591760176117621763176417651766176717681769177017711772177317741775177617771778177917801781178217831784178517861787178817891790179117921793179417951796179717981799180018011802180318041805180618071808180918101811181218131814181518161817181818191820182118221823182418251826182718281829183018311832183318341835183618371838183918401841184218431844184518461847184818491850185118521853185418551856185718581859186018611862186318641865186618671868186918701871187218731874187518761877187818791880188118821883188418851886188718881889189018911892189318941895189618971898189919001901190219031904190519061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192919301931193219331934193519361937193819391940194119421943194419451946194719481949195019511952195319541955195619571958195919601961196219631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994199519961997199819992000200120022003200420052006200720082009201020112012201320142015201620172018201920202021202220232024202520262027202820292030203120322033203420352036203720382039204020412042204320442045204620472048204920502051205220532054205520562057205820592060206120622063206420652066206720682069207020712072207320742075207620772078207920802081208220832084208520862087208820892090209120922093209420952096209720982099210021012102210321042105210621072108210921102111211221132114211521162117211821192120212121222123212421252126212721282129213021312132213321342135213621372138213921402141214221432144214521462147214821492150215121522153215421552156215721582159216021612162216321642165216621672168216921702171217221732174217521762177217821792180218121822183218421852186218721882189219021912192219321942195219621972198219922002201220222032204220522062207220822092210221122122213221422152216221722182219222022212222222322242225222622272228222922302231223222332234223522362237223822392240224122422243224422452246224722482249225022512252225322542255225622572258225922602261226222632264226522662267226822692270227122722273227422752276227722782279228022812282228322842285228622872288228922902291229222932294229522962297229822992300230123022303230423052306230723082309231023112312231323142315231623172318231923202321232223232324232523262327232823292330233123322333233423352336233723382339234023412342234323442345234623472348234923502351235223532354235523562357235823592360236123622363236423652366236723682369237023712372237323742375237623772378237923802381238223832384238523862387238823892390239123922393239423952396239723982399240024012402240324042405240624072408240924102411241224132414241524162417241824192420242124222423242424252426242724282429243024312432243324342435243624372438243924402441244224432444244524462447244824492450245124522453245424552456245724582459246024612462246324642465246624672468246924702471247224732474247524762477247824792480248124822483248424852486248724882489249024912492249324942495249624972498249925002501250225032504250525062507250825092510251125122513251425152516251725182519252025212522252325242525252625272528252925302531253225332534253525362537253825392540254125422543254425452546254725482549255025512552255325542555255625572558255925602561256225632564256525662567256825692570257125722573257425752576257725782579258025812582258325842585258625872588258925902591259225932594259525962597259825992600260126022603260426052606260726082609261026112612261326142615261626172618261926202621262226232624262526262627262826292630263126322633263426352636263726382639264026412642264326442645264626472648264926502651265226532654265526562657265826592660266126622663266426652666266726682669267026712672267326742675267626772678267926802681268226832684268526862687268826892690269126922693269426952696269726982699270027012702270327042705270627072708270927102711271227132714271527162717271827192720272127222723272427252726272727282729273027312732273327342735273627372738273927402741274227432744274527462747274827492750275127522753275427552756275727582759276027612762276327642765276627672768276927702771277227732774277527762777277827792780278127822783278427852786278727882789279027912792279327942795279627972798279928002801280228032804280528062807280828092810281128122813281428152816281728182819282028212822282328242825282628272828282928302831283228332834283528362837283828392840284128422843284428452846284728482849285028512852285328542855285628572858285928602861286228632864286528662867286828692870287128722873287428752876287728782879288028812882288328842885288628872888288928902891289228932894289528962897289828992900290129022903290429052906290729082909291029112912291329142915291629172918291929202921292229232924292529262927292829292930293129322933293429352936293729382939294029412942294329442945294629472948294929502951295229532954295529562957295829592960296129622963296429652966296729682969297029712972297329742975297629772978297929802981298229832984298529862987298829892990299129922993299429952996299729982999300030013002300330043005300630073008300930103011301230133014301530163017301830193020302130223023302430253026302730283029303030313032303330343035303630373038303930403041304230433044304530463047304830493050
  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 != "html")
  242. return DOM::QuirksMode::Yes;
  243. auto const& public_identifier = doctype_token.m_doctype.public_identifier;
  244. auto const& system_identifier = doctype_token.m_doctype.system_identifier;
  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.comment()));
  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);
  288. doctype->set_public_id(token.m_doctype.public_identifier);
  289. doctype->set_system_id(token.m_doctype.system_identifier);
  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.comment()));
  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, attribute.value);
  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 adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  450. adjusted_insertion_location.parent->insert_before(adopt_ref(*new DOM::Comment(document(), token.comment())), adjusted_insertion_location.insert_before_sibling);
  451. }
  452. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  453. {
  454. if (token.is_parser_whitespace()) {
  455. insert_character(token.code_point());
  456. return;
  457. }
  458. if (token.is_comment()) {
  459. insert_comment(token);
  460. return;
  461. }
  462. if (token.is_doctype()) {
  463. log_parse_error();
  464. return;
  465. }
  466. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  467. process_using_the_rules_for(InsertionMode::InBody, token);
  468. return;
  469. }
  470. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  471. insert_html_element(token);
  472. m_stack_of_open_elements.pop();
  473. token.acknowledge_self_closing_flag_if_set();
  474. return;
  475. }
  476. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  477. auto element = insert_html_element(token);
  478. m_stack_of_open_elements.pop();
  479. token.acknowledge_self_closing_flag_if_set();
  480. return;
  481. }
  482. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  483. insert_html_element(token);
  484. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  485. m_original_insertion_mode = m_insertion_mode;
  486. m_insertion_mode = InsertionMode::Text;
  487. return;
  488. }
  489. 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)) {
  490. parse_generic_raw_text_element(token);
  491. return;
  492. }
  493. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noscript && !m_scripting_enabled) {
  494. insert_html_element(token);
  495. m_insertion_mode = InsertionMode::InHeadNoscript;
  496. return;
  497. }
  498. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  499. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  500. auto element = create_element_for(token, Namespace::HTML);
  501. auto& script_element = verify_cast<HTMLScriptElement>(*element);
  502. script_element.set_parser_document({}, document());
  503. script_element.set_non_blocking({}, false);
  504. if (m_parsing_fragment) {
  505. script_element.set_already_started({}, true);
  506. }
  507. if (m_invoked_via_document_write) {
  508. TODO();
  509. }
  510. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling, false);
  511. m_stack_of_open_elements.push(element);
  512. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  513. m_original_insertion_mode = m_insertion_mode;
  514. m_insertion_mode = InsertionMode::Text;
  515. return;
  516. }
  517. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  518. m_stack_of_open_elements.pop();
  519. m_insertion_mode = InsertionMode::AfterHead;
  520. return;
  521. }
  522. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  523. goto AnythingElse;
  524. }
  525. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  526. insert_html_element(token);
  527. m_list_of_active_formatting_elements.add_marker();
  528. m_frameset_ok = false;
  529. m_insertion_mode = InsertionMode::InTemplate;
  530. m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  531. return;
  532. }
  533. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  534. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  535. log_parse_error();
  536. return;
  537. }
  538. generate_all_implied_end_tags_thoroughly();
  539. if (current_node().local_name() != HTML::TagNames::template_)
  540. log_parse_error();
  541. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  542. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  543. m_stack_of_template_insertion_modes.take_last();
  544. reset_the_insertion_mode_appropriately();
  545. return;
  546. }
  547. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  548. log_parse_error();
  549. return;
  550. }
  551. AnythingElse:
  552. m_stack_of_open_elements.pop();
  553. m_insertion_mode = InsertionMode::AfterHead;
  554. process_using_the_rules_for(m_insertion_mode, token);
  555. }
  556. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken& token)
  557. {
  558. if (token.is_doctype()) {
  559. log_parse_error();
  560. return;
  561. }
  562. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  563. process_using_the_rules_for(InsertionMode::InBody, token);
  564. return;
  565. }
  566. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::noscript) {
  567. m_stack_of_open_elements.pop();
  568. m_insertion_mode = InsertionMode::InHead;
  569. return;
  570. }
  571. 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))) {
  572. process_using_the_rules_for(InsertionMode::InHead, token);
  573. return;
  574. }
  575. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  576. goto AnythingElse;
  577. }
  578. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::noscript)) {
  579. log_parse_error();
  580. return;
  581. }
  582. AnythingElse:
  583. log_parse_error();
  584. m_stack_of_open_elements.pop();
  585. m_insertion_mode = InsertionMode::InHead;
  586. process_using_the_rules_for(m_insertion_mode, token);
  587. }
  588. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  589. {
  590. insert_html_element(token);
  591. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  592. m_original_insertion_mode = m_insertion_mode;
  593. m_insertion_mode = InsertionMode::Text;
  594. }
  595. DOM::Text* HTMLDocumentParser::find_character_insertion_node()
  596. {
  597. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  598. if (adjusted_insertion_location.insert_before_sibling) {
  599. TODO();
  600. }
  601. if (adjusted_insertion_location.parent->is_document())
  602. return nullptr;
  603. if (adjusted_insertion_location.parent->last_child() && adjusted_insertion_location.parent->last_child()->is_text())
  604. return verify_cast<DOM::Text>(adjusted_insertion_location.parent->last_child());
  605. auto new_text_node = adopt_ref(*new DOM::Text(document(), ""));
  606. adjusted_insertion_location.parent->append_child(new_text_node);
  607. return new_text_node;
  608. }
  609. void HTMLDocumentParser::flush_character_insertions()
  610. {
  611. if (m_character_insertion_builder.is_empty())
  612. return;
  613. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  614. m_character_insertion_node->parent()->children_changed();
  615. m_character_insertion_builder.clear();
  616. }
  617. void HTMLDocumentParser::insert_character(u32 data)
  618. {
  619. auto node = find_character_insertion_node();
  620. if (node == m_character_insertion_node) {
  621. m_character_insertion_builder.append(Utf32View { &data, 1 });
  622. return;
  623. }
  624. if (!m_character_insertion_node) {
  625. m_character_insertion_node = node;
  626. m_character_insertion_builder.append(Utf32View { &data, 1 });
  627. return;
  628. }
  629. flush_character_insertions();
  630. m_character_insertion_node = node;
  631. m_character_insertion_builder.append(Utf32View { &data, 1 });
  632. }
  633. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  634. {
  635. if (token.is_character() && token.is_parser_whitespace()) {
  636. insert_character(token.code_point());
  637. return;
  638. }
  639. if (token.is_comment()) {
  640. insert_comment(token);
  641. return;
  642. }
  643. if (token.is_doctype()) {
  644. log_parse_error();
  645. return;
  646. }
  647. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  648. process_using_the_rules_for(InsertionMode::InBody, token);
  649. return;
  650. }
  651. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  652. insert_html_element(token);
  653. m_frameset_ok = false;
  654. m_insertion_mode = InsertionMode::InBody;
  655. return;
  656. }
  657. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  658. insert_html_element(token);
  659. m_insertion_mode = InsertionMode::InFrameset;
  660. return;
  661. }
  662. 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)) {
  663. log_parse_error();
  664. m_stack_of_open_elements.push(*m_head_element);
  665. process_using_the_rules_for(InsertionMode::InHead, token);
  666. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  667. return entry.ptr() == m_head_element;
  668. });
  669. return;
  670. }
  671. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  672. process_using_the_rules_for(InsertionMode::InHead, token);
  673. return;
  674. }
  675. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  676. goto AnythingElse;
  677. }
  678. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  679. log_parse_error();
  680. return;
  681. }
  682. AnythingElse:
  683. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::body));
  684. m_insertion_mode = InsertionMode::InBody;
  685. process_using_the_rules_for(m_insertion_mode, token);
  686. }
  687. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  688. {
  689. 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))
  690. m_stack_of_open_elements.pop();
  691. }
  692. void HTMLDocumentParser::generate_all_implied_end_tags_thoroughly()
  693. {
  694. 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))
  695. m_stack_of_open_elements.pop();
  696. }
  697. void HTMLDocumentParser::close_a_p_element()
  698. {
  699. generate_implied_end_tags(HTML::TagNames::p);
  700. if (current_node().local_name() != HTML::TagNames::p) {
  701. log_parse_error();
  702. }
  703. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  704. }
  705. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  706. {
  707. if (token.is_character() && token.is_parser_whitespace()) {
  708. process_using_the_rules_for(InsertionMode::InBody, token);
  709. return;
  710. }
  711. if (token.is_comment()) {
  712. auto& insertion_location = m_stack_of_open_elements.first();
  713. insertion_location.append_child(adopt_ref(*new DOM::Comment(document(), token.comment())));
  714. return;
  715. }
  716. if (token.is_doctype()) {
  717. log_parse_error();
  718. return;
  719. }
  720. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  721. process_using_the_rules_for(InsertionMode::InBody, token);
  722. return;
  723. }
  724. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  725. if (m_parsing_fragment) {
  726. log_parse_error();
  727. return;
  728. }
  729. m_insertion_mode = InsertionMode::AfterAfterBody;
  730. return;
  731. }
  732. if (token.is_end_of_file()) {
  733. stop_parsing();
  734. return;
  735. }
  736. log_parse_error();
  737. m_insertion_mode = InsertionMode::InBody;
  738. process_using_the_rules_for(InsertionMode::InBody, token);
  739. }
  740. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  741. {
  742. if (token.is_comment()) {
  743. auto comment = adopt_ref(*new DOM::Comment(document(), token.comment()));
  744. document().append_child(move(comment));
  745. return;
  746. }
  747. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  748. process_using_the_rules_for(InsertionMode::InBody, token);
  749. return;
  750. }
  751. if (token.is_end_of_file()) {
  752. stop_parsing();
  753. return;
  754. }
  755. log_parse_error();
  756. m_insertion_mode = InsertionMode::InBody;
  757. process_using_the_rules_for(m_insertion_mode, token);
  758. }
  759. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  760. {
  761. // FIXME: This needs to care about "markers"
  762. if (m_list_of_active_formatting_elements.is_empty())
  763. return;
  764. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  765. return;
  766. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  767. return;
  768. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  769. RefPtr<DOM::Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  770. VERIFY(entry);
  771. Rewind:
  772. if (index == 0) {
  773. goto Create;
  774. }
  775. --index;
  776. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  777. VERIFY(entry);
  778. if (!m_stack_of_open_elements.contains(*entry))
  779. goto Rewind;
  780. Advance:
  781. ++index;
  782. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  783. VERIFY(entry);
  784. Create:
  785. // FIXME: Hold on to the real token!
  786. auto new_element = insert_html_element(HTMLToken::make_start_tag(entry->local_name()));
  787. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  788. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  789. goto Advance;
  790. }
  791. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  792. {
  793. auto subject = token.tag_name();
  794. // If the current node is an HTML element whose tag name is subject,
  795. // and the current node is not in the list of active formatting elements,
  796. // then pop the current node off the stack of open elements, and return.
  797. if (current_node().local_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  798. m_stack_of_open_elements.pop();
  799. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  800. }
  801. size_t outer_loop_counter = 0;
  802. //OuterLoop:
  803. if (outer_loop_counter >= 8)
  804. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  805. ++outer_loop_counter;
  806. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  807. if (!formatting_element)
  808. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  809. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  810. log_parse_error();
  811. m_list_of_active_formatting_elements.remove(*formatting_element);
  812. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  813. }
  814. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  815. log_parse_error();
  816. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  817. }
  818. if (formatting_element != &current_node()) {
  819. log_parse_error();
  820. }
  821. RefPtr<DOM::Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  822. if (!furthest_block) {
  823. while (&current_node() != formatting_element)
  824. m_stack_of_open_elements.pop();
  825. m_stack_of_open_elements.pop();
  826. m_list_of_active_formatting_elements.remove(*formatting_element);
  827. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  828. }
  829. // FIXME: Implement the rest of the AAA :^)
  830. TODO();
  831. }
  832. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name, const FlyString& namespace_)
  833. {
  834. if (namespace_ == Namespace::HTML) {
  835. return tag_name.is_one_of(
  836. HTML::TagNames::address,
  837. HTML::TagNames::applet,
  838. HTML::TagNames::area,
  839. HTML::TagNames::article,
  840. HTML::TagNames::aside,
  841. HTML::TagNames::base,
  842. HTML::TagNames::basefont,
  843. HTML::TagNames::bgsound,
  844. HTML::TagNames::blockquote,
  845. HTML::TagNames::body,
  846. HTML::TagNames::br,
  847. HTML::TagNames::button,
  848. HTML::TagNames::caption,
  849. HTML::TagNames::center,
  850. HTML::TagNames::col,
  851. HTML::TagNames::colgroup,
  852. HTML::TagNames::dd,
  853. HTML::TagNames::details,
  854. HTML::TagNames::dir,
  855. HTML::TagNames::div,
  856. HTML::TagNames::dl,
  857. HTML::TagNames::dt,
  858. HTML::TagNames::embed,
  859. HTML::TagNames::fieldset,
  860. HTML::TagNames::figcaption,
  861. HTML::TagNames::figure,
  862. HTML::TagNames::footer,
  863. HTML::TagNames::form,
  864. HTML::TagNames::frame,
  865. HTML::TagNames::frameset,
  866. HTML::TagNames::h1,
  867. HTML::TagNames::h2,
  868. HTML::TagNames::h3,
  869. HTML::TagNames::h4,
  870. HTML::TagNames::h5,
  871. HTML::TagNames::h6,
  872. HTML::TagNames::head,
  873. HTML::TagNames::header,
  874. HTML::TagNames::hgroup,
  875. HTML::TagNames::hr,
  876. HTML::TagNames::html,
  877. HTML::TagNames::iframe,
  878. HTML::TagNames::img,
  879. HTML::TagNames::input,
  880. HTML::TagNames::keygen,
  881. HTML::TagNames::li,
  882. HTML::TagNames::link,
  883. HTML::TagNames::listing,
  884. HTML::TagNames::main,
  885. HTML::TagNames::marquee,
  886. HTML::TagNames::menu,
  887. HTML::TagNames::meta,
  888. HTML::TagNames::nav,
  889. HTML::TagNames::noembed,
  890. HTML::TagNames::noframes,
  891. HTML::TagNames::noscript,
  892. HTML::TagNames::object,
  893. HTML::TagNames::ol,
  894. HTML::TagNames::p,
  895. HTML::TagNames::param,
  896. HTML::TagNames::plaintext,
  897. HTML::TagNames::pre,
  898. HTML::TagNames::script,
  899. HTML::TagNames::section,
  900. HTML::TagNames::select,
  901. HTML::TagNames::source,
  902. HTML::TagNames::style,
  903. HTML::TagNames::summary,
  904. HTML::TagNames::table,
  905. HTML::TagNames::tbody,
  906. HTML::TagNames::td,
  907. HTML::TagNames::template_,
  908. HTML::TagNames::textarea,
  909. HTML::TagNames::tfoot,
  910. HTML::TagNames::th,
  911. HTML::TagNames::thead,
  912. HTML::TagNames::title,
  913. HTML::TagNames::tr,
  914. HTML::TagNames::track,
  915. HTML::TagNames::ul,
  916. HTML::TagNames::wbr,
  917. HTML::TagNames::xmp);
  918. } else if (namespace_ == Namespace::SVG) {
  919. return tag_name.is_one_of(
  920. SVG::TagNames::desc,
  921. SVG::TagNames::foreignObject,
  922. SVG::TagNames::title);
  923. } else if (namespace_ == Namespace::MathML) {
  924. TODO();
  925. }
  926. return false;
  927. }
  928. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  929. {
  930. if (token.is_character()) {
  931. if (token.code_point() == 0) {
  932. log_parse_error();
  933. return;
  934. }
  935. if (token.is_parser_whitespace()) {
  936. reconstruct_the_active_formatting_elements();
  937. insert_character(token.code_point());
  938. return;
  939. }
  940. reconstruct_the_active_formatting_elements();
  941. insert_character(token.code_point());
  942. m_frameset_ok = false;
  943. return;
  944. }
  945. if (token.is_comment()) {
  946. insert_comment(token);
  947. return;
  948. }
  949. if (token.is_doctype()) {
  950. log_parse_error();
  951. return;
  952. }
  953. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  954. log_parse_error();
  955. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  956. return;
  957. for (auto& attribute : token.m_tag.attributes) {
  958. if (current_node().has_attribute(attribute.local_name))
  959. continue;
  960. current_node().set_attribute(attribute.local_name, attribute.value);
  961. }
  962. return;
  963. }
  964. 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)) {
  965. process_using_the_rules_for(InsertionMode::InHead, token);
  966. return;
  967. }
  968. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  969. process_using_the_rules_for(InsertionMode::InHead, token);
  970. return;
  971. }
  972. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  973. log_parse_error();
  974. if (m_stack_of_open_elements.elements().size() == 1
  975. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body
  976. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  977. VERIFY(m_parsing_fragment);
  978. return;
  979. }
  980. m_frameset_ok = false;
  981. auto& body_element = m_stack_of_open_elements.elements().at(1);
  982. for (auto& attribute : token.m_tag.attributes) {
  983. if (body_element.has_attribute(attribute.local_name))
  984. continue;
  985. body_element.set_attribute(attribute.local_name, attribute.value);
  986. }
  987. return;
  988. }
  989. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  990. log_parse_error();
  991. if (m_stack_of_open_elements.elements().size() == 1
  992. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body) {
  993. VERIFY(m_parsing_fragment);
  994. return;
  995. }
  996. if (!m_frameset_ok)
  997. return;
  998. TODO();
  999. }
  1000. if (token.is_end_of_file()) {
  1001. if (!m_stack_of_template_insertion_modes.is_empty()) {
  1002. process_using_the_rules_for(InsertionMode::InTemplate, token);
  1003. return;
  1004. }
  1005. for (auto& node : m_stack_of_open_elements.elements()) {
  1006. 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)) {
  1007. log_parse_error();
  1008. break;
  1009. }
  1010. }
  1011. stop_parsing();
  1012. return;
  1013. }
  1014. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  1015. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1016. log_parse_error();
  1017. return;
  1018. }
  1019. for (auto& node : m_stack_of_open_elements.elements()) {
  1020. 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)) {
  1021. log_parse_error();
  1022. break;
  1023. }
  1024. }
  1025. m_insertion_mode = InsertionMode::AfterBody;
  1026. return;
  1027. }
  1028. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  1029. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1030. log_parse_error();
  1031. return;
  1032. }
  1033. for (auto& node : m_stack_of_open_elements.elements()) {
  1034. 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)) {
  1035. log_parse_error();
  1036. break;
  1037. }
  1038. }
  1039. m_insertion_mode = InsertionMode::AfterBody;
  1040. process_using_the_rules_for(m_insertion_mode, token);
  1041. return;
  1042. }
  1043. 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)) {
  1044. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1045. close_a_p_element();
  1046. insert_html_element(token);
  1047. return;
  1048. }
  1049. 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)) {
  1050. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1051. close_a_p_element();
  1052. 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)) {
  1053. log_parse_error();
  1054. m_stack_of_open_elements.pop();
  1055. }
  1056. insert_html_element(token);
  1057. return;
  1058. }
  1059. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  1060. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1061. close_a_p_element();
  1062. insert_html_element(token);
  1063. m_frameset_ok = false;
  1064. // If the next token is a U+000A LINE FEED (LF) character token,
  1065. // then ignore that token and move on to the next one.
  1066. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1067. auto next_token = m_tokenizer.next_token();
  1068. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1069. // Ignore it.
  1070. } else {
  1071. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1072. }
  1073. return;
  1074. }
  1075. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1076. if (m_form_element && !m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1077. log_parse_error();
  1078. return;
  1079. }
  1080. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1081. close_a_p_element();
  1082. auto element = insert_html_element(token);
  1083. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  1084. m_form_element = verify_cast<HTMLFormElement>(*element);
  1085. return;
  1086. }
  1087. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  1088. m_frameset_ok = false;
  1089. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1090. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1091. if (node->local_name() == HTML::TagNames::li) {
  1092. generate_implied_end_tags(HTML::TagNames::li);
  1093. if (current_node().local_name() != HTML::TagNames::li) {
  1094. log_parse_error();
  1095. }
  1096. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1097. break;
  1098. }
  1099. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1100. break;
  1101. }
  1102. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1103. close_a_p_element();
  1104. insert_html_element(token);
  1105. return;
  1106. }
  1107. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1108. m_frameset_ok = false;
  1109. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1110. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1111. if (node->local_name() == HTML::TagNames::dd) {
  1112. generate_implied_end_tags(HTML::TagNames::dd);
  1113. if (current_node().local_name() != HTML::TagNames::dd) {
  1114. log_parse_error();
  1115. }
  1116. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  1117. break;
  1118. }
  1119. if (node->local_name() == HTML::TagNames::dt) {
  1120. generate_implied_end_tags(HTML::TagNames::dt);
  1121. if (current_node().local_name() != HTML::TagNames::dt) {
  1122. log_parse_error();
  1123. }
  1124. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  1125. break;
  1126. }
  1127. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1128. break;
  1129. }
  1130. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1131. close_a_p_element();
  1132. insert_html_element(token);
  1133. return;
  1134. }
  1135. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  1136. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1137. close_a_p_element();
  1138. insert_html_element(token);
  1139. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  1140. return;
  1141. }
  1142. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  1143. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  1144. log_parse_error();
  1145. generate_implied_end_tags();
  1146. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  1147. }
  1148. reconstruct_the_active_formatting_elements();
  1149. insert_html_element(token);
  1150. m_frameset_ok = false;
  1151. return;
  1152. }
  1153. 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)) {
  1154. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1155. log_parse_error();
  1156. return;
  1157. }
  1158. generate_implied_end_tags();
  1159. if (current_node().local_name() != token.tag_name()) {
  1160. log_parse_error();
  1161. }
  1162. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1163. return;
  1164. }
  1165. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  1166. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1167. auto node = m_form_element;
  1168. m_form_element = nullptr;
  1169. if (!node || !m_stack_of_open_elements.has_in_scope(*node)) {
  1170. log_parse_error();
  1171. return;
  1172. }
  1173. generate_implied_end_tags();
  1174. if (&current_node() != node) {
  1175. log_parse_error();
  1176. }
  1177. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  1178. } else {
  1179. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  1180. log_parse_error();
  1181. return;
  1182. }
  1183. generate_implied_end_tags();
  1184. if (current_node().local_name() != HTML::TagNames::form) {
  1185. log_parse_error();
  1186. }
  1187. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  1188. }
  1189. return;
  1190. }
  1191. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  1192. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1193. log_parse_error();
  1194. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::p));
  1195. }
  1196. close_a_p_element();
  1197. return;
  1198. }
  1199. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  1200. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  1201. log_parse_error();
  1202. return;
  1203. }
  1204. generate_implied_end_tags(HTML::TagNames::li);
  1205. if (current_node().local_name() != HTML::TagNames::li) {
  1206. log_parse_error();
  1207. dbgln("Expected <li> current node, but had <{}>", current_node().local_name());
  1208. }
  1209. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1210. return;
  1211. }
  1212. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1213. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1214. log_parse_error();
  1215. return;
  1216. }
  1217. generate_implied_end_tags(token.tag_name());
  1218. if (current_node().local_name() != token.tag_name()) {
  1219. log_parse_error();
  1220. }
  1221. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1222. return;
  1223. }
  1224. 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)) {
  1225. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  1226. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  1227. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  1228. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  1229. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  1230. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  1231. log_parse_error();
  1232. return;
  1233. }
  1234. generate_implied_end_tags();
  1235. if (current_node().local_name() != token.tag_name()) {
  1236. log_parse_error();
  1237. }
  1238. for (;;) {
  1239. auto popped_element = m_stack_of_open_elements.pop();
  1240. 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))
  1241. break;
  1242. }
  1243. return;
  1244. }
  1245. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1246. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1247. log_parse_error();
  1248. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1249. goto AnyOtherEndTag;
  1250. m_list_of_active_formatting_elements.remove(*element);
  1251. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1252. return entry.ptr() == element;
  1253. });
  1254. }
  1255. reconstruct_the_active_formatting_elements();
  1256. auto element = insert_html_element(token);
  1257. m_list_of_active_formatting_elements.add(*element);
  1258. return;
  1259. }
  1260. 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)) {
  1261. reconstruct_the_active_formatting_elements();
  1262. auto element = insert_html_element(token);
  1263. m_list_of_active_formatting_elements.add(*element);
  1264. return;
  1265. }
  1266. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1267. reconstruct_the_active_formatting_elements();
  1268. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1269. log_parse_error();
  1270. run_the_adoption_agency_algorithm(token);
  1271. reconstruct_the_active_formatting_elements();
  1272. }
  1273. auto element = insert_html_element(token);
  1274. m_list_of_active_formatting_elements.add(*element);
  1275. return;
  1276. }
  1277. 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)) {
  1278. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1279. goto AnyOtherEndTag;
  1280. return;
  1281. }
  1282. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1283. reconstruct_the_active_formatting_elements();
  1284. insert_html_element(token);
  1285. m_list_of_active_formatting_elements.add_marker();
  1286. m_frameset_ok = false;
  1287. return;
  1288. }
  1289. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1290. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1291. log_parse_error();
  1292. return;
  1293. }
  1294. generate_implied_end_tags();
  1295. if (current_node().local_name() != token.tag_name()) {
  1296. log_parse_error();
  1297. }
  1298. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1299. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1300. return;
  1301. }
  1302. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1303. if (!document().in_quirks_mode()) {
  1304. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1305. close_a_p_element();
  1306. }
  1307. insert_html_element(token);
  1308. m_frameset_ok = false;
  1309. m_insertion_mode = InsertionMode::InTable;
  1310. return;
  1311. }
  1312. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1313. token.drop_attributes();
  1314. goto BRStartTag;
  1315. }
  1316. 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)) {
  1317. BRStartTag:
  1318. reconstruct_the_active_formatting_elements();
  1319. insert_html_element(token);
  1320. m_stack_of_open_elements.pop();
  1321. token.acknowledge_self_closing_flag_if_set();
  1322. m_frameset_ok = false;
  1323. return;
  1324. }
  1325. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1326. reconstruct_the_active_formatting_elements();
  1327. insert_html_element(token);
  1328. m_stack_of_open_elements.pop();
  1329. token.acknowledge_self_closing_flag_if_set();
  1330. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1331. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1332. m_frameset_ok = false;
  1333. }
  1334. return;
  1335. }
  1336. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1337. insert_html_element(token);
  1338. m_stack_of_open_elements.pop();
  1339. token.acknowledge_self_closing_flag_if_set();
  1340. return;
  1341. }
  1342. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1343. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1344. close_a_p_element();
  1345. insert_html_element(token);
  1346. m_stack_of_open_elements.pop();
  1347. token.acknowledge_self_closing_flag_if_set();
  1348. m_frameset_ok = false;
  1349. return;
  1350. }
  1351. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::image) {
  1352. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1353. log_parse_error();
  1354. token.set_tag_name("img");
  1355. process_using_the_rules_for(m_insertion_mode, token);
  1356. return;
  1357. }
  1358. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1359. insert_html_element(token);
  1360. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1361. // If the next token is a U+000A LINE FEED (LF) character token,
  1362. // then ignore that token and move on to the next one.
  1363. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1364. auto next_token = m_tokenizer.next_token();
  1365. m_original_insertion_mode = m_insertion_mode;
  1366. m_frameset_ok = false;
  1367. m_insertion_mode = InsertionMode::Text;
  1368. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1369. // Ignore it.
  1370. } else {
  1371. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1372. }
  1373. return;
  1374. }
  1375. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1376. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1377. close_a_p_element();
  1378. }
  1379. reconstruct_the_active_formatting_elements();
  1380. m_frameset_ok = false;
  1381. parse_generic_raw_text_element(token);
  1382. return;
  1383. }
  1384. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1385. m_frameset_ok = false;
  1386. parse_generic_raw_text_element(token);
  1387. return;
  1388. }
  1389. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1390. parse_generic_raw_text_element(token);
  1391. return;
  1392. }
  1393. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1394. reconstruct_the_active_formatting_elements();
  1395. insert_html_element(token);
  1396. m_frameset_ok = false;
  1397. switch (m_insertion_mode) {
  1398. case InsertionMode::InTable:
  1399. case InsertionMode::InCaption:
  1400. case InsertionMode::InTableBody:
  1401. case InsertionMode::InRow:
  1402. case InsertionMode::InCell:
  1403. m_insertion_mode = InsertionMode::InSelectInTable;
  1404. break;
  1405. default:
  1406. m_insertion_mode = InsertionMode::InSelect;
  1407. break;
  1408. }
  1409. return;
  1410. }
  1411. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1412. if (current_node().local_name() == HTML::TagNames::option)
  1413. m_stack_of_open_elements.pop();
  1414. reconstruct_the_active_formatting_elements();
  1415. insert_html_element(token);
  1416. return;
  1417. }
  1418. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1419. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1420. generate_implied_end_tags();
  1421. if (current_node().local_name() != HTML::TagNames::ruby)
  1422. log_parse_error();
  1423. insert_html_element(token);
  1424. return;
  1425. }
  1426. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1427. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1428. generate_implied_end_tags(HTML::TagNames::rtc);
  1429. if (current_node().local_name() != HTML::TagNames::rtc || current_node().local_name() != HTML::TagNames::ruby)
  1430. log_parse_error();
  1431. insert_html_element(token);
  1432. return;
  1433. }
  1434. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::math) {
  1435. reconstruct_the_active_formatting_elements();
  1436. adjust_mathml_attributes(token);
  1437. adjust_foreign_attributes(token);
  1438. insert_foreign_element(token, Namespace::MathML);
  1439. if (token.is_self_closing()) {
  1440. m_stack_of_open_elements.pop();
  1441. token.acknowledge_self_closing_flag_if_set();
  1442. }
  1443. return;
  1444. }
  1445. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::svg) {
  1446. reconstruct_the_active_formatting_elements();
  1447. adjust_svg_attributes(token);
  1448. adjust_foreign_attributes(token);
  1449. insert_foreign_element(token, Namespace::SVG);
  1450. if (token.is_self_closing()) {
  1451. m_stack_of_open_elements.pop();
  1452. token.acknowledge_self_closing_flag_if_set();
  1453. }
  1454. return;
  1455. }
  1456. 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))) {
  1457. log_parse_error();
  1458. return;
  1459. }
  1460. // Any other start tag
  1461. if (token.is_start_tag()) {
  1462. reconstruct_the_active_formatting_elements();
  1463. insert_html_element(token);
  1464. return;
  1465. }
  1466. if (token.is_end_tag()) {
  1467. AnyOtherEndTag:
  1468. RefPtr<DOM::Element> node;
  1469. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1470. node = m_stack_of_open_elements.elements()[i];
  1471. if (node->local_name() == token.tag_name()) {
  1472. generate_implied_end_tags(token.tag_name());
  1473. if (node != current_node()) {
  1474. log_parse_error();
  1475. }
  1476. while (&current_node() != node) {
  1477. m_stack_of_open_elements.pop();
  1478. }
  1479. m_stack_of_open_elements.pop();
  1480. break;
  1481. }
  1482. if (is_special_tag(node->local_name(), node->namespace_())) {
  1483. log_parse_error();
  1484. return;
  1485. }
  1486. }
  1487. return;
  1488. }
  1489. }
  1490. void HTMLDocumentParser::adjust_mathml_attributes(HTMLToken& token)
  1491. {
  1492. token.adjust_attribute_name("definitionurl", "definitionURL");
  1493. }
  1494. void HTMLDocumentParser::adjust_svg_tag_names(HTMLToken& token)
  1495. {
  1496. token.adjust_tag_name("altglyph", "altGlyph");
  1497. token.adjust_tag_name("altglyphdef", "altGlyphDef");
  1498. token.adjust_tag_name("altglyphitem", "altGlyphItem");
  1499. token.adjust_tag_name("animatecolor", "animateColor");
  1500. token.adjust_tag_name("animatemotion", "animateMotion");
  1501. token.adjust_tag_name("animatetransform", "animateTransform");
  1502. token.adjust_tag_name("clippath", "clipPath");
  1503. token.adjust_tag_name("feblend", "feBlend");
  1504. token.adjust_tag_name("fecolormatrix", "feColorMatrix");
  1505. token.adjust_tag_name("fecomponenttransfer", "feComponentTransfer");
  1506. token.adjust_tag_name("fecomposite", "feComposite");
  1507. token.adjust_tag_name("feconvolvematrix", "feConvolveMatrix");
  1508. token.adjust_tag_name("fediffuselighting", "feDiffuseLighting");
  1509. token.adjust_tag_name("fedisplacementmap", "feDisplacementMap");
  1510. token.adjust_tag_name("fedistantlight", "feDistantLight");
  1511. token.adjust_tag_name("fedropshadow", "feDropShadow");
  1512. token.adjust_tag_name("feflood", "feFlood");
  1513. token.adjust_tag_name("fefunca", "feFuncA");
  1514. token.adjust_tag_name("fefuncb", "feFuncB");
  1515. token.adjust_tag_name("fefuncg", "feFuncG");
  1516. token.adjust_tag_name("fefuncr", "feFuncR");
  1517. token.adjust_tag_name("fegaussianblur", "feGaussianBlur");
  1518. token.adjust_tag_name("feimage", "feImage");
  1519. token.adjust_tag_name("femerge", "feMerge");
  1520. token.adjust_tag_name("femergenode", "feMergeNode");
  1521. token.adjust_tag_name("femorphology", "feMorphology");
  1522. token.adjust_tag_name("feoffset", "feOffset");
  1523. token.adjust_tag_name("fepointlight", "fePointLight");
  1524. token.adjust_tag_name("fespecularlighting", "feSpecularLighting");
  1525. token.adjust_tag_name("fespotlight", "feSpotlight");
  1526. token.adjust_tag_name("glyphref", "glyphRef");
  1527. token.adjust_tag_name("lineargradient", "linearGradient");
  1528. token.adjust_tag_name("radialgradient", "radialGradient");
  1529. token.adjust_tag_name("textpath", "textPath");
  1530. }
  1531. void HTMLDocumentParser::adjust_svg_attributes(HTMLToken& token)
  1532. {
  1533. token.adjust_attribute_name("attributename", "attributeName");
  1534. token.adjust_attribute_name("attributetype", "attributeType");
  1535. token.adjust_attribute_name("basefrequency", "baseFrequency");
  1536. token.adjust_attribute_name("baseprofile", "baseProfile");
  1537. token.adjust_attribute_name("calcmode", "calcMode");
  1538. token.adjust_attribute_name("clippathunits", "clipPathUnits");
  1539. token.adjust_attribute_name("diffuseconstant", "diffuseConstant");
  1540. token.adjust_attribute_name("edgemode", "edgeMode");
  1541. token.adjust_attribute_name("filterunits", "filterUnits");
  1542. token.adjust_attribute_name("glyphref", "glyphRef");
  1543. token.adjust_attribute_name("gradienttransform", "gradientTransform");
  1544. token.adjust_attribute_name("gradientunits", "gradientUnits");
  1545. token.adjust_attribute_name("kernelmatrix", "kernelMatrix");
  1546. token.adjust_attribute_name("kernelunitlength", "kernelUnitLength");
  1547. token.adjust_attribute_name("keypoints", "keyPoints");
  1548. token.adjust_attribute_name("keysplines", "keySplines");
  1549. token.adjust_attribute_name("keytimes", "keyTimes");
  1550. token.adjust_attribute_name("lengthadjust", "lengthAdjust");
  1551. token.adjust_attribute_name("limitingconeangle", "limitingConeAngle");
  1552. token.adjust_attribute_name("markerheight", "markerHeight");
  1553. token.adjust_attribute_name("markerunits", "markerUnits");
  1554. token.adjust_attribute_name("markerwidth", "markerWidth");
  1555. token.adjust_attribute_name("maskcontentunits", "maskContentUnits");
  1556. token.adjust_attribute_name("maskunits", "maskUnits");
  1557. token.adjust_attribute_name("numoctaves", "numOctaves");
  1558. token.adjust_attribute_name("pathlength", "pathLength");
  1559. token.adjust_attribute_name("patterncontentunits", "patternContentUnits");
  1560. token.adjust_attribute_name("patterntransform", "patternTransform");
  1561. token.adjust_attribute_name("patternunits", "patternUnits");
  1562. token.adjust_attribute_name("pointsatx", "pointsAtX");
  1563. token.adjust_attribute_name("pointsaty", "pointsAtY");
  1564. token.adjust_attribute_name("pointsatz", "pointsAtZ");
  1565. token.adjust_attribute_name("preservealpha", "preserveAlpha");
  1566. token.adjust_attribute_name("preserveaspectratio", "preserveAspectRatio");
  1567. token.adjust_attribute_name("primitiveunits", "primitiveUnits");
  1568. token.adjust_attribute_name("refx", "refX");
  1569. token.adjust_attribute_name("refy", "refY");
  1570. token.adjust_attribute_name("repeatcount", "repeatCount");
  1571. token.adjust_attribute_name("repeatdur", "repeatDur");
  1572. token.adjust_attribute_name("requiredextensions", "requiredExtensions");
  1573. token.adjust_attribute_name("requiredfeatures", "requiredFeatures");
  1574. token.adjust_attribute_name("specularconstant", "specularConstant");
  1575. token.adjust_attribute_name("specularexponent", "specularExponent");
  1576. token.adjust_attribute_name("spreadmethod", "spreadMethod");
  1577. token.adjust_attribute_name("startoffset", "startOffset");
  1578. token.adjust_attribute_name("stddeviation", "stdDeviation");
  1579. token.adjust_attribute_name("stitchtiles", "stitchTiles");
  1580. token.adjust_attribute_name("surfacescale", "surfaceScale");
  1581. token.adjust_attribute_name("systemlanguage", "systemLanguage");
  1582. token.adjust_attribute_name("tablevalues", "tableValues");
  1583. token.adjust_attribute_name("targetx", "targetX");
  1584. token.adjust_attribute_name("targety", "targetY");
  1585. token.adjust_attribute_name("textlength", "textLength");
  1586. token.adjust_attribute_name("viewbox", "viewBox");
  1587. token.adjust_attribute_name("viewtarget", "viewTarget");
  1588. token.adjust_attribute_name("xchannelselector", "xChannelSelector");
  1589. token.adjust_attribute_name("ychannelselector", "yChannelSelector");
  1590. token.adjust_attribute_name("zoomandpan", "zoomAndPan");
  1591. }
  1592. void HTMLDocumentParser::adjust_foreign_attributes(HTMLToken& token)
  1593. {
  1594. token.adjust_foreign_attribute("xlink:actuate", "xlink", "actuate", Namespace::XLink);
  1595. token.adjust_foreign_attribute("xlink:arcrole", "xlink", "arcrole", Namespace::XLink);
  1596. token.adjust_foreign_attribute("xlink:href", "xlink", "href", Namespace::XLink);
  1597. token.adjust_foreign_attribute("xlink:role", "xlink", "role", Namespace::XLink);
  1598. token.adjust_foreign_attribute("xlink:show", "xlink", "show", Namespace::XLink);
  1599. token.adjust_foreign_attribute("xlink:title", "xlink", "title", Namespace::XLink);
  1600. token.adjust_foreign_attribute("xlink:type", "xlink", "type", Namespace::XLink);
  1601. token.adjust_foreign_attribute("xml:lang", "xml", "lang", Namespace::XML);
  1602. token.adjust_foreign_attribute("xml:space", "xml", "space", Namespace::XML);
  1603. token.adjust_foreign_attribute("xmlns", "", "xmlns", Namespace::XMLNS);
  1604. token.adjust_foreign_attribute("xmlns:xlink", "xmlns", "xlink", Namespace::XMLNS);
  1605. }
  1606. void HTMLDocumentParser::increment_script_nesting_level()
  1607. {
  1608. ++m_script_nesting_level;
  1609. }
  1610. void HTMLDocumentParser::decrement_script_nesting_level()
  1611. {
  1612. VERIFY(m_script_nesting_level);
  1613. --m_script_nesting_level;
  1614. }
  1615. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1616. {
  1617. if (token.is_character()) {
  1618. insert_character(token.code_point());
  1619. return;
  1620. }
  1621. if (token.is_end_of_file()) {
  1622. log_parse_error();
  1623. if (current_node().local_name() == HTML::TagNames::script)
  1624. verify_cast<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1625. m_stack_of_open_elements.pop();
  1626. m_insertion_mode = m_original_insertion_mode;
  1627. process_using_the_rules_for(m_insertion_mode, token);
  1628. return;
  1629. }
  1630. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1631. // Make sure the <script> element has up-to-date text content before preparing the script.
  1632. flush_character_insertions();
  1633. NonnullRefPtr<HTMLScriptElement> script = verify_cast<HTMLScriptElement>(current_node());
  1634. m_stack_of_open_elements.pop();
  1635. m_insertion_mode = m_original_insertion_mode;
  1636. // FIXME: Handle tokenizer insertion point stuff here.
  1637. increment_script_nesting_level();
  1638. script->prepare_script({});
  1639. decrement_script_nesting_level();
  1640. if (script_nesting_level() == 0)
  1641. m_parser_pause_flag = false;
  1642. // FIXME: Handle tokenizer insertion point stuff here too.
  1643. while (document().pending_parsing_blocking_script()) {
  1644. if (script_nesting_level() != 0) {
  1645. m_parser_pause_flag = true;
  1646. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1647. // yielding control back to the caller. (Tokenization will resume when
  1648. // the caller returns to the "outer" tree construction stage.)
  1649. TODO();
  1650. } else {
  1651. auto the_script = document().take_pending_parsing_blocking_script({});
  1652. m_tokenizer.set_blocked(true);
  1653. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1654. // or the script's "ready to be parser-executed" flag is not set:
  1655. // spin the event loop until the parser's Document has no style sheet
  1656. // that is blocking scripts and the script's "ready to be parser-executed"
  1657. // flag is set.
  1658. if (the_script->failed_to_load())
  1659. return;
  1660. VERIFY(the_script->is_ready_to_be_parser_executed());
  1661. if (m_aborted)
  1662. return;
  1663. m_tokenizer.set_blocked(false);
  1664. // FIXME: Handle tokenizer insertion point stuff here too.
  1665. VERIFY(script_nesting_level() == 0);
  1666. increment_script_nesting_level();
  1667. the_script->execute_script();
  1668. decrement_script_nesting_level();
  1669. VERIFY(script_nesting_level() == 0);
  1670. m_parser_pause_flag = false;
  1671. // FIXME: Handle tokenizer insertion point stuff here too.
  1672. }
  1673. }
  1674. return;
  1675. }
  1676. if (token.is_end_tag()) {
  1677. m_stack_of_open_elements.pop();
  1678. m_insertion_mode = m_original_insertion_mode;
  1679. return;
  1680. }
  1681. TODO();
  1682. }
  1683. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1684. {
  1685. while (!current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1686. m_stack_of_open_elements.pop();
  1687. if (current_node().local_name() == HTML::TagNames::html)
  1688. VERIFY(m_parsing_fragment);
  1689. }
  1690. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1691. {
  1692. while (!current_node().local_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1693. m_stack_of_open_elements.pop();
  1694. if (current_node().local_name() == HTML::TagNames::html)
  1695. VERIFY(m_parsing_fragment);
  1696. }
  1697. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1698. {
  1699. while (!current_node().local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1700. m_stack_of_open_elements.pop();
  1701. if (current_node().local_name() == HTML::TagNames::html)
  1702. VERIFY(m_parsing_fragment);
  1703. }
  1704. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1705. {
  1706. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1707. clear_the_stack_back_to_a_table_row_context();
  1708. insert_html_element(token);
  1709. m_insertion_mode = InsertionMode::InCell;
  1710. m_list_of_active_formatting_elements.add_marker();
  1711. return;
  1712. }
  1713. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1714. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1715. log_parse_error();
  1716. return;
  1717. }
  1718. clear_the_stack_back_to_a_table_row_context();
  1719. m_stack_of_open_elements.pop();
  1720. m_insertion_mode = InsertionMode::InTableBody;
  1721. return;
  1722. }
  1723. 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))
  1724. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1725. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1726. log_parse_error();
  1727. return;
  1728. }
  1729. clear_the_stack_back_to_a_table_row_context();
  1730. m_stack_of_open_elements.pop();
  1731. m_insertion_mode = InsertionMode::InTableBody;
  1732. process_using_the_rules_for(m_insertion_mode, token);
  1733. return;
  1734. }
  1735. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1736. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1737. log_parse_error();
  1738. return;
  1739. }
  1740. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1741. return;
  1742. }
  1743. clear_the_stack_back_to_a_table_row_context();
  1744. m_stack_of_open_elements.pop();
  1745. m_insertion_mode = InsertionMode::InTableBody;
  1746. process_using_the_rules_for(m_insertion_mode, token);
  1747. return;
  1748. }
  1749. 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)) {
  1750. log_parse_error();
  1751. return;
  1752. }
  1753. process_using_the_rules_for(InsertionMode::InTable, token);
  1754. }
  1755. void HTMLDocumentParser::close_the_cell()
  1756. {
  1757. generate_implied_end_tags();
  1758. if (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1759. log_parse_error();
  1760. }
  1761. while (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1762. m_stack_of_open_elements.pop();
  1763. m_stack_of_open_elements.pop();
  1764. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1765. m_insertion_mode = InsertionMode::InRow;
  1766. }
  1767. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1768. {
  1769. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1770. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1771. log_parse_error();
  1772. return;
  1773. }
  1774. generate_implied_end_tags();
  1775. if (current_node().local_name() != token.tag_name()) {
  1776. log_parse_error();
  1777. }
  1778. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1779. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1780. m_insertion_mode = InsertionMode::InRow;
  1781. return;
  1782. }
  1783. 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)) {
  1784. 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)) {
  1785. VERIFY(m_parsing_fragment);
  1786. log_parse_error();
  1787. return;
  1788. }
  1789. close_the_cell();
  1790. process_using_the_rules_for(m_insertion_mode, token);
  1791. return;
  1792. }
  1793. 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)) {
  1794. log_parse_error();
  1795. return;
  1796. }
  1797. 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)) {
  1798. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1799. log_parse_error();
  1800. return;
  1801. }
  1802. close_the_cell();
  1803. // Reprocess the token.
  1804. process_using_the_rules_for(m_insertion_mode, token);
  1805. return;
  1806. }
  1807. process_using_the_rules_for(InsertionMode::InBody, token);
  1808. }
  1809. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1810. {
  1811. if (token.is_character()) {
  1812. if (token.code_point() == 0) {
  1813. log_parse_error();
  1814. return;
  1815. }
  1816. m_pending_table_character_tokens.append(token);
  1817. return;
  1818. }
  1819. for (auto& pending_token : m_pending_table_character_tokens) {
  1820. VERIFY(pending_token.is_character());
  1821. if (!pending_token.is_parser_whitespace()) {
  1822. // If any of the tokens in the pending table character tokens list
  1823. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1824. // reprocess the character tokens in the pending table character tokens list using
  1825. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1826. log_parse_error();
  1827. m_foster_parenting = true;
  1828. process_using_the_rules_for(InsertionMode::InBody, token);
  1829. m_foster_parenting = false;
  1830. return;
  1831. }
  1832. }
  1833. for (auto& pending_token : m_pending_table_character_tokens) {
  1834. insert_character(pending_token.code_point());
  1835. }
  1836. m_insertion_mode = m_original_insertion_mode;
  1837. process_using_the_rules_for(m_insertion_mode, token);
  1838. }
  1839. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1840. {
  1841. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1842. clear_the_stack_back_to_a_table_body_context();
  1843. insert_html_element(token);
  1844. m_insertion_mode = InsertionMode::InRow;
  1845. return;
  1846. }
  1847. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1848. log_parse_error();
  1849. clear_the_stack_back_to_a_table_body_context();
  1850. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tr));
  1851. m_insertion_mode = InsertionMode::InRow;
  1852. process_using_the_rules_for(m_insertion_mode, token);
  1853. return;
  1854. }
  1855. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1856. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1857. log_parse_error();
  1858. return;
  1859. }
  1860. clear_the_stack_back_to_a_table_body_context();
  1861. m_stack_of_open_elements.pop();
  1862. m_insertion_mode = InsertionMode::InTable;
  1863. return;
  1864. }
  1865. 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))
  1866. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1867. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tbody)
  1868. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::thead)
  1869. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tfoot)) {
  1870. log_parse_error();
  1871. return;
  1872. }
  1873. clear_the_stack_back_to_a_table_body_context();
  1874. m_stack_of_open_elements.pop();
  1875. m_insertion_mode = InsertionMode::InTable;
  1876. process_using_the_rules_for(InsertionMode::InTable, token);
  1877. return;
  1878. }
  1879. 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)) {
  1880. log_parse_error();
  1881. return;
  1882. }
  1883. process_using_the_rules_for(InsertionMode::InTable, token);
  1884. }
  1885. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1886. {
  1887. 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)) {
  1888. m_pending_table_character_tokens.clear();
  1889. m_original_insertion_mode = m_insertion_mode;
  1890. m_insertion_mode = InsertionMode::InTableText;
  1891. process_using_the_rules_for(InsertionMode::InTableText, token);
  1892. return;
  1893. }
  1894. if (token.is_comment()) {
  1895. insert_comment(token);
  1896. return;
  1897. }
  1898. if (token.is_doctype()) {
  1899. log_parse_error();
  1900. return;
  1901. }
  1902. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1903. clear_the_stack_back_to_a_table_context();
  1904. m_list_of_active_formatting_elements.add_marker();
  1905. insert_html_element(token);
  1906. m_insertion_mode = InsertionMode::InCaption;
  1907. return;
  1908. }
  1909. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1910. clear_the_stack_back_to_a_table_context();
  1911. insert_html_element(token);
  1912. m_insertion_mode = InsertionMode::InColumnGroup;
  1913. return;
  1914. }
  1915. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1916. clear_the_stack_back_to_a_table_context();
  1917. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::colgroup));
  1918. m_insertion_mode = InsertionMode::InColumnGroup;
  1919. process_using_the_rules_for(m_insertion_mode, token);
  1920. return;
  1921. }
  1922. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1923. clear_the_stack_back_to_a_table_context();
  1924. insert_html_element(token);
  1925. m_insertion_mode = InsertionMode::InTableBody;
  1926. return;
  1927. }
  1928. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1929. clear_the_stack_back_to_a_table_context();
  1930. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tbody));
  1931. m_insertion_mode = InsertionMode::InTableBody;
  1932. process_using_the_rules_for(m_insertion_mode, token);
  1933. return;
  1934. }
  1935. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1936. log_parse_error();
  1937. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table))
  1938. return;
  1939. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1940. reset_the_insertion_mode_appropriately();
  1941. process_using_the_rules_for(m_insertion_mode, token);
  1942. return;
  1943. }
  1944. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::table) {
  1945. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1946. log_parse_error();
  1947. return;
  1948. }
  1949. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1950. reset_the_insertion_mode_appropriately();
  1951. return;
  1952. }
  1953. 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)) {
  1954. log_parse_error();
  1955. return;
  1956. }
  1957. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::style, HTML::TagNames::script, HTML::TagNames::template_))
  1958. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_)) {
  1959. process_using_the_rules_for(InsertionMode::InHead, token);
  1960. return;
  1961. }
  1962. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1963. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1964. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1965. goto AnythingElse;
  1966. }
  1967. log_parse_error();
  1968. insert_html_element(token);
  1969. // FIXME: Is this the correct interpretation of "Pop that input element off the stack of open elements."?
  1970. // Because this wording is the first time it's seen in the spec.
  1971. // Other times it's worded as: "Immediately pop the current node off the stack of open elements."
  1972. m_stack_of_open_elements.pop();
  1973. token.acknowledge_self_closing_flag_if_set();
  1974. return;
  1975. }
  1976. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1977. log_parse_error();
  1978. if (m_form_element || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1979. return;
  1980. }
  1981. m_form_element = verify_cast<HTMLFormElement>(*insert_html_element(token));
  1982. // FIXME: See previous FIXME, as this is the same situation but for form.
  1983. m_stack_of_open_elements.pop();
  1984. return;
  1985. }
  1986. if (token.is_end_of_file()) {
  1987. process_using_the_rules_for(InsertionMode::InBody, token);
  1988. return;
  1989. }
  1990. AnythingElse:
  1991. log_parse_error();
  1992. m_foster_parenting = true;
  1993. process_using_the_rules_for(InsertionMode::InBody, token);
  1994. m_foster_parenting = false;
  1995. }
  1996. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1997. {
  1998. 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)) {
  1999. log_parse_error();
  2000. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2001. reset_the_insertion_mode_appropriately();
  2002. process_using_the_rules_for(m_insertion_mode, token);
  2003. return;
  2004. }
  2005. 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)) {
  2006. log_parse_error();
  2007. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name()))
  2008. return;
  2009. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2010. reset_the_insertion_mode_appropriately();
  2011. process_using_the_rules_for(m_insertion_mode, token);
  2012. return;
  2013. }
  2014. process_using_the_rules_for(InsertionMode::InSelect, token);
  2015. }
  2016. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  2017. {
  2018. if (token.is_character()) {
  2019. if (token.code_point() == 0) {
  2020. log_parse_error();
  2021. return;
  2022. }
  2023. insert_character(token.code_point());
  2024. return;
  2025. }
  2026. if (token.is_comment()) {
  2027. insert_comment(token);
  2028. return;
  2029. }
  2030. if (token.is_doctype()) {
  2031. log_parse_error();
  2032. return;
  2033. }
  2034. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2035. process_using_the_rules_for(InsertionMode::InBody, token);
  2036. return;
  2037. }
  2038. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  2039. if (current_node().local_name() == HTML::TagNames::option) {
  2040. m_stack_of_open_elements.pop();
  2041. }
  2042. insert_html_element(token);
  2043. return;
  2044. }
  2045. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2046. if (current_node().local_name() == HTML::TagNames::option) {
  2047. m_stack_of_open_elements.pop();
  2048. }
  2049. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2050. m_stack_of_open_elements.pop();
  2051. }
  2052. insert_html_element(token);
  2053. return;
  2054. }
  2055. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2056. if (current_node().local_name() == HTML::TagNames::option && node_before_current_node().local_name() == HTML::TagNames::optgroup)
  2057. m_stack_of_open_elements.pop();
  2058. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2059. m_stack_of_open_elements.pop();
  2060. } else {
  2061. log_parse_error();
  2062. return;
  2063. }
  2064. return;
  2065. }
  2066. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  2067. if (current_node().local_name() == HTML::TagNames::option) {
  2068. m_stack_of_open_elements.pop();
  2069. } else {
  2070. log_parse_error();
  2071. return;
  2072. }
  2073. return;
  2074. }
  2075. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  2076. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2077. VERIFY(m_parsing_fragment);
  2078. log_parse_error();
  2079. return;
  2080. }
  2081. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2082. reset_the_insertion_mode_appropriately();
  2083. return;
  2084. }
  2085. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  2086. log_parse_error();
  2087. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2088. VERIFY(m_parsing_fragment);
  2089. return;
  2090. }
  2091. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2092. reset_the_insertion_mode_appropriately();
  2093. return;
  2094. }
  2095. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  2096. log_parse_error();
  2097. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2098. VERIFY(m_parsing_fragment);
  2099. return;
  2100. }
  2101. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2102. reset_the_insertion_mode_appropriately();
  2103. process_using_the_rules_for(m_insertion_mode, token);
  2104. return;
  2105. }
  2106. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  2107. process_using_the_rules_for(InsertionMode::InHead, token);
  2108. return;
  2109. }
  2110. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2111. process_using_the_rules_for(InsertionMode::InHead, token);
  2112. return;
  2113. }
  2114. if (token.is_end_of_file()) {
  2115. process_using_the_rules_for(InsertionMode::InBody, token);
  2116. return;
  2117. }
  2118. log_parse_error();
  2119. }
  2120. void HTMLDocumentParser::handle_in_caption(HTMLToken& token)
  2121. {
  2122. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::caption) {
  2123. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2124. VERIFY(m_parsing_fragment);
  2125. log_parse_error();
  2126. return;
  2127. }
  2128. generate_implied_end_tags();
  2129. if (current_node().local_name() != HTML::TagNames::caption)
  2130. log_parse_error();
  2131. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2132. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2133. m_insertion_mode = InsertionMode::InTable;
  2134. return;
  2135. }
  2136. 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))
  2137. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  2138. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2139. VERIFY(m_parsing_fragment);
  2140. log_parse_error();
  2141. return;
  2142. }
  2143. generate_implied_end_tags();
  2144. if (current_node().local_name() != HTML::TagNames::caption)
  2145. log_parse_error();
  2146. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2147. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2148. m_insertion_mode = InsertionMode::InTable;
  2149. process_using_the_rules_for(m_insertion_mode, token);
  2150. return;
  2151. }
  2152. 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)) {
  2153. log_parse_error();
  2154. return;
  2155. }
  2156. process_using_the_rules_for(InsertionMode::InBody, token);
  2157. }
  2158. void HTMLDocumentParser::handle_in_column_group(HTMLToken& token)
  2159. {
  2160. if (token.is_character() && token.is_parser_whitespace()) {
  2161. insert_character(token.code_point());
  2162. return;
  2163. }
  2164. if (token.is_comment()) {
  2165. insert_comment(token);
  2166. return;
  2167. }
  2168. if (token.is_doctype()) {
  2169. log_parse_error();
  2170. return;
  2171. }
  2172. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2173. process_using_the_rules_for(InsertionMode::InBody, token);
  2174. return;
  2175. }
  2176. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2177. insert_html_element(token);
  2178. m_stack_of_open_elements.pop();
  2179. token.acknowledge_self_closing_flag_if_set();
  2180. return;
  2181. }
  2182. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  2183. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2184. log_parse_error();
  2185. return;
  2186. }
  2187. m_stack_of_open_elements.pop();
  2188. m_insertion_mode = InsertionMode::InTable;
  2189. return;
  2190. }
  2191. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::col) {
  2192. log_parse_error();
  2193. return;
  2194. }
  2195. if ((token.is_start_tag() || token.is_end_tag()) && token.tag_name() == HTML::TagNames::template_) {
  2196. process_using_the_rules_for(InsertionMode::InHead, token);
  2197. return;
  2198. }
  2199. if (token.is_end_of_file()) {
  2200. process_using_the_rules_for(InsertionMode::InBody, token);
  2201. return;
  2202. }
  2203. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2204. log_parse_error();
  2205. return;
  2206. }
  2207. m_stack_of_open_elements.pop();
  2208. m_insertion_mode = InsertionMode::InTable;
  2209. process_using_the_rules_for(m_insertion_mode, token);
  2210. }
  2211. void HTMLDocumentParser::handle_in_template(HTMLToken& token)
  2212. {
  2213. if (token.is_character() || token.is_comment() || token.is_doctype()) {
  2214. process_using_the_rules_for(InsertionMode::InBody, token);
  2215. return;
  2216. }
  2217. 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)) {
  2218. process_using_the_rules_for(InsertionMode::InHead, token);
  2219. return;
  2220. }
  2221. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2222. process_using_the_rules_for(InsertionMode::InHead, token);
  2223. return;
  2224. }
  2225. 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)) {
  2226. m_stack_of_template_insertion_modes.take_last();
  2227. m_stack_of_template_insertion_modes.append(InsertionMode::InTable);
  2228. m_insertion_mode = InsertionMode::InTable;
  2229. process_using_the_rules_for(m_insertion_mode, token);
  2230. return;
  2231. }
  2232. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2233. m_stack_of_template_insertion_modes.take_last();
  2234. m_stack_of_template_insertion_modes.append(InsertionMode::InColumnGroup);
  2235. m_insertion_mode = InsertionMode::InColumnGroup;
  2236. process_using_the_rules_for(m_insertion_mode, token);
  2237. return;
  2238. }
  2239. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  2240. m_stack_of_template_insertion_modes.take_last();
  2241. m_stack_of_template_insertion_modes.append(InsertionMode::InTableBody);
  2242. m_insertion_mode = InsertionMode::InTableBody;
  2243. process_using_the_rules_for(m_insertion_mode, token);
  2244. return;
  2245. }
  2246. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2247. m_stack_of_template_insertion_modes.take_last();
  2248. m_stack_of_template_insertion_modes.append(InsertionMode::InRow);
  2249. m_insertion_mode = InsertionMode::InRow;
  2250. process_using_the_rules_for(m_insertion_mode, token);
  2251. return;
  2252. }
  2253. if (token.is_start_tag()) {
  2254. m_stack_of_template_insertion_modes.take_last();
  2255. m_stack_of_template_insertion_modes.append(InsertionMode::InBody);
  2256. m_insertion_mode = InsertionMode::InBody;
  2257. process_using_the_rules_for(m_insertion_mode, token);
  2258. return;
  2259. }
  2260. if (token.is_end_tag()) {
  2261. log_parse_error();
  2262. return;
  2263. }
  2264. if (token.is_end_of_file()) {
  2265. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  2266. VERIFY(m_parsing_fragment);
  2267. stop_parsing();
  2268. return;
  2269. }
  2270. log_parse_error();
  2271. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  2272. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2273. m_stack_of_template_insertion_modes.take_last();
  2274. reset_the_insertion_mode_appropriately();
  2275. process_using_the_rules_for(m_insertion_mode, token);
  2276. }
  2277. }
  2278. void HTMLDocumentParser::handle_in_frameset(HTMLToken& token)
  2279. {
  2280. if (token.is_character() && token.is_parser_whitespace()) {
  2281. insert_character(token.code_point());
  2282. return;
  2283. }
  2284. if (token.is_comment()) {
  2285. insert_comment(token);
  2286. return;
  2287. }
  2288. if (token.is_doctype()) {
  2289. log_parse_error();
  2290. return;
  2291. }
  2292. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2293. process_using_the_rules_for(InsertionMode::InBody, token);
  2294. return;
  2295. }
  2296. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2297. insert_html_element(token);
  2298. return;
  2299. }
  2300. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2301. // FIXME: If the current node is the root html element, then this is a parse error; ignore the token. (fragment case)
  2302. m_stack_of_open_elements.pop();
  2303. if (!m_parsing_fragment && current_node().local_name() != HTML::TagNames::frameset) {
  2304. m_insertion_mode = InsertionMode::AfterFrameset;
  2305. }
  2306. return;
  2307. }
  2308. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frame) {
  2309. insert_html_element(token);
  2310. m_stack_of_open_elements.pop();
  2311. token.acknowledge_self_closing_flag_if_set();
  2312. return;
  2313. }
  2314. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2315. process_using_the_rules_for(InsertionMode::InHead, token);
  2316. return;
  2317. }
  2318. if (token.is_end_of_file()) {
  2319. //FIXME: If the current node is not the root html element, then this is a parse error.
  2320. stop_parsing();
  2321. return;
  2322. }
  2323. log_parse_error();
  2324. }
  2325. void HTMLDocumentParser::handle_after_frameset(HTMLToken& token)
  2326. {
  2327. if (token.is_character() && token.is_parser_whitespace()) {
  2328. insert_character(token.code_point());
  2329. return;
  2330. }
  2331. if (token.is_comment()) {
  2332. insert_comment(token);
  2333. return;
  2334. }
  2335. if (token.is_doctype()) {
  2336. log_parse_error();
  2337. return;
  2338. }
  2339. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2340. process_using_the_rules_for(InsertionMode::InBody, token);
  2341. return;
  2342. }
  2343. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  2344. m_insertion_mode = InsertionMode::AfterAfterFrameset;
  2345. return;
  2346. }
  2347. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2348. process_using_the_rules_for(InsertionMode::InHead, token);
  2349. return;
  2350. }
  2351. if (token.is_end_of_file()) {
  2352. stop_parsing();
  2353. return;
  2354. }
  2355. log_parse_error();
  2356. }
  2357. void HTMLDocumentParser::handle_after_after_frameset(HTMLToken& token)
  2358. {
  2359. if (token.is_comment()) {
  2360. auto comment = adopt_ref(*new DOM::Comment(document(), token.comment()));
  2361. document().append_child(move(comment));
  2362. return;
  2363. }
  2364. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  2365. process_using_the_rules_for(InsertionMode::InBody, token);
  2366. return;
  2367. }
  2368. if (token.is_end_of_file()) {
  2369. stop_parsing();
  2370. return;
  2371. }
  2372. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2373. process_using_the_rules_for(InsertionMode::InHead, token);
  2374. return;
  2375. }
  2376. log_parse_error();
  2377. }
  2378. void HTMLDocumentParser::process_using_the_rules_for_foreign_content(HTMLToken& token)
  2379. {
  2380. if (token.is_character()) {
  2381. if (token.code_point() == 0) {
  2382. log_parse_error();
  2383. insert_character(0xFFFD);
  2384. return;
  2385. }
  2386. if (token.is_parser_whitespace()) {
  2387. insert_character(token.code_point());
  2388. return;
  2389. }
  2390. insert_character(token.code_point());
  2391. m_frameset_ok = false;
  2392. return;
  2393. }
  2394. if (token.is_comment()) {
  2395. insert_comment(token);
  2396. return;
  2397. }
  2398. if (token.is_doctype()) {
  2399. log_parse_error();
  2400. return;
  2401. }
  2402. 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))
  2403. || (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)))) {
  2404. log_parse_error();
  2405. if (m_parsing_fragment) {
  2406. goto AnyOtherStartTag;
  2407. }
  2408. TODO();
  2409. }
  2410. if (token.is_start_tag()) {
  2411. AnyOtherStartTag:
  2412. if (adjusted_current_node().namespace_() == Namespace::MathML) {
  2413. adjust_mathml_attributes(token);
  2414. } else if (adjusted_current_node().namespace_() == Namespace::SVG) {
  2415. adjust_svg_tag_names(token);
  2416. adjust_svg_attributes(token);
  2417. }
  2418. adjust_foreign_attributes(token);
  2419. insert_foreign_element(token, adjusted_current_node().namespace_());
  2420. if (token.is_self_closing()) {
  2421. if (token.tag_name() == SVG::TagNames::script && current_node().namespace_() == Namespace::SVG) {
  2422. token.acknowledge_self_closing_flag_if_set();
  2423. goto ScriptEndTag;
  2424. }
  2425. m_stack_of_open_elements.pop();
  2426. token.acknowledge_self_closing_flag_if_set();
  2427. }
  2428. return;
  2429. }
  2430. if (token.is_end_tag() && current_node().namespace_() == Namespace::SVG && current_node().tag_name() == SVG::TagNames::script) {
  2431. ScriptEndTag:
  2432. m_stack_of_open_elements.pop();
  2433. TODO();
  2434. }
  2435. if (token.is_end_tag()) {
  2436. RefPtr<DOM::Element> node = current_node();
  2437. // FIXME: Not sure if this is the correct to_lowercase, as the specification says "to ASCII lowercase"
  2438. if (node->tag_name().to_lowercase() != token.tag_name())
  2439. log_parse_error();
  2440. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2441. if (node == m_stack_of_open_elements.first()) {
  2442. VERIFY(m_parsing_fragment);
  2443. return;
  2444. }
  2445. // FIXME: See the above FIXME
  2446. if (node->tag_name().to_lowercase() == token.tag_name()) {
  2447. while (current_node() != node)
  2448. m_stack_of_open_elements.pop();
  2449. m_stack_of_open_elements.pop();
  2450. return;
  2451. }
  2452. node = m_stack_of_open_elements.elements().at(i - 1);
  2453. if (node->namespace_() != Namespace::HTML)
  2454. continue;
  2455. process_using_the_rules_for(m_insertion_mode, token);
  2456. return;
  2457. }
  2458. }
  2459. VERIFY_NOT_REACHED();
  2460. }
  2461. // https://html.spec.whatwg.org/multipage/parsing.html#reset-the-insertion-mode-appropriately
  2462. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  2463. {
  2464. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2465. bool last = i == 0;
  2466. // NOTE: When parsing fragments, we substitute the context element for the root of the stack of open elements.
  2467. RefPtr<DOM::Element> node;
  2468. if (last && m_parsing_fragment) {
  2469. node = m_context_element;
  2470. } else {
  2471. node = m_stack_of_open_elements.elements().at(i);
  2472. }
  2473. if (node->local_name() == HTML::TagNames::select) {
  2474. if (!last) {
  2475. for (ssize_t j = i; j > 0; --j) {
  2476. auto& ancestor = m_stack_of_open_elements.elements().at(j - 1);
  2477. if (is<HTMLTemplateElement>(ancestor))
  2478. break;
  2479. if (is<HTMLTableElement>(ancestor)) {
  2480. m_insertion_mode = InsertionMode::InSelectInTable;
  2481. return;
  2482. }
  2483. }
  2484. }
  2485. m_insertion_mode = InsertionMode::InSelect;
  2486. return;
  2487. }
  2488. if (!last && node->local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2489. m_insertion_mode = InsertionMode::InCell;
  2490. return;
  2491. }
  2492. if (node->local_name() == HTML::TagNames::tr) {
  2493. m_insertion_mode = InsertionMode::InRow;
  2494. return;
  2495. }
  2496. if (node->local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  2497. m_insertion_mode = InsertionMode::InTableBody;
  2498. return;
  2499. }
  2500. if (node->local_name() == HTML::TagNames::caption) {
  2501. m_insertion_mode = InsertionMode::InCaption;
  2502. return;
  2503. }
  2504. if (node->local_name() == HTML::TagNames::colgroup) {
  2505. m_insertion_mode = InsertionMode::InColumnGroup;
  2506. return;
  2507. }
  2508. if (node->local_name() == HTML::TagNames::table) {
  2509. m_insertion_mode = InsertionMode::InTable;
  2510. return;
  2511. }
  2512. if (node->local_name() == HTML::TagNames::template_) {
  2513. m_insertion_mode = m_stack_of_template_insertion_modes.last();
  2514. return;
  2515. }
  2516. if (!last && node->local_name() == HTML::TagNames::head) {
  2517. m_insertion_mode = InsertionMode::InHead;
  2518. return;
  2519. }
  2520. if (node->local_name() == HTML::TagNames::body) {
  2521. m_insertion_mode = InsertionMode::InBody;
  2522. return;
  2523. }
  2524. if (node->local_name() == HTML::TagNames::frameset) {
  2525. VERIFY(m_parsing_fragment);
  2526. m_insertion_mode = InsertionMode::InFrameset;
  2527. return;
  2528. }
  2529. if (node->local_name() == HTML::TagNames::html) {
  2530. if (!m_head_element) {
  2531. VERIFY(m_parsing_fragment);
  2532. m_insertion_mode = InsertionMode::BeforeHead;
  2533. return;
  2534. }
  2535. m_insertion_mode = InsertionMode::AfterHead;
  2536. return;
  2537. }
  2538. }
  2539. VERIFY(m_parsing_fragment);
  2540. m_insertion_mode = InsertionMode::InBody;
  2541. }
  2542. const char* HTMLDocumentParser::insertion_mode_name() const
  2543. {
  2544. switch (m_insertion_mode) {
  2545. #define __ENUMERATE_INSERTION_MODE(mode) \
  2546. case InsertionMode::mode: \
  2547. return #mode;
  2548. ENUMERATE_INSERTION_MODES
  2549. #undef __ENUMERATE_INSERTION_MODE
  2550. }
  2551. VERIFY_NOT_REACHED();
  2552. }
  2553. DOM::Document& HTMLDocumentParser::document()
  2554. {
  2555. return *m_document;
  2556. }
  2557. NonnullRefPtrVector<DOM::Node> HTMLDocumentParser::parse_html_fragment(DOM::Element& context_element, const StringView& markup)
  2558. {
  2559. auto temp_document = DOM::Document::create();
  2560. HTMLDocumentParser parser(*temp_document, markup, "utf-8");
  2561. parser.m_context_element = context_element;
  2562. parser.m_parsing_fragment = true;
  2563. parser.document().set_quirks_mode(context_element.document().mode());
  2564. if (context_element.local_name().is_one_of(HTML::TagNames::title, HTML::TagNames::textarea)) {
  2565. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  2566. } else if (context_element.local_name().is_one_of(HTML::TagNames::style, HTML::TagNames::xmp, HTML::TagNames::iframe, HTML::TagNames::noembed, HTML::TagNames::noframes)) {
  2567. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2568. } else if (context_element.local_name().is_one_of(HTML::TagNames::script)) {
  2569. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  2570. } else if (context_element.local_name().is_one_of(HTML::TagNames::noscript)) {
  2571. if (context_element.document().is_scripting_enabled())
  2572. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2573. } else if (context_element.local_name().is_one_of(HTML::TagNames::plaintext)) {
  2574. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  2575. }
  2576. auto root = create_element(context_element.document(), HTML::TagNames::html, Namespace::HTML);
  2577. parser.document().append_child(root);
  2578. parser.m_stack_of_open_elements.push(root);
  2579. if (context_element.local_name() == HTML::TagNames::template_) {
  2580. parser.m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  2581. }
  2582. // FIXME: Create a start tag token whose name is the local name of context and whose attributes are the attributes of context.
  2583. parser.reset_the_insertion_mode_appropriately();
  2584. for (auto* form_candidate = &context_element; form_candidate; form_candidate = form_candidate->parent_element()) {
  2585. if (is<HTMLFormElement>(*form_candidate)) {
  2586. parser.m_form_element = verify_cast<HTMLFormElement>(*form_candidate);
  2587. break;
  2588. }
  2589. }
  2590. parser.run(context_element.document().url());
  2591. NonnullRefPtrVector<DOM::Node> children;
  2592. while (RefPtr<DOM::Node> child = root->first_child()) {
  2593. root->remove_child(*child);
  2594. context_element.document().adopt_node(*child);
  2595. children.append(*child);
  2596. }
  2597. return children;
  2598. }
  2599. NonnullOwnPtr<HTMLDocumentParser> HTMLDocumentParser::create_with_uncertain_encoding(DOM::Document& document, const ByteBuffer& input)
  2600. {
  2601. if (document.has_encoding())
  2602. return make<HTMLDocumentParser>(document, input, document.encoding().value());
  2603. auto encoding = run_encoding_sniffing_algorithm(input);
  2604. dbgln("The encoding sniffing algorithm returned encoding '{}'", encoding);
  2605. return make<HTMLDocumentParser>(document, input, encoding);
  2606. }
  2607. }