HTMLDocumentParser.cpp 107 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658165916601661166216631664166516661667166816691670167116721673167416751676167716781679168016811682168316841685168616871688168916901691169216931694169516961697169816991700170117021703170417051706170717081709171017111712171317141715171617171718171917201721172217231724172517261727172817291730173117321733173417351736173717381739174017411742174317441745174617471748174917501751175217531754175517561757175817591760176117621763176417651766176717681769177017711772177317741775177617771778177917801781178217831784178517861787178817891790179117921793179417951796179717981799180018011802180318041805180618071808180918101811181218131814181518161817181818191820182118221823182418251826182718281829183018311832183318341835183618371838183918401841184218431844184518461847184818491850185118521853185418551856185718581859186018611862186318641865186618671868186918701871187218731874187518761877187818791880188118821883188418851886188718881889189018911892189318941895189618971898189919001901190219031904190519061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192919301931193219331934193519361937193819391940194119421943194419451946194719481949195019511952195319541955195619571958195919601961196219631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994199519961997199819992000200120022003200420052006200720082009201020112012201320142015201620172018201920202021202220232024202520262027202820292030203120322033203420352036203720382039204020412042204320442045204620472048204920502051205220532054205520562057205820592060206120622063206420652066206720682069207020712072207320742075207620772078207920802081208220832084208520862087208820892090209120922093209420952096209720982099210021012102210321042105210621072108210921102111211221132114211521162117211821192120212121222123212421252126212721282129213021312132213321342135213621372138213921402141214221432144214521462147214821492150215121522153215421552156215721582159216021612162216321642165216621672168216921702171217221732174217521762177217821792180218121822183218421852186218721882189219021912192219321942195219621972198219922002201220222032204220522062207220822092210221122122213221422152216221722182219222022212222222322242225222622272228222922302231223222332234223522362237223822392240224122422243224422452246224722482249225022512252225322542255225622572258225922602261226222632264226522662267226822692270227122722273227422752276227722782279228022812282228322842285228622872288228922902291229222932294229522962297229822992300230123022303230423052306230723082309231023112312231323142315231623172318231923202321232223232324232523262327232823292330233123322333233423352336233723382339234023412342234323442345234623472348234923502351235223532354235523562357235823592360236123622363236423652366236723682369237023712372237323742375237623772378237923802381238223832384238523862387238823892390239123922393239423952396239723982399240024012402240324042405240624072408240924102411241224132414241524162417241824192420242124222423242424252426242724282429243024312432243324342435243624372438243924402441244224432444244524462447244824492450245124522453245424552456245724582459246024612462246324642465246624672468246924702471247224732474247524762477247824792480248124822483248424852486248724882489249024912492249324942495249624972498249925002501250225032504250525062507250825092510251125122513251425152516251725182519252025212522252325242525252625272528252925302531253225332534253525362537253825392540254125422543254425452546254725482549255025512552255325542555255625572558255925602561256225632564256525662567256825692570257125722573257425752576257725782579258025812582258325842585258625872588258925902591259225932594259525962597259825992600260126022603260426052606260726082609261026112612261326142615261626172618261926202621262226232624262526262627262826292630263126322633263426352636263726382639264026412642264326442645264626472648264926502651265226532654265526562657265826592660266126622663266426652666266726682669267026712672267326742675267626772678267926802681268226832684268526862687268826892690269126922693269426952696269726982699270027012702270327042705270627072708270927102711271227132714271527162717271827192720272127222723272427252726272727282729273027312732273327342735273627372738273927402741274227432744274527462747274827492750275127522753275427552756275727582759276027612762276327642765276627672768276927702771277227732774277527762777277827792780278127822783278427852786278727882789279027912792279327942795279627972798279928002801280228032804280528062807280828092810281128122813281428152816281728182819282028212822282328242825282628272828282928302831283228332834283528362837283828392840284128422843284428452846
  1. /*
  2. * Copyright (c) 2020, Andreas Kling <kling@serenityos.org>
  3. * All rights reserved.
  4. *
  5. * Redistribution and use in source and binary forms, with or without
  6. * modification, are permitted provided that the following conditions are met:
  7. *
  8. * 1. Redistributions of source code must retain the above copyright notice, this
  9. * list of conditions and the following disclaimer.
  10. *
  11. * 2. Redistributions in binary form must reproduce the above copyright notice,
  12. * this list of conditions and the following disclaimer in the documentation
  13. * and/or other materials provided with the distribution.
  14. *
  15. * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
  16. * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  17. * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
  18. * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
  19. * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  20. * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
  21. * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
  22. * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
  23. * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  24. * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  25. */
  26. //#define PARSER_DEBUG
  27. #include <AK/Utf32View.h>
  28. #include <LibWeb/DOM/Comment.h>
  29. #include <LibWeb/DOM/Document.h>
  30. #include <LibWeb/DOM/DocumentType.h>
  31. #include <LibWeb/DOM/ElementFactory.h>
  32. #include <LibWeb/DOM/Event.h>
  33. #include <LibWeb/DOM/Text.h>
  34. #include <LibWeb/HTML/HTMLFormElement.h>
  35. #include <LibWeb/HTML/HTMLHeadElement.h>
  36. #include <LibWeb/HTML/HTMLScriptElement.h>
  37. #include <LibWeb/HTML/HTMLTemplateElement.h>
  38. #include <LibWeb/HTML/Parser/HTMLDocumentParser.h>
  39. #include <LibWeb/HTML/Parser/HTMLToken.h>
  40. namespace Web::HTML {
  41. #define PARSE_ERROR() \
  42. do { \
  43. dbg() << "Parse error! " << __PRETTY_FUNCTION__ << " @ " << __LINE__; \
  44. } while (0)
  45. static Vector<FlyString> s_quirks_public_ids = {
  46. "+//Silmaril//dtd html Pro v0r11 19970101//",
  47. "-//AS//DTD HTML 3.0 asWedit + extensions//",
  48. "-//AdvaSoft Ltd//DTD HTML 3.0 asWedit + extensions//",
  49. "-//IETF//DTD HTML 2.0 Level 1//",
  50. "-//IETF//DTD HTML 2.0 Level 2//",
  51. "-//IETF//DTD HTML 2.0 Strict Level 1//",
  52. "-//IETF//DTD HTML 2.0 Strict Level 2//",
  53. "-//IETF//DTD HTML 2.0 Strict//",
  54. "-//IETF//DTD HTML 2.0//",
  55. "-//IETF//DTD HTML 2.1E//",
  56. "-//IETF//DTD HTML 3.0//",
  57. "-//IETF//DTD HTML 3.2 Final//",
  58. "-//IETF//DTD HTML 3.2//",
  59. "-//IETF//DTD HTML 3//",
  60. "-//IETF//DTD HTML Level 0//",
  61. "-//IETF//DTD HTML Level 1//",
  62. "-//IETF//DTD HTML Level 2//",
  63. "-//IETF//DTD HTML Level 3//",
  64. "-//IETF//DTD HTML Strict Level 0//",
  65. "-//IETF//DTD HTML Strict Level 1//",
  66. "-//IETF//DTD HTML Strict Level 2//",
  67. "-//IETF//DTD HTML Strict Level 3//",
  68. "-//IETF//DTD HTML Strict//",
  69. "-//IETF//DTD HTML//",
  70. "-//Metrius//DTD Metrius Presentational//",
  71. "-//Microsoft//DTD Internet Explorer 2.0 HTML Strict//",
  72. "-//Microsoft//DTD Internet Explorer 2.0 HTML//",
  73. "-//Microsoft//DTD Internet Explorer 2.0 Tables//",
  74. "-//Microsoft//DTD Internet Explorer 3.0 HTML Strict//",
  75. "-//Microsoft//DTD Internet Explorer 3.0 HTML//",
  76. "-//Microsoft//DTD Internet Explorer 3.0 Tables//",
  77. "-//Netscape Comm. Corp.//DTD HTML//",
  78. "-//Netscape Comm. Corp.//DTD Strict HTML//",
  79. "-//O'Reilly and Associates//DTD HTML 2.0//",
  80. "-//O'Reilly and Associates//DTD HTML Extended 1.0//",
  81. "-//O'Reilly and Associates//DTD HTML Extended Relaxed 1.0//",
  82. "-//SQ//DTD HTML 2.0 HoTMetaL + extensions//",
  83. "-//SoftQuad Software//DTD HoTMetaL PRO 6.0::19990601::extensions to HTML 4.0//",
  84. "-//SoftQuad//DTD HoTMetaL PRO 4.0::19971010::extensions to HTML 4.0//",
  85. "-//Spyglass//DTD HTML 2.0 Extended//",
  86. "-//Sun Microsystems Corp.//DTD HotJava HTML//",
  87. "-//Sun Microsystems Corp.//DTD HotJava Strict HTML//",
  88. "-//W3C//DTD HTML 3 1995-03-24//",
  89. "-//W3C//DTD HTML 3.2 Draft//",
  90. "-//W3C//DTD HTML 3.2 Final//",
  91. "-//W3C//DTD HTML 3.2//",
  92. "-//W3C//DTD HTML 3.2S Draft//",
  93. "-//W3C//DTD HTML 4.0 Frameset//",
  94. "-//W3C//DTD HTML 4.0 Transitional//",
  95. "-//W3C//DTD HTML Experimental 19960712//",
  96. "-//W3C//DTD HTML Experimental 970421//",
  97. "-//W3C//DTD W3 HTML//",
  98. "-//W3O//DTD W3 HTML 3.0//",
  99. "-//WebTechs//DTD Mozilla HTML 2.0//",
  100. "-//WebTechs//DTD Mozilla HTML//"
  101. };
  102. RefPtr<DOM::Document> parse_html_document(const StringView& data, const URL& url, const String& encoding)
  103. {
  104. HTMLDocumentParser parser(data, encoding);
  105. parser.run(url);
  106. return parser.document();
  107. }
  108. HTMLDocumentParser::HTMLDocumentParser(const StringView& input, const String& encoding)
  109. : m_tokenizer(input, encoding)
  110. {
  111. m_document = adopt(*new DOM::Document);
  112. }
  113. HTMLDocumentParser::HTMLDocumentParser(const StringView& input, const String& encoding, DOM::Document& existing_document)
  114. : m_tokenizer(input, encoding)
  115. , m_document(existing_document)
  116. {
  117. }
  118. HTMLDocumentParser::~HTMLDocumentParser()
  119. {
  120. }
  121. void HTMLDocumentParser::run(const URL& url)
  122. {
  123. m_document->set_url(url);
  124. m_document->set_source(m_tokenizer.source());
  125. for (;;) {
  126. auto optional_token = m_tokenizer.next_token();
  127. if (!optional_token.has_value())
  128. break;
  129. auto& token = optional_token.value();
  130. #ifdef PARSER_DEBUG
  131. dbg() << "[" << insertion_mode_name() << "] " << token.to_string();
  132. #endif
  133. process_using_the_rules_for(m_insertion_mode, token);
  134. if (m_stop_parsing) {
  135. #ifdef PARSER_DEBUG
  136. dbg() << "Stop parsing" << (m_parsing_fragment ? " fragment" : "") << "! :^)";
  137. #endif
  138. break;
  139. }
  140. }
  141. flush_character_insertions();
  142. // "The end"
  143. auto scripts_to_execute_when_parsing_has_finished = m_document->take_scripts_to_execute_when_parsing_has_finished({});
  144. for (auto& script : scripts_to_execute_when_parsing_has_finished) {
  145. script.execute_script();
  146. }
  147. m_document->dispatch_event(DOM::Event::create("DOMContentLoaded"));
  148. auto scripts_to_execute_as_soon_as_possible = m_document->take_scripts_to_execute_as_soon_as_possible({});
  149. for (auto& script : scripts_to_execute_as_soon_as_possible) {
  150. script.execute_script();
  151. }
  152. }
  153. void HTMLDocumentParser::process_using_the_rules_for(InsertionMode mode, HTMLToken& token)
  154. {
  155. switch (mode) {
  156. case InsertionMode::Initial:
  157. handle_initial(token);
  158. break;
  159. case InsertionMode::BeforeHTML:
  160. handle_before_html(token);
  161. break;
  162. case InsertionMode::BeforeHead:
  163. handle_before_head(token);
  164. break;
  165. case InsertionMode::InHead:
  166. handle_in_head(token);
  167. break;
  168. case InsertionMode::InHeadNoscript:
  169. handle_in_head_noscript(token);
  170. break;
  171. case InsertionMode::AfterHead:
  172. handle_after_head(token);
  173. break;
  174. case InsertionMode::InBody:
  175. handle_in_body(token);
  176. break;
  177. case InsertionMode::AfterBody:
  178. handle_after_body(token);
  179. break;
  180. case InsertionMode::AfterAfterBody:
  181. handle_after_after_body(token);
  182. break;
  183. case InsertionMode::Text:
  184. handle_text(token);
  185. break;
  186. case InsertionMode::InTable:
  187. handle_in_table(token);
  188. break;
  189. case InsertionMode::InTableBody:
  190. handle_in_table_body(token);
  191. break;
  192. case InsertionMode::InRow:
  193. handle_in_row(token);
  194. break;
  195. case InsertionMode::InCell:
  196. handle_in_cell(token);
  197. break;
  198. case InsertionMode::InTableText:
  199. handle_in_table_text(token);
  200. break;
  201. case InsertionMode::InSelectInTable:
  202. handle_in_select_in_table(token);
  203. break;
  204. case InsertionMode::InSelect:
  205. handle_in_select(token);
  206. break;
  207. case InsertionMode::InCaption:
  208. handle_in_caption(token);
  209. break;
  210. case InsertionMode::InColumnGroup:
  211. handle_in_column_group(token);
  212. break;
  213. case InsertionMode::InTemplate:
  214. handle_in_template(token);
  215. break;
  216. case InsertionMode::InFrameset:
  217. handle_in_frameset(token);
  218. break;
  219. case InsertionMode::AfterFrameset:
  220. handle_after_frameset(token);
  221. break;
  222. case InsertionMode::AfterAfterFrameset:
  223. handle_after_after_frameset(token);
  224. break;
  225. default:
  226. ASSERT_NOT_REACHED();
  227. }
  228. }
  229. DOM::QuirksMode HTMLDocumentParser::which_quirks_mode(const HTMLToken& doctype_token) const
  230. {
  231. if (doctype_token.m_doctype.force_quirks)
  232. return DOM::QuirksMode::Yes;
  233. // NOTE: The tokenizer puts the name into lower case for us.
  234. if (doctype_token.m_doctype.name.to_string() != "html")
  235. return DOM::QuirksMode::Yes;
  236. auto public_identifier = doctype_token.m_doctype.public_identifier.to_string();
  237. auto system_identifier = doctype_token.m_doctype.system_identifier.to_string();
  238. if (public_identifier.equals_ignoring_case("-//W3O//DTD W3 HTML Strict 3.0//EN//"))
  239. return DOM::QuirksMode::Yes;
  240. if (public_identifier.equals_ignoring_case("-/W3C/DTD HTML 4.0 Transitional/EN"))
  241. return DOM::QuirksMode::Yes;
  242. if (public_identifier.equals_ignoring_case("HTML"))
  243. return DOM::QuirksMode::Yes;
  244. if (system_identifier.equals_ignoring_case("http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd"))
  245. return DOM::QuirksMode::Yes;
  246. for (auto& public_id : s_quirks_public_ids) {
  247. if (public_identifier.starts_with(public_id, CaseSensitivity::CaseInsensitive))
  248. return DOM::QuirksMode::Yes;
  249. }
  250. if (doctype_token.m_doctype.missing_system_identifier) {
  251. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  252. return DOM::QuirksMode::Yes;
  253. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  254. return DOM::QuirksMode::Yes;
  255. }
  256. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Frameset//", CaseSensitivity::CaseInsensitive))
  257. return DOM::QuirksMode::Limited;
  258. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Transitional//", CaseSensitivity::CaseInsensitive))
  259. return DOM::QuirksMode::Limited;
  260. if (!doctype_token.m_doctype.missing_system_identifier) {
  261. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  262. return DOM::QuirksMode::Limited;
  263. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  264. return DOM::QuirksMode::Limited;
  265. }
  266. return DOM::QuirksMode::No;
  267. }
  268. void HTMLDocumentParser::handle_initial(HTMLToken& token)
  269. {
  270. if (token.is_character() && token.is_parser_whitespace()) {
  271. return;
  272. }
  273. if (token.is_comment()) {
  274. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  275. document().append_child(move(comment));
  276. return;
  277. }
  278. if (token.is_doctype()) {
  279. auto doctype = adopt(*new DOM::DocumentType(document()));
  280. doctype->set_name(token.m_doctype.name.to_string());
  281. doctype->set_public_id(token.m_doctype.public_identifier.to_string());
  282. doctype->set_system_id(token.m_doctype.system_identifier.to_string());
  283. document().append_child(move(doctype));
  284. document().set_quirks_mode(which_quirks_mode(token));
  285. m_insertion_mode = InsertionMode::BeforeHTML;
  286. return;
  287. }
  288. PARSE_ERROR();
  289. document().set_quirks_mode(DOM::QuirksMode::Yes);
  290. m_insertion_mode = InsertionMode::BeforeHTML;
  291. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  292. }
  293. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  294. {
  295. if (token.is_doctype()) {
  296. PARSE_ERROR();
  297. return;
  298. }
  299. if (token.is_comment()) {
  300. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  301. document().append_child(move(comment));
  302. return;
  303. }
  304. if (token.is_character() && token.is_parser_whitespace()) {
  305. return;
  306. }
  307. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  308. auto element = create_element_for(token);
  309. document().append_child(element);
  310. m_stack_of_open_elements.push(move(element));
  311. m_insertion_mode = InsertionMode::BeforeHead;
  312. return;
  313. }
  314. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  315. goto AnythingElse;
  316. }
  317. if (token.is_end_tag()) {
  318. PARSE_ERROR();
  319. return;
  320. }
  321. AnythingElse:
  322. auto element = create_element(document(), HTML::TagNames::html);
  323. document().append_child(element);
  324. m_stack_of_open_elements.push(element);
  325. // 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.
  326. m_insertion_mode = InsertionMode::BeforeHead;
  327. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  328. return;
  329. }
  330. DOM::Element& HTMLDocumentParser::current_node()
  331. {
  332. return m_stack_of_open_elements.current_node();
  333. }
  334. DOM::Element& HTMLDocumentParser::node_before_current_node()
  335. {
  336. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  337. }
  338. HTMLDocumentParser::AdjustedInsertionLocation HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  339. {
  340. auto& target = current_node();
  341. HTMLDocumentParser::AdjustedInsertionLocation adjusted_insertion_location;
  342. 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)) {
  343. auto last_template = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::template_);
  344. auto last_table = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::table);
  345. if (last_template.element && (!last_table.element || last_template.index > last_table.index)) {
  346. // This returns the template content, so no need to check the parent is a template.
  347. return { downcast<HTMLTemplateElement>(last_template.element)->content(), nullptr };
  348. }
  349. if (!last_table.element) {
  350. ASSERT(m_parsing_fragment);
  351. // Guaranteed not to be a template element (it will be the html element),
  352. // so no need to check the parent is a template.
  353. return { m_stack_of_open_elements.elements().first(), nullptr };
  354. }
  355. if (last_table.element->parent_node())
  356. adjusted_insertion_location = { last_table.element->parent_node(), last_table.element };
  357. else
  358. adjusted_insertion_location = { m_stack_of_open_elements.element_before(*last_table.element), nullptr };
  359. } else {
  360. adjusted_insertion_location = { target, nullptr };
  361. }
  362. if (is<HTMLTemplateElement>(*adjusted_insertion_location.parent))
  363. return { downcast<HTMLTemplateElement>(*adjusted_insertion_location.parent).content(), nullptr };
  364. return adjusted_insertion_location;
  365. }
  366. NonnullRefPtr<DOM::Element> HTMLDocumentParser::create_element_for(const HTMLToken& token)
  367. {
  368. auto element = create_element(document(), token.tag_name());
  369. for (auto& attribute : token.m_tag.attributes) {
  370. element->set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  371. }
  372. return element;
  373. }
  374. RefPtr<DOM::Element> HTMLDocumentParser::insert_html_element(const HTMLToken& token)
  375. {
  376. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  377. auto element = create_element_for(token);
  378. // FIXME: Check if it's possible to insert `element` at `adjusted_insertion_location`
  379. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling);
  380. m_stack_of_open_elements.push(element);
  381. return element;
  382. }
  383. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  384. {
  385. if (token.is_character() && token.is_parser_whitespace()) {
  386. return;
  387. }
  388. if (token.is_comment()) {
  389. insert_comment(token);
  390. return;
  391. }
  392. if (token.is_doctype()) {
  393. PARSE_ERROR();
  394. return;
  395. }
  396. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  397. process_using_the_rules_for(InsertionMode::InBody, token);
  398. return;
  399. }
  400. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  401. auto element = insert_html_element(token);
  402. m_head_element = downcast<HTMLHeadElement>(*element);
  403. m_insertion_mode = InsertionMode::InHead;
  404. return;
  405. }
  406. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  407. goto AnythingElse;
  408. }
  409. if (token.is_end_tag()) {
  410. PARSE_ERROR();
  411. return;
  412. }
  413. AnythingElse:
  414. m_head_element = downcast<HTMLHeadElement>(*insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::head)));
  415. m_insertion_mode = InsertionMode::InHead;
  416. process_using_the_rules_for(InsertionMode::InHead, token);
  417. return;
  418. }
  419. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  420. {
  421. auto data = token.m_comment_or_character.data.to_string();
  422. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  423. adjusted_insertion_location.parent->insert_before(adopt(*new DOM::Comment(document(), data)), adjusted_insertion_location.insert_before_sibling);
  424. }
  425. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  426. {
  427. if (token.is_parser_whitespace()) {
  428. insert_character(token.code_point());
  429. return;
  430. }
  431. if (token.is_comment()) {
  432. insert_comment(token);
  433. return;
  434. }
  435. if (token.is_doctype()) {
  436. PARSE_ERROR();
  437. return;
  438. }
  439. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  440. process_using_the_rules_for(InsertionMode::InBody, token);
  441. return;
  442. }
  443. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  444. insert_html_element(token);
  445. m_stack_of_open_elements.pop();
  446. token.acknowledge_self_closing_flag_if_set();
  447. return;
  448. }
  449. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  450. auto element = insert_html_element(token);
  451. m_stack_of_open_elements.pop();
  452. token.acknowledge_self_closing_flag_if_set();
  453. return;
  454. }
  455. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  456. insert_html_element(token);
  457. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  458. m_original_insertion_mode = m_insertion_mode;
  459. m_insertion_mode = InsertionMode::Text;
  460. return;
  461. }
  462. 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)) {
  463. parse_generic_raw_text_element(token);
  464. return;
  465. }
  466. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noscript && !m_scripting_enabled) {
  467. insert_html_element(token);
  468. m_insertion_mode = InsertionMode::InHeadNoscript;
  469. return;
  470. }
  471. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  472. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  473. auto element = create_element_for(token);
  474. auto& script_element = downcast<HTMLScriptElement>(*element);
  475. script_element.set_parser_document({}, document());
  476. script_element.set_non_blocking({}, false);
  477. if (m_parsing_fragment) {
  478. TODO();
  479. }
  480. if (m_invoked_via_document_write) {
  481. TODO();
  482. }
  483. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling, false);
  484. m_stack_of_open_elements.push(element);
  485. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  486. m_original_insertion_mode = m_insertion_mode;
  487. m_insertion_mode = InsertionMode::Text;
  488. return;
  489. }
  490. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  491. m_stack_of_open_elements.pop();
  492. m_insertion_mode = InsertionMode::AfterHead;
  493. return;
  494. }
  495. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  496. goto AnythingElse;
  497. }
  498. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  499. insert_html_element(token);
  500. m_list_of_active_formatting_elements.add_marker();
  501. m_frameset_ok = false;
  502. m_insertion_mode = InsertionMode::InTemplate;
  503. m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  504. return;
  505. }
  506. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  507. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  508. PARSE_ERROR();
  509. return;
  510. }
  511. generate_all_implied_end_tags_thoroughly();
  512. if (current_node().local_name() != HTML::TagNames::template_)
  513. PARSE_ERROR();
  514. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  515. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  516. m_stack_of_template_insertion_modes.take_last();
  517. reset_the_insertion_mode_appropriately();
  518. return;
  519. }
  520. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  521. PARSE_ERROR();
  522. return;
  523. }
  524. AnythingElse:
  525. m_stack_of_open_elements.pop();
  526. m_insertion_mode = InsertionMode::AfterHead;
  527. process_using_the_rules_for(m_insertion_mode, token);
  528. }
  529. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken& token)
  530. {
  531. if (token.is_doctype()) {
  532. PARSE_ERROR();
  533. return;
  534. }
  535. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  536. process_using_the_rules_for(InsertionMode::InBody, token);
  537. return;
  538. }
  539. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::noscript) {
  540. m_stack_of_open_elements.pop();
  541. m_insertion_mode = InsertionMode::InHead;
  542. return;
  543. }
  544. 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))) {
  545. process_using_the_rules_for(InsertionMode::InHead, token);
  546. return;
  547. }
  548. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  549. goto AnythingElse;
  550. }
  551. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::noscript)) {
  552. PARSE_ERROR();
  553. return;
  554. }
  555. AnythingElse:
  556. PARSE_ERROR();
  557. m_stack_of_open_elements.pop();
  558. m_insertion_mode = InsertionMode::InHead;
  559. process_using_the_rules_for(m_insertion_mode, token);
  560. }
  561. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  562. {
  563. insert_html_element(token);
  564. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  565. m_original_insertion_mode = m_insertion_mode;
  566. m_insertion_mode = InsertionMode::Text;
  567. }
  568. DOM::Text* HTMLDocumentParser::find_character_insertion_node()
  569. {
  570. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  571. if (adjusted_insertion_location.insert_before_sibling) {
  572. TODO();
  573. }
  574. if (adjusted_insertion_location.parent->is_document())
  575. return nullptr;
  576. if (adjusted_insertion_location.parent->last_child() && adjusted_insertion_location.parent->last_child()->is_text())
  577. return downcast<DOM::Text>(adjusted_insertion_location.parent->last_child());
  578. auto new_text_node = adopt(*new DOM::Text(document(), ""));
  579. adjusted_insertion_location.parent->append_child(new_text_node);
  580. return new_text_node;
  581. }
  582. void HTMLDocumentParser::flush_character_insertions()
  583. {
  584. if (m_character_insertion_builder.is_empty())
  585. return;
  586. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  587. m_character_insertion_node->parent()->children_changed();
  588. m_character_insertion_builder.clear();
  589. }
  590. void HTMLDocumentParser::insert_character(u32 data)
  591. {
  592. auto node = find_character_insertion_node();
  593. if (node == m_character_insertion_node) {
  594. m_character_insertion_builder.append(Utf32View { &data, 1 });
  595. return;
  596. }
  597. if (!m_character_insertion_node) {
  598. m_character_insertion_node = node;
  599. m_character_insertion_builder.append(Utf32View { &data, 1 });
  600. return;
  601. }
  602. flush_character_insertions();
  603. m_character_insertion_node = node;
  604. m_character_insertion_builder.append(Utf32View { &data, 1 });
  605. }
  606. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  607. {
  608. if (token.is_character() && token.is_parser_whitespace()) {
  609. insert_character(token.code_point());
  610. return;
  611. }
  612. if (token.is_comment()) {
  613. insert_comment(token);
  614. return;
  615. }
  616. if (token.is_doctype()) {
  617. PARSE_ERROR();
  618. return;
  619. }
  620. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  621. process_using_the_rules_for(InsertionMode::InBody, token);
  622. return;
  623. }
  624. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  625. insert_html_element(token);
  626. m_frameset_ok = false;
  627. m_insertion_mode = InsertionMode::InBody;
  628. return;
  629. }
  630. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  631. insert_html_element(token);
  632. m_insertion_mode = InsertionMode::InFrameset;
  633. return;
  634. }
  635. 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)) {
  636. PARSE_ERROR();
  637. m_stack_of_open_elements.push(*m_head_element);
  638. process_using_the_rules_for(InsertionMode::InHead, token);
  639. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  640. return entry.ptr() == m_head_element;
  641. });
  642. return;
  643. }
  644. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  645. process_using_the_rules_for(InsertionMode::InHead, token);
  646. return;
  647. }
  648. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  649. goto AnythingElse;
  650. }
  651. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  652. PARSE_ERROR();
  653. return;
  654. }
  655. AnythingElse:
  656. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::body));
  657. m_insertion_mode = InsertionMode::InBody;
  658. process_using_the_rules_for(m_insertion_mode, token);
  659. }
  660. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  661. {
  662. 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))
  663. m_stack_of_open_elements.pop();
  664. }
  665. void HTMLDocumentParser::generate_all_implied_end_tags_thoroughly()
  666. {
  667. 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))
  668. m_stack_of_open_elements.pop();
  669. }
  670. void HTMLDocumentParser::close_a_p_element()
  671. {
  672. generate_implied_end_tags(HTML::TagNames::p);
  673. if (current_node().local_name() != HTML::TagNames::p) {
  674. PARSE_ERROR();
  675. }
  676. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  677. }
  678. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  679. {
  680. if (token.is_character() && token.is_parser_whitespace()) {
  681. process_using_the_rules_for(InsertionMode::InBody, token);
  682. return;
  683. }
  684. if (token.is_comment()) {
  685. auto data = token.m_comment_or_character.data.to_string();
  686. auto& insertion_location = m_stack_of_open_elements.first();
  687. insertion_location.append_child(adopt(*new DOM::Comment(document(), data)));
  688. return;
  689. }
  690. if (token.is_doctype()) {
  691. PARSE_ERROR();
  692. return;
  693. }
  694. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  695. process_using_the_rules_for(InsertionMode::InBody, token);
  696. return;
  697. }
  698. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  699. if (m_parsing_fragment) {
  700. PARSE_ERROR();
  701. return;
  702. }
  703. m_insertion_mode = InsertionMode::AfterAfterBody;
  704. return;
  705. }
  706. if (token.is_end_of_file()) {
  707. stop_parsing();
  708. return;
  709. }
  710. PARSE_ERROR();
  711. m_insertion_mode = InsertionMode::InBody;
  712. process_using_the_rules_for(InsertionMode::InBody, token);
  713. }
  714. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  715. {
  716. if (token.is_comment()) {
  717. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  718. document().append_child(move(comment));
  719. return;
  720. }
  721. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  722. process_using_the_rules_for(InsertionMode::InBody, token);
  723. return;
  724. }
  725. if (token.is_end_of_file()) {
  726. stop_parsing();
  727. return;
  728. }
  729. PARSE_ERROR();
  730. m_insertion_mode = InsertionMode::InBody;
  731. process_using_the_rules_for(m_insertion_mode, token);
  732. }
  733. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  734. {
  735. // FIXME: This needs to care about "markers"
  736. if (m_list_of_active_formatting_elements.is_empty())
  737. return;
  738. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  739. return;
  740. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  741. return;
  742. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  743. RefPtr<DOM::Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  744. ASSERT(entry);
  745. Rewind:
  746. if (index == 0) {
  747. goto Create;
  748. }
  749. --index;
  750. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  751. ASSERT(entry);
  752. if (!m_stack_of_open_elements.contains(*entry))
  753. goto Rewind;
  754. Advance:
  755. ++index;
  756. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  757. ASSERT(entry);
  758. Create:
  759. // FIXME: Hold on to the real token!
  760. auto new_element = insert_html_element(HTMLToken::make_start_tag(entry->local_name()));
  761. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  762. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  763. goto Advance;
  764. }
  765. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  766. {
  767. auto subject = token.tag_name();
  768. // If the current node is an HTML element whose tag name is subject,
  769. // and the current node is not in the list of active formatting elements,
  770. // then pop the current node off the stack of open elements, and return.
  771. if (current_node().local_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  772. m_stack_of_open_elements.pop();
  773. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  774. }
  775. size_t outer_loop_counter = 0;
  776. //OuterLoop:
  777. if (outer_loop_counter >= 8)
  778. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  779. ++outer_loop_counter;
  780. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  781. if (!formatting_element)
  782. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  783. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  784. PARSE_ERROR();
  785. // FIXME: If formatting element is not in the stack of open elements,
  786. // then this is a parse error; remove the element from the list, and return.
  787. TODO();
  788. }
  789. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  790. PARSE_ERROR();
  791. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  792. }
  793. if (formatting_element != &current_node()) {
  794. PARSE_ERROR();
  795. }
  796. RefPtr<DOM::Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  797. if (!furthest_block) {
  798. while (&current_node() != formatting_element)
  799. m_stack_of_open_elements.pop();
  800. m_stack_of_open_elements.pop();
  801. m_list_of_active_formatting_elements.remove(*formatting_element);
  802. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  803. }
  804. // FIXME: Implement the rest of the AAA :^)
  805. TODO();
  806. }
  807. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name)
  808. {
  809. return tag_name.is_one_of(
  810. HTML::TagNames::address,
  811. HTML::TagNames::applet,
  812. HTML::TagNames::area,
  813. HTML::TagNames::article,
  814. HTML::TagNames::aside,
  815. HTML::TagNames::base,
  816. HTML::TagNames::basefont,
  817. HTML::TagNames::bgsound,
  818. HTML::TagNames::blockquote,
  819. HTML::TagNames::body,
  820. HTML::TagNames::br,
  821. HTML::TagNames::button,
  822. HTML::TagNames::caption,
  823. HTML::TagNames::center,
  824. HTML::TagNames::col,
  825. HTML::TagNames::colgroup,
  826. HTML::TagNames::dd,
  827. HTML::TagNames::details,
  828. HTML::TagNames::dir,
  829. HTML::TagNames::div,
  830. HTML::TagNames::dl,
  831. HTML::TagNames::dt,
  832. HTML::TagNames::embed,
  833. HTML::TagNames::fieldset,
  834. HTML::TagNames::figcaption,
  835. HTML::TagNames::figure,
  836. HTML::TagNames::footer,
  837. HTML::TagNames::form,
  838. HTML::TagNames::frame,
  839. HTML::TagNames::frameset,
  840. HTML::TagNames::h1,
  841. HTML::TagNames::h2,
  842. HTML::TagNames::h3,
  843. HTML::TagNames::h4,
  844. HTML::TagNames::h5,
  845. HTML::TagNames::h6,
  846. HTML::TagNames::head,
  847. HTML::TagNames::header,
  848. HTML::TagNames::hgroup,
  849. HTML::TagNames::hr,
  850. HTML::TagNames::html,
  851. HTML::TagNames::iframe,
  852. HTML::TagNames::img,
  853. HTML::TagNames::input,
  854. HTML::TagNames::keygen,
  855. HTML::TagNames::li,
  856. HTML::TagNames::link,
  857. HTML::TagNames::listing,
  858. HTML::TagNames::main,
  859. HTML::TagNames::marquee,
  860. HTML::TagNames::menu,
  861. HTML::TagNames::meta,
  862. HTML::TagNames::nav,
  863. HTML::TagNames::noembed,
  864. HTML::TagNames::noframes,
  865. HTML::TagNames::noscript,
  866. HTML::TagNames::object,
  867. HTML::TagNames::ol,
  868. HTML::TagNames::p,
  869. HTML::TagNames::param,
  870. HTML::TagNames::plaintext,
  871. HTML::TagNames::pre,
  872. HTML::TagNames::script,
  873. HTML::TagNames::section,
  874. HTML::TagNames::select,
  875. HTML::TagNames::source,
  876. HTML::TagNames::style,
  877. HTML::TagNames::summary,
  878. HTML::TagNames::table,
  879. HTML::TagNames::tbody,
  880. HTML::TagNames::td,
  881. HTML::TagNames::template_,
  882. HTML::TagNames::textarea,
  883. HTML::TagNames::tfoot,
  884. HTML::TagNames::th,
  885. HTML::TagNames::thead,
  886. HTML::TagNames::title,
  887. HTML::TagNames::tr,
  888. HTML::TagNames::track,
  889. HTML::TagNames::ul,
  890. HTML::TagNames::wbr,
  891. HTML::TagNames::xmp);
  892. }
  893. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  894. {
  895. if (token.is_character()) {
  896. if (token.code_point() == 0) {
  897. PARSE_ERROR();
  898. return;
  899. }
  900. if (token.is_parser_whitespace()) {
  901. reconstruct_the_active_formatting_elements();
  902. insert_character(token.code_point());
  903. return;
  904. }
  905. reconstruct_the_active_formatting_elements();
  906. insert_character(token.code_point());
  907. m_frameset_ok = false;
  908. return;
  909. }
  910. if (token.is_comment()) {
  911. insert_comment(token);
  912. return;
  913. }
  914. if (token.is_doctype()) {
  915. PARSE_ERROR();
  916. return;
  917. }
  918. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  919. PARSE_ERROR();
  920. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  921. return;
  922. for (auto& attribute : token.m_tag.attributes) {
  923. if (current_node().has_attribute(attribute.local_name_builder.string_view()))
  924. continue;
  925. current_node().set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  926. }
  927. return;
  928. }
  929. 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)) {
  930. process_using_the_rules_for(InsertionMode::InHead, token);
  931. return;
  932. }
  933. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  934. process_using_the_rules_for(InsertionMode::InHead, token);
  935. return;
  936. }
  937. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  938. PARSE_ERROR();
  939. if (m_stack_of_open_elements.elements().size() == 1
  940. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body
  941. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  942. ASSERT(m_parsing_fragment);
  943. return;
  944. }
  945. m_frameset_ok = false;
  946. auto& body_element = m_stack_of_open_elements.elements().at(1);
  947. for (auto& attribute : token.m_tag.attributes) {
  948. if (body_element.has_attribute(attribute.local_name_builder.string_view()))
  949. continue;
  950. body_element.set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  951. }
  952. return;
  953. }
  954. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  955. PARSE_ERROR();
  956. if (m_stack_of_open_elements.elements().size() == 1
  957. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body) {
  958. ASSERT(m_parsing_fragment);
  959. return;
  960. }
  961. if (!m_frameset_ok)
  962. return;
  963. TODO();
  964. }
  965. if (token.is_end_of_file()) {
  966. if (!m_stack_of_template_insertion_modes.is_empty()) {
  967. process_using_the_rules_for(InsertionMode::InTemplate, token);
  968. return;
  969. }
  970. for (auto& node : m_stack_of_open_elements.elements()) {
  971. 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)) {
  972. PARSE_ERROR();
  973. break;
  974. }
  975. }
  976. stop_parsing();
  977. return;
  978. }
  979. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  980. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  981. PARSE_ERROR();
  982. return;
  983. }
  984. for (auto& node : m_stack_of_open_elements.elements()) {
  985. 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)) {
  986. PARSE_ERROR();
  987. break;
  988. }
  989. }
  990. m_insertion_mode = InsertionMode::AfterBody;
  991. return;
  992. }
  993. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  994. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  995. PARSE_ERROR();
  996. return;
  997. }
  998. for (auto& node : m_stack_of_open_elements.elements()) {
  999. 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)) {
  1000. PARSE_ERROR();
  1001. break;
  1002. }
  1003. }
  1004. m_insertion_mode = InsertionMode::AfterBody;
  1005. process_using_the_rules_for(m_insertion_mode, token);
  1006. return;
  1007. }
  1008. 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)) {
  1009. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1010. close_a_p_element();
  1011. insert_html_element(token);
  1012. return;
  1013. }
  1014. 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)) {
  1015. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1016. close_a_p_element();
  1017. 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)) {
  1018. PARSE_ERROR();
  1019. m_stack_of_open_elements.pop();
  1020. }
  1021. insert_html_element(token);
  1022. return;
  1023. }
  1024. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  1025. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1026. close_a_p_element();
  1027. insert_html_element(token);
  1028. m_frameset_ok = false;
  1029. // If the next token is a U+000A LINE FEED (LF) character token,
  1030. // then ignore that token and move on to the next one.
  1031. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1032. auto next_token = m_tokenizer.next_token();
  1033. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1034. // Ignore it.
  1035. } else {
  1036. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1037. }
  1038. return;
  1039. }
  1040. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1041. if (m_form_element && !m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1042. PARSE_ERROR();
  1043. return;
  1044. }
  1045. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1046. close_a_p_element();
  1047. auto element = insert_html_element(token);
  1048. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  1049. m_form_element = downcast<HTMLFormElement>(*element);
  1050. return;
  1051. }
  1052. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  1053. m_frameset_ok = false;
  1054. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1055. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1056. if (node->local_name() == HTML::TagNames::li) {
  1057. generate_implied_end_tags(HTML::TagNames::li);
  1058. if (current_node().local_name() != HTML::TagNames::li) {
  1059. PARSE_ERROR();
  1060. }
  1061. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1062. break;
  1063. }
  1064. if (is_special_tag(node->local_name()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1065. break;
  1066. }
  1067. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1068. close_a_p_element();
  1069. insert_html_element(token);
  1070. return;
  1071. }
  1072. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1073. m_frameset_ok = false;
  1074. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1075. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1076. if (node->local_name() == HTML::TagNames::dd) {
  1077. generate_implied_end_tags(HTML::TagNames::dd);
  1078. if (current_node().local_name() != HTML::TagNames::dd) {
  1079. PARSE_ERROR();
  1080. }
  1081. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  1082. break;
  1083. }
  1084. if (node->local_name() == HTML::TagNames::dt) {
  1085. generate_implied_end_tags(HTML::TagNames::dt);
  1086. if (current_node().local_name() != HTML::TagNames::dt) {
  1087. PARSE_ERROR();
  1088. }
  1089. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  1090. break;
  1091. }
  1092. if (is_special_tag(node->local_name()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1093. break;
  1094. }
  1095. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1096. close_a_p_element();
  1097. insert_html_element(token);
  1098. return;
  1099. }
  1100. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  1101. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1102. close_a_p_element();
  1103. insert_html_element(token);
  1104. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  1105. return;
  1106. }
  1107. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  1108. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  1109. PARSE_ERROR();
  1110. generate_implied_end_tags();
  1111. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  1112. }
  1113. reconstruct_the_active_formatting_elements();
  1114. insert_html_element(token);
  1115. m_frameset_ok = false;
  1116. return;
  1117. }
  1118. 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)) {
  1119. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1120. PARSE_ERROR();
  1121. return;
  1122. }
  1123. generate_implied_end_tags();
  1124. if (current_node().local_name() != token.tag_name()) {
  1125. PARSE_ERROR();
  1126. }
  1127. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1128. return;
  1129. }
  1130. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  1131. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1132. auto node = m_form_element;
  1133. m_form_element = nullptr;
  1134. if (!node || !m_stack_of_open_elements.has_in_scope(*node)) {
  1135. PARSE_ERROR();
  1136. return;
  1137. }
  1138. generate_implied_end_tags();
  1139. if (&current_node() != node) {
  1140. PARSE_ERROR();
  1141. }
  1142. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  1143. } else {
  1144. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  1145. PARSE_ERROR();
  1146. return;
  1147. }
  1148. generate_implied_end_tags();
  1149. if (current_node().local_name() != HTML::TagNames::form) {
  1150. PARSE_ERROR();
  1151. }
  1152. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  1153. }
  1154. return;
  1155. }
  1156. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  1157. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1158. PARSE_ERROR();
  1159. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::p));
  1160. }
  1161. close_a_p_element();
  1162. return;
  1163. }
  1164. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  1165. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  1166. PARSE_ERROR();
  1167. return;
  1168. }
  1169. generate_implied_end_tags(HTML::TagNames::li);
  1170. if (current_node().local_name() != HTML::TagNames::li) {
  1171. PARSE_ERROR();
  1172. dbg() << "Expected <li> current node, but had <" << current_node().local_name() << ">";
  1173. }
  1174. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1175. return;
  1176. }
  1177. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1178. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1179. PARSE_ERROR();
  1180. return;
  1181. }
  1182. generate_implied_end_tags(token.tag_name());
  1183. if (current_node().local_name() != token.tag_name()) {
  1184. PARSE_ERROR();
  1185. }
  1186. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1187. return;
  1188. }
  1189. 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)) {
  1190. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  1191. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  1192. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  1193. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  1194. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  1195. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  1196. PARSE_ERROR();
  1197. return;
  1198. }
  1199. generate_implied_end_tags();
  1200. if (current_node().local_name() != token.tag_name()) {
  1201. PARSE_ERROR();
  1202. }
  1203. for (;;) {
  1204. auto popped_element = m_stack_of_open_elements.pop();
  1205. 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))
  1206. break;
  1207. }
  1208. return;
  1209. }
  1210. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1211. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1212. PARSE_ERROR();
  1213. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1214. goto AnyOtherEndTag;
  1215. m_list_of_active_formatting_elements.remove(*element);
  1216. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1217. return entry.ptr() == element;
  1218. });
  1219. }
  1220. reconstruct_the_active_formatting_elements();
  1221. auto element = insert_html_element(token);
  1222. m_list_of_active_formatting_elements.add(*element);
  1223. return;
  1224. }
  1225. 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)) {
  1226. reconstruct_the_active_formatting_elements();
  1227. auto element = insert_html_element(token);
  1228. m_list_of_active_formatting_elements.add(*element);
  1229. return;
  1230. }
  1231. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1232. reconstruct_the_active_formatting_elements();
  1233. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1234. PARSE_ERROR();
  1235. run_the_adoption_agency_algorithm(token);
  1236. reconstruct_the_active_formatting_elements();
  1237. }
  1238. auto element = insert_html_element(token);
  1239. m_list_of_active_formatting_elements.add(*element);
  1240. return;
  1241. }
  1242. 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)) {
  1243. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1244. goto AnyOtherEndTag;
  1245. return;
  1246. }
  1247. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1248. reconstruct_the_active_formatting_elements();
  1249. insert_html_element(token);
  1250. m_list_of_active_formatting_elements.add_marker();
  1251. m_frameset_ok = false;
  1252. return;
  1253. }
  1254. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1255. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1256. PARSE_ERROR();
  1257. return;
  1258. }
  1259. generate_implied_end_tags();
  1260. if (current_node().local_name() != token.tag_name()) {
  1261. PARSE_ERROR();
  1262. }
  1263. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1264. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1265. return;
  1266. }
  1267. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1268. if (!document().in_quirks_mode()) {
  1269. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1270. close_a_p_element();
  1271. }
  1272. insert_html_element(token);
  1273. m_frameset_ok = false;
  1274. m_insertion_mode = InsertionMode::InTable;
  1275. return;
  1276. }
  1277. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1278. token.drop_attributes();
  1279. goto BRStartTag;
  1280. }
  1281. 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)) {
  1282. BRStartTag:
  1283. reconstruct_the_active_formatting_elements();
  1284. insert_html_element(token);
  1285. m_stack_of_open_elements.pop();
  1286. token.acknowledge_self_closing_flag_if_set();
  1287. m_frameset_ok = false;
  1288. return;
  1289. }
  1290. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1291. reconstruct_the_active_formatting_elements();
  1292. insert_html_element(token);
  1293. m_stack_of_open_elements.pop();
  1294. token.acknowledge_self_closing_flag_if_set();
  1295. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1296. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1297. m_frameset_ok = false;
  1298. }
  1299. return;
  1300. }
  1301. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1302. insert_html_element(token);
  1303. m_stack_of_open_elements.pop();
  1304. token.acknowledge_self_closing_flag_if_set();
  1305. return;
  1306. }
  1307. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1308. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1309. close_a_p_element();
  1310. insert_html_element(token);
  1311. m_stack_of_open_elements.pop();
  1312. token.acknowledge_self_closing_flag_if_set();
  1313. m_frameset_ok = false;
  1314. return;
  1315. }
  1316. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::image) {
  1317. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1318. PARSE_ERROR();
  1319. token.m_tag.tag_name.clear();
  1320. token.m_tag.tag_name.append(HTML::TagNames::img);
  1321. process_using_the_rules_for(m_insertion_mode, token);
  1322. return;
  1323. }
  1324. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1325. insert_html_element(token);
  1326. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1327. // If the next token is a U+000A LINE FEED (LF) character token,
  1328. // then ignore that token and move on to the next one.
  1329. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1330. auto next_token = m_tokenizer.next_token();
  1331. m_original_insertion_mode = m_insertion_mode;
  1332. m_frameset_ok = false;
  1333. m_insertion_mode = InsertionMode::Text;
  1334. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1335. // Ignore it.
  1336. } else {
  1337. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1338. }
  1339. return;
  1340. }
  1341. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1342. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1343. close_a_p_element();
  1344. }
  1345. reconstruct_the_active_formatting_elements();
  1346. m_frameset_ok = false;
  1347. parse_generic_raw_text_element(token);
  1348. return;
  1349. }
  1350. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1351. m_frameset_ok = false;
  1352. parse_generic_raw_text_element(token);
  1353. return;
  1354. }
  1355. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1356. parse_generic_raw_text_element(token);
  1357. return;
  1358. }
  1359. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1360. reconstruct_the_active_formatting_elements();
  1361. insert_html_element(token);
  1362. m_frameset_ok = false;
  1363. switch (m_insertion_mode) {
  1364. case InsertionMode::InTable:
  1365. case InsertionMode::InCaption:
  1366. case InsertionMode::InTableBody:
  1367. case InsertionMode::InRow:
  1368. case InsertionMode::InCell:
  1369. m_insertion_mode = InsertionMode::InSelectInTable;
  1370. break;
  1371. default:
  1372. m_insertion_mode = InsertionMode::InSelect;
  1373. break;
  1374. }
  1375. return;
  1376. }
  1377. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1378. if (current_node().local_name() == HTML::TagNames::option)
  1379. m_stack_of_open_elements.pop();
  1380. reconstruct_the_active_formatting_elements();
  1381. insert_html_element(token);
  1382. return;
  1383. }
  1384. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1385. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1386. generate_implied_end_tags();
  1387. if (current_node().local_name() != HTML::TagNames::ruby)
  1388. PARSE_ERROR();
  1389. insert_html_element(token);
  1390. return;
  1391. }
  1392. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1393. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1394. generate_implied_end_tags(HTML::TagNames::rtc);
  1395. if (current_node().local_name() != HTML::TagNames::rtc || current_node().local_name() != HTML::TagNames::ruby)
  1396. PARSE_ERROR();
  1397. insert_html_element(token);
  1398. return;
  1399. }
  1400. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::math) {
  1401. dbg() << "<math> element encountered.";
  1402. reconstruct_the_active_formatting_elements();
  1403. adjust_mathml_attributes(token);
  1404. adjust_foreign_attributes(token);
  1405. // FIXME: this should insert a foreign element, but lets just insert it normally for now :^)
  1406. insert_html_element(token);
  1407. if (token.is_self_closing()) {
  1408. m_stack_of_open_elements.pop();
  1409. token.acknowledge_self_closing_flag_if_set();
  1410. }
  1411. return;
  1412. }
  1413. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::svg) {
  1414. dbg() << "<svg> element encountered.";
  1415. reconstruct_the_active_formatting_elements();
  1416. adjust_svg_attributes(token);
  1417. adjust_foreign_attributes(token);
  1418. // FIXME: this should insert a foreign element, but lets just insert it normally for now :^)
  1419. insert_html_element(token);
  1420. if (token.is_self_closing()) {
  1421. m_stack_of_open_elements.pop();
  1422. token.acknowledge_self_closing_flag_if_set();
  1423. }
  1424. return;
  1425. }
  1426. 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))) {
  1427. PARSE_ERROR();
  1428. return;
  1429. }
  1430. // Any other start tag
  1431. if (token.is_start_tag()) {
  1432. reconstruct_the_active_formatting_elements();
  1433. insert_html_element(token);
  1434. return;
  1435. }
  1436. if (token.is_end_tag()) {
  1437. AnyOtherEndTag:
  1438. RefPtr<DOM::Element> node;
  1439. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1440. node = m_stack_of_open_elements.elements()[i];
  1441. if (node->local_name() == token.tag_name()) {
  1442. generate_implied_end_tags(token.tag_name());
  1443. if (node != current_node()) {
  1444. PARSE_ERROR();
  1445. }
  1446. while (&current_node() != node) {
  1447. m_stack_of_open_elements.pop();
  1448. }
  1449. m_stack_of_open_elements.pop();
  1450. break;
  1451. }
  1452. if (is_special_tag(node->local_name())) {
  1453. PARSE_ERROR();
  1454. return;
  1455. }
  1456. }
  1457. return;
  1458. }
  1459. }
  1460. void HTMLDocumentParser::adjust_mathml_attributes(HTMLToken& token)
  1461. {
  1462. token.adjust_attribute_name("definitionurl", "definitionURL");
  1463. }
  1464. void HTMLDocumentParser::adjust_svg_attributes(HTMLToken& token)
  1465. {
  1466. token.adjust_attribute_name("attributename", "attributeName");
  1467. token.adjust_attribute_name("attributetype", "attributeType");
  1468. token.adjust_attribute_name("basefrequency", "baseFrequency");
  1469. token.adjust_attribute_name("baseprofile", "baseProfile");
  1470. token.adjust_attribute_name("calcmode", "calcMode");
  1471. token.adjust_attribute_name("clippathunits", "clipPathUnits");
  1472. token.adjust_attribute_name("diffuseconstant", "diffuseConstant");
  1473. token.adjust_attribute_name("edgemode", "edgeMode");
  1474. token.adjust_attribute_name("filterunits", "filterUnits");
  1475. token.adjust_attribute_name("glyphref", "glyphRef");
  1476. token.adjust_attribute_name("gradienttransform", "gradientTransform");
  1477. token.adjust_attribute_name("gradientunits", "gradientUnits");
  1478. token.adjust_attribute_name("kernelmatrix", "kernelMatrix");
  1479. token.adjust_attribute_name("kernelunitlength", "kernelUnitLength");
  1480. token.adjust_attribute_name("keypoints", "keyPoints");
  1481. token.adjust_attribute_name("keysplines", "keySplines");
  1482. token.adjust_attribute_name("keytimes", "keyTimes");
  1483. token.adjust_attribute_name("lengthadjust", "lengthAdjust");
  1484. token.adjust_attribute_name("limitingconeangle", "limitingConeAngle");
  1485. token.adjust_attribute_name("markerheight", "markerHeight");
  1486. token.adjust_attribute_name("markerunits", "markerUnits");
  1487. token.adjust_attribute_name("markerwidth", "markerWidth");
  1488. token.adjust_attribute_name("maskcontentunits", "maskContentUnits");
  1489. token.adjust_attribute_name("maskunits", "maskUnits");
  1490. token.adjust_attribute_name("numoctaves", "numOctaves");
  1491. token.adjust_attribute_name("pathlength", "pathLength");
  1492. token.adjust_attribute_name("patterncontentunits", "patternContentUnits");
  1493. token.adjust_attribute_name("patterntransform", "patternTransform");
  1494. token.adjust_attribute_name("patternunits", "patternUnits");
  1495. token.adjust_attribute_name("pointsatx", "pointsAtX");
  1496. token.adjust_attribute_name("pointsaty", "pointsAtY");
  1497. token.adjust_attribute_name("pointsatz", "pointsAtZ");
  1498. token.adjust_attribute_name("preservealpha", "preserveAlpha");
  1499. token.adjust_attribute_name("preserveaspectratio", "preserveAspectRatio");
  1500. token.adjust_attribute_name("primitiveunits", "primitiveUnits");
  1501. token.adjust_attribute_name("refx", "refX");
  1502. token.adjust_attribute_name("refy", "refY");
  1503. token.adjust_attribute_name("repeatcount", "repeatCount");
  1504. token.adjust_attribute_name("repeatdur", "repeatDur");
  1505. token.adjust_attribute_name("requiredextensions", "requiredExtensions");
  1506. token.adjust_attribute_name("requiredfeatures", "requiredFeatures");
  1507. token.adjust_attribute_name("specularconstant", "specularConstant");
  1508. token.adjust_attribute_name("specularexponent", "specularExponent");
  1509. token.adjust_attribute_name("spreadmethod", "spreadMethod");
  1510. token.adjust_attribute_name("startoffset", "startOffset");
  1511. token.adjust_attribute_name("stddeviation", "stdDeviation");
  1512. token.adjust_attribute_name("stitchtiles", "stitchTiles");
  1513. token.adjust_attribute_name("surfacescale", "surfaceScale");
  1514. token.adjust_attribute_name("systemlanguage", "systemLanguage");
  1515. token.adjust_attribute_name("tablevalues", "tableValues");
  1516. token.adjust_attribute_name("targetx", "targetX");
  1517. token.adjust_attribute_name("targety", "targetY");
  1518. token.adjust_attribute_name("textlength", "textLength");
  1519. token.adjust_attribute_name("viewbox", "viewBox");
  1520. token.adjust_attribute_name("viewtarget", "viewTarget");
  1521. token.adjust_attribute_name("xchannelselector", "xChannelSelector");
  1522. token.adjust_attribute_name("ychannelselector", "yChannelSelector");
  1523. token.adjust_attribute_name("zoomandpan", "zoomAndPan");
  1524. }
  1525. void HTMLDocumentParser::adjust_foreign_attributes(HTMLToken& token)
  1526. {
  1527. auto xlink_namespace = "http://www.w3.org/1999/xlink";
  1528. token.adjust_foreign_attribute("xlink:actuate", "xlink", "actuate", xlink_namespace);
  1529. token.adjust_foreign_attribute("xlink:arcrole", "xlink", "arcrole", xlink_namespace);
  1530. token.adjust_foreign_attribute("xlink:href", "xlink", "href", xlink_namespace);
  1531. token.adjust_foreign_attribute("xlink:role", "xlink", "role", xlink_namespace);
  1532. token.adjust_foreign_attribute("xlink:show", "xlink", "show", xlink_namespace);
  1533. token.adjust_foreign_attribute("xlink:title", "xlink", "title", xlink_namespace);
  1534. token.adjust_foreign_attribute("xlink:type", "xlink", "type", xlink_namespace);
  1535. auto xml_namespace = "http://www.w3.org/XML/1998/namespace";
  1536. token.adjust_foreign_attribute("xml:lang", "xml", "lang", xml_namespace);
  1537. token.adjust_foreign_attribute("xml:space", "xml", "space", xml_namespace);
  1538. auto xmlns_namespace = "http://www.w3.org/2000/xmlns/";
  1539. token.adjust_foreign_attribute("xmlns", "", "xmlns", xmlns_namespace);
  1540. token.adjust_foreign_attribute("xmlns:xlink", "xmlns", "xlink", xmlns_namespace);
  1541. }
  1542. void HTMLDocumentParser::increment_script_nesting_level()
  1543. {
  1544. ++m_script_nesting_level;
  1545. }
  1546. void HTMLDocumentParser::decrement_script_nesting_level()
  1547. {
  1548. ASSERT(m_script_nesting_level);
  1549. --m_script_nesting_level;
  1550. }
  1551. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1552. {
  1553. if (token.is_character()) {
  1554. insert_character(token.code_point());
  1555. return;
  1556. }
  1557. if (token.is_end_of_file()) {
  1558. PARSE_ERROR();
  1559. if (current_node().local_name() == HTML::TagNames::script)
  1560. downcast<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1561. m_stack_of_open_elements.pop();
  1562. m_insertion_mode = m_original_insertion_mode;
  1563. process_using_the_rules_for(m_insertion_mode, token);
  1564. return;
  1565. }
  1566. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1567. // Make sure the <script> element has up-to-date text content before preparing the script.
  1568. flush_character_insertions();
  1569. NonnullRefPtr<HTMLScriptElement> script = downcast<HTMLScriptElement>(current_node());
  1570. m_stack_of_open_elements.pop();
  1571. m_insertion_mode = m_original_insertion_mode;
  1572. // FIXME: Handle tokenizer insertion point stuff here.
  1573. increment_script_nesting_level();
  1574. script->prepare_script({});
  1575. decrement_script_nesting_level();
  1576. if (script_nesting_level() == 0)
  1577. m_parser_pause_flag = false;
  1578. // FIXME: Handle tokenizer insertion point stuff here too.
  1579. while (document().pending_parsing_blocking_script()) {
  1580. if (script_nesting_level() != 0) {
  1581. m_parser_pause_flag = true;
  1582. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1583. // yielding control back to the caller. (Tokenization will resume when
  1584. // the caller returns to the "outer" tree construction stage.)
  1585. TODO();
  1586. } else {
  1587. auto the_script = document().take_pending_parsing_blocking_script({});
  1588. m_tokenizer.set_blocked(true);
  1589. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1590. // or the script's "ready to be parser-executed" flag is not set:
  1591. // spin the event loop until the parser's Document has no style sheet
  1592. // that is blocking scripts and the script's "ready to be parser-executed"
  1593. // flag is set.
  1594. if (the_script->failed_to_load())
  1595. return;
  1596. ASSERT(the_script->is_ready_to_be_parser_executed());
  1597. if (m_aborted)
  1598. return;
  1599. m_tokenizer.set_blocked(false);
  1600. // FIXME: Handle tokenizer insertion point stuff here too.
  1601. ASSERT(script_nesting_level() == 0);
  1602. increment_script_nesting_level();
  1603. the_script->execute_script();
  1604. decrement_script_nesting_level();
  1605. ASSERT(script_nesting_level() == 0);
  1606. m_parser_pause_flag = false;
  1607. // FIXME: Handle tokenizer insertion point stuff here too.
  1608. }
  1609. }
  1610. return;
  1611. }
  1612. if (token.is_end_tag()) {
  1613. m_stack_of_open_elements.pop();
  1614. m_insertion_mode = m_original_insertion_mode;
  1615. return;
  1616. }
  1617. TODO();
  1618. }
  1619. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1620. {
  1621. while (!current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1622. m_stack_of_open_elements.pop();
  1623. if (current_node().local_name() == HTML::TagNames::html)
  1624. ASSERT(m_parsing_fragment);
  1625. }
  1626. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1627. {
  1628. while (!current_node().local_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1629. m_stack_of_open_elements.pop();
  1630. if (current_node().local_name() == HTML::TagNames::html)
  1631. ASSERT(m_parsing_fragment);
  1632. }
  1633. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1634. {
  1635. while (!current_node().local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1636. m_stack_of_open_elements.pop();
  1637. if (current_node().local_name() == HTML::TagNames::html)
  1638. ASSERT(m_parsing_fragment);
  1639. }
  1640. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1641. {
  1642. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1643. clear_the_stack_back_to_a_table_row_context();
  1644. insert_html_element(token);
  1645. m_insertion_mode = InsertionMode::InCell;
  1646. m_list_of_active_formatting_elements.add_marker();
  1647. return;
  1648. }
  1649. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1650. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1651. PARSE_ERROR();
  1652. return;
  1653. }
  1654. clear_the_stack_back_to_a_table_row_context();
  1655. m_stack_of_open_elements.pop();
  1656. m_insertion_mode = InsertionMode::InTableBody;
  1657. return;
  1658. }
  1659. 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))
  1660. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1661. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1662. PARSE_ERROR();
  1663. return;
  1664. }
  1665. clear_the_stack_back_to_a_table_row_context();
  1666. m_stack_of_open_elements.pop();
  1667. m_insertion_mode = InsertionMode::InTableBody;
  1668. process_using_the_rules_for(m_insertion_mode, token);
  1669. return;
  1670. }
  1671. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1672. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1673. PARSE_ERROR();
  1674. return;
  1675. }
  1676. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1677. return;
  1678. }
  1679. clear_the_stack_back_to_a_table_row_context();
  1680. m_stack_of_open_elements.pop();
  1681. m_insertion_mode = InsertionMode::InTableBody;
  1682. process_using_the_rules_for(m_insertion_mode, token);
  1683. return;
  1684. }
  1685. 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)) {
  1686. PARSE_ERROR();
  1687. return;
  1688. }
  1689. process_using_the_rules_for(InsertionMode::InTable, token);
  1690. }
  1691. void HTMLDocumentParser::close_the_cell()
  1692. {
  1693. generate_implied_end_tags();
  1694. if (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1695. PARSE_ERROR();
  1696. }
  1697. while (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1698. m_stack_of_open_elements.pop();
  1699. m_stack_of_open_elements.pop();
  1700. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1701. m_insertion_mode = InsertionMode::InRow;
  1702. }
  1703. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1704. {
  1705. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1706. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1707. PARSE_ERROR();
  1708. return;
  1709. }
  1710. generate_implied_end_tags();
  1711. if (current_node().local_name() != token.tag_name()) {
  1712. PARSE_ERROR();
  1713. }
  1714. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1715. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1716. m_insertion_mode = InsertionMode::InRow;
  1717. return;
  1718. }
  1719. 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)) {
  1720. 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)) {
  1721. ASSERT(m_parsing_fragment);
  1722. PARSE_ERROR();
  1723. return;
  1724. }
  1725. close_the_cell();
  1726. process_using_the_rules_for(m_insertion_mode, token);
  1727. return;
  1728. }
  1729. 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)) {
  1730. PARSE_ERROR();
  1731. return;
  1732. }
  1733. 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)) {
  1734. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1735. PARSE_ERROR();
  1736. return;
  1737. }
  1738. close_the_cell();
  1739. // Reprocess the token.
  1740. process_using_the_rules_for(m_insertion_mode, token);
  1741. return;
  1742. }
  1743. process_using_the_rules_for(InsertionMode::InBody, token);
  1744. }
  1745. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1746. {
  1747. if (token.is_character()) {
  1748. if (token.code_point() == 0) {
  1749. PARSE_ERROR();
  1750. return;
  1751. }
  1752. m_pending_table_character_tokens.append(token);
  1753. return;
  1754. }
  1755. for (auto& pending_token : m_pending_table_character_tokens) {
  1756. ASSERT(pending_token.is_character());
  1757. if (!pending_token.is_parser_whitespace()) {
  1758. // If any of the tokens in the pending table character tokens list
  1759. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1760. // reprocess the character tokens in the pending table character tokens list using
  1761. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1762. PARSE_ERROR();
  1763. m_foster_parenting = true;
  1764. process_using_the_rules_for(InsertionMode::InBody, token);
  1765. m_foster_parenting = false;
  1766. return;
  1767. }
  1768. }
  1769. for (auto& pending_token : m_pending_table_character_tokens) {
  1770. insert_character(pending_token.code_point());
  1771. }
  1772. m_insertion_mode = m_original_insertion_mode;
  1773. process_using_the_rules_for(m_insertion_mode, token);
  1774. }
  1775. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1776. {
  1777. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1778. clear_the_stack_back_to_a_table_body_context();
  1779. insert_html_element(token);
  1780. m_insertion_mode = InsertionMode::InRow;
  1781. return;
  1782. }
  1783. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1784. PARSE_ERROR();
  1785. clear_the_stack_back_to_a_table_body_context();
  1786. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tr));
  1787. m_insertion_mode = InsertionMode::InRow;
  1788. process_using_the_rules_for(m_insertion_mode, token);
  1789. return;
  1790. }
  1791. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1792. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1793. PARSE_ERROR();
  1794. return;
  1795. }
  1796. clear_the_stack_back_to_a_table_body_context();
  1797. m_stack_of_open_elements.pop();
  1798. m_insertion_mode = InsertionMode::InTable;
  1799. return;
  1800. }
  1801. 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))
  1802. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1803. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tbody)
  1804. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::thead)
  1805. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tfoot)) {
  1806. PARSE_ERROR();
  1807. return;
  1808. }
  1809. clear_the_stack_back_to_a_table_body_context();
  1810. m_stack_of_open_elements.pop();
  1811. m_insertion_mode = InsertionMode::InTable;
  1812. process_using_the_rules_for(InsertionMode::InTable, token);
  1813. return;
  1814. }
  1815. 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)) {
  1816. PARSE_ERROR();
  1817. return;
  1818. }
  1819. process_using_the_rules_for(InsertionMode::InTable, token);
  1820. }
  1821. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1822. {
  1823. 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)) {
  1824. m_pending_table_character_tokens.clear();
  1825. m_original_insertion_mode = m_insertion_mode;
  1826. m_insertion_mode = InsertionMode::InTableText;
  1827. process_using_the_rules_for(InsertionMode::InTableText, token);
  1828. return;
  1829. }
  1830. if (token.is_comment()) {
  1831. insert_comment(token);
  1832. return;
  1833. }
  1834. if (token.is_doctype()) {
  1835. PARSE_ERROR();
  1836. return;
  1837. }
  1838. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1839. clear_the_stack_back_to_a_table_context();
  1840. m_list_of_active_formatting_elements.add_marker();
  1841. insert_html_element(token);
  1842. m_insertion_mode = InsertionMode::InCaption;
  1843. return;
  1844. }
  1845. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1846. clear_the_stack_back_to_a_table_context();
  1847. insert_html_element(token);
  1848. m_insertion_mode = InsertionMode::InColumnGroup;
  1849. return;
  1850. }
  1851. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1852. clear_the_stack_back_to_a_table_context();
  1853. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::colgroup));
  1854. m_insertion_mode = InsertionMode::InColumnGroup;
  1855. process_using_the_rules_for(m_insertion_mode, token);
  1856. return;
  1857. }
  1858. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1859. clear_the_stack_back_to_a_table_context();
  1860. insert_html_element(token);
  1861. m_insertion_mode = InsertionMode::InTableBody;
  1862. return;
  1863. }
  1864. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1865. clear_the_stack_back_to_a_table_context();
  1866. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tbody));
  1867. m_insertion_mode = InsertionMode::InTableBody;
  1868. process_using_the_rules_for(m_insertion_mode, token);
  1869. return;
  1870. }
  1871. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1872. PARSE_ERROR();
  1873. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table))
  1874. return;
  1875. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1876. reset_the_insertion_mode_appropriately();
  1877. process_using_the_rules_for(m_insertion_mode, token);
  1878. return;
  1879. }
  1880. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::table) {
  1881. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1882. PARSE_ERROR();
  1883. return;
  1884. }
  1885. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1886. reset_the_insertion_mode_appropriately();
  1887. return;
  1888. }
  1889. 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)) {
  1890. PARSE_ERROR();
  1891. return;
  1892. }
  1893. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::style, HTML::TagNames::script, HTML::TagNames::template_))
  1894. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_)) {
  1895. process_using_the_rules_for(InsertionMode::InHead, token);
  1896. return;
  1897. }
  1898. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1899. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1900. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1901. goto AnythingElse;
  1902. }
  1903. PARSE_ERROR();
  1904. insert_html_element(token);
  1905. // FIXME: Is this the correct interpretation of "Pop that input element off the stack of open elements."?
  1906. // Because this wording is the first time it's seen in the spec.
  1907. // Other times it's worded as: "Immediately pop the current node off the stack of open elements."
  1908. m_stack_of_open_elements.pop();
  1909. token.acknowledge_self_closing_flag_if_set();
  1910. return;
  1911. }
  1912. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1913. PARSE_ERROR();
  1914. if (m_form_element || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1915. return;
  1916. }
  1917. m_form_element = downcast<HTMLFormElement>(*insert_html_element(token));
  1918. // FIXME: See previous FIXME, as this is the same situation but for form.
  1919. m_stack_of_open_elements.pop();
  1920. return;
  1921. }
  1922. if (token.is_end_of_file()) {
  1923. process_using_the_rules_for(InsertionMode::InBody, token);
  1924. return;
  1925. }
  1926. AnythingElse:
  1927. PARSE_ERROR();
  1928. m_foster_parenting = true;
  1929. process_using_the_rules_for(InsertionMode::InBody, token);
  1930. m_foster_parenting = false;
  1931. }
  1932. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1933. {
  1934. 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)) {
  1935. PARSE_ERROR();
  1936. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1937. reset_the_insertion_mode_appropriately();
  1938. process_using_the_rules_for(m_insertion_mode, token);
  1939. return;
  1940. }
  1941. 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)) {
  1942. PARSE_ERROR();
  1943. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name()))
  1944. return;
  1945. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1946. reset_the_insertion_mode_appropriately();
  1947. process_using_the_rules_for(m_insertion_mode, token);
  1948. return;
  1949. }
  1950. process_using_the_rules_for(InsertionMode::InSelect, token);
  1951. }
  1952. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  1953. {
  1954. if (token.is_character()) {
  1955. if (token.code_point() == 0) {
  1956. PARSE_ERROR();
  1957. return;
  1958. }
  1959. insert_character(token.code_point());
  1960. return;
  1961. }
  1962. if (token.is_comment()) {
  1963. insert_comment(token);
  1964. return;
  1965. }
  1966. if (token.is_doctype()) {
  1967. PARSE_ERROR();
  1968. return;
  1969. }
  1970. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  1971. process_using_the_rules_for(InsertionMode::InBody, token);
  1972. return;
  1973. }
  1974. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  1975. if (current_node().local_name() == HTML::TagNames::option) {
  1976. m_stack_of_open_elements.pop();
  1977. }
  1978. insert_html_element(token);
  1979. return;
  1980. }
  1981. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1982. if (current_node().local_name() == HTML::TagNames::option) {
  1983. m_stack_of_open_elements.pop();
  1984. }
  1985. if (current_node().local_name() == HTML::TagNames::optgroup) {
  1986. m_stack_of_open_elements.pop();
  1987. }
  1988. insert_html_element(token);
  1989. return;
  1990. }
  1991. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1992. if (current_node().local_name() == HTML::TagNames::option && node_before_current_node().local_name() == HTML::TagNames::optgroup)
  1993. m_stack_of_open_elements.pop();
  1994. if (current_node().local_name() == HTML::TagNames::optgroup) {
  1995. m_stack_of_open_elements.pop();
  1996. } else {
  1997. PARSE_ERROR();
  1998. return;
  1999. }
  2000. return;
  2001. }
  2002. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  2003. if (current_node().local_name() == HTML::TagNames::option) {
  2004. m_stack_of_open_elements.pop();
  2005. } else {
  2006. PARSE_ERROR();
  2007. return;
  2008. }
  2009. return;
  2010. }
  2011. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  2012. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2013. ASSERT(m_parsing_fragment);
  2014. PARSE_ERROR();
  2015. return;
  2016. }
  2017. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2018. reset_the_insertion_mode_appropriately();
  2019. return;
  2020. }
  2021. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  2022. PARSE_ERROR();
  2023. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2024. ASSERT(m_parsing_fragment);
  2025. return;
  2026. }
  2027. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2028. reset_the_insertion_mode_appropriately();
  2029. return;
  2030. }
  2031. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  2032. PARSE_ERROR();
  2033. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2034. ASSERT(m_parsing_fragment);
  2035. return;
  2036. }
  2037. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2038. reset_the_insertion_mode_appropriately();
  2039. process_using_the_rules_for(m_insertion_mode, token);
  2040. return;
  2041. }
  2042. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  2043. process_using_the_rules_for(InsertionMode::InHead, token);
  2044. return;
  2045. }
  2046. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2047. process_using_the_rules_for(InsertionMode::InHead, token);
  2048. return;
  2049. }
  2050. if (token.is_end_of_file()) {
  2051. process_using_the_rules_for(InsertionMode::InBody, token);
  2052. return;
  2053. }
  2054. PARSE_ERROR();
  2055. }
  2056. void HTMLDocumentParser::handle_in_caption(HTMLToken& token)
  2057. {
  2058. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::caption) {
  2059. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2060. ASSERT(m_parsing_fragment);
  2061. PARSE_ERROR();
  2062. return;
  2063. }
  2064. generate_implied_end_tags();
  2065. if (current_node().local_name() != HTML::TagNames::caption)
  2066. PARSE_ERROR();
  2067. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2068. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2069. m_insertion_mode = InsertionMode::InTable;
  2070. return;
  2071. }
  2072. 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))
  2073. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  2074. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2075. ASSERT(m_parsing_fragment);
  2076. PARSE_ERROR();
  2077. return;
  2078. }
  2079. generate_implied_end_tags();
  2080. if (current_node().local_name() != HTML::TagNames::caption)
  2081. PARSE_ERROR();
  2082. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2083. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2084. m_insertion_mode = InsertionMode::InTable;
  2085. process_using_the_rules_for(m_insertion_mode, token);
  2086. return;
  2087. }
  2088. 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)) {
  2089. PARSE_ERROR();
  2090. return;
  2091. }
  2092. process_using_the_rules_for(InsertionMode::InBody, token);
  2093. }
  2094. void HTMLDocumentParser::handle_in_column_group(HTMLToken& token)
  2095. {
  2096. if (token.is_character() && token.is_parser_whitespace()) {
  2097. insert_character(token.code_point());
  2098. return;
  2099. }
  2100. if (token.is_comment()) {
  2101. insert_comment(token);
  2102. return;
  2103. }
  2104. if (token.is_doctype()) {
  2105. PARSE_ERROR();
  2106. return;
  2107. }
  2108. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2109. process_using_the_rules_for(InsertionMode::InBody, token);
  2110. return;
  2111. }
  2112. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2113. insert_html_element(token);
  2114. m_stack_of_open_elements.pop();
  2115. token.acknowledge_self_closing_flag_if_set();
  2116. return;
  2117. }
  2118. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  2119. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2120. PARSE_ERROR();
  2121. return;
  2122. }
  2123. m_stack_of_open_elements.pop();
  2124. m_insertion_mode = InsertionMode::InTable;
  2125. return;
  2126. }
  2127. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::col) {
  2128. PARSE_ERROR();
  2129. return;
  2130. }
  2131. if ((token.is_start_tag() || token.is_end_tag()) && token.tag_name() == HTML::TagNames::template_) {
  2132. process_using_the_rules_for(InsertionMode::InHead, token);
  2133. return;
  2134. }
  2135. if (token.is_end_of_file()) {
  2136. process_using_the_rules_for(InsertionMode::InBody, token);
  2137. return;
  2138. }
  2139. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2140. PARSE_ERROR();
  2141. return;
  2142. }
  2143. m_stack_of_open_elements.pop();
  2144. m_insertion_mode = InsertionMode::InTable;
  2145. process_using_the_rules_for(m_insertion_mode, token);
  2146. }
  2147. void HTMLDocumentParser::handle_in_template(HTMLToken& token)
  2148. {
  2149. if (token.is_character() || token.is_comment() || token.is_doctype()) {
  2150. process_using_the_rules_for(InsertionMode::InBody, token);
  2151. return;
  2152. }
  2153. 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)) {
  2154. process_using_the_rules_for(InsertionMode::InHead, token);
  2155. return;
  2156. }
  2157. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2158. process_using_the_rules_for(InsertionMode::InHead, token);
  2159. return;
  2160. }
  2161. 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)) {
  2162. m_stack_of_template_insertion_modes.take_last();
  2163. m_stack_of_template_insertion_modes.append(InsertionMode::InTable);
  2164. m_insertion_mode = InsertionMode::InTable;
  2165. process_using_the_rules_for(m_insertion_mode, token);
  2166. return;
  2167. }
  2168. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2169. m_stack_of_template_insertion_modes.take_last();
  2170. m_stack_of_template_insertion_modes.append(InsertionMode::InColumnGroup);
  2171. m_insertion_mode = InsertionMode::InColumnGroup;
  2172. process_using_the_rules_for(m_insertion_mode, token);
  2173. return;
  2174. }
  2175. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  2176. m_stack_of_template_insertion_modes.take_last();
  2177. m_stack_of_template_insertion_modes.append(InsertionMode::InTableBody);
  2178. m_insertion_mode = InsertionMode::InTableBody;
  2179. process_using_the_rules_for(m_insertion_mode, token);
  2180. return;
  2181. }
  2182. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2183. m_stack_of_template_insertion_modes.take_last();
  2184. m_stack_of_template_insertion_modes.append(InsertionMode::InRow);
  2185. m_insertion_mode = InsertionMode::InRow;
  2186. process_using_the_rules_for(m_insertion_mode, token);
  2187. return;
  2188. }
  2189. if (token.is_start_tag()) {
  2190. m_stack_of_template_insertion_modes.take_last();
  2191. m_stack_of_template_insertion_modes.append(InsertionMode::InBody);
  2192. m_insertion_mode = InsertionMode::InBody;
  2193. process_using_the_rules_for(m_insertion_mode, token);
  2194. return;
  2195. }
  2196. if (token.is_end_tag()) {
  2197. PARSE_ERROR();
  2198. return;
  2199. }
  2200. if (token.is_end_of_file()) {
  2201. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  2202. ASSERT(m_parsing_fragment);
  2203. stop_parsing();
  2204. return;
  2205. }
  2206. PARSE_ERROR();
  2207. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  2208. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2209. m_stack_of_template_insertion_modes.take_last();
  2210. reset_the_insertion_mode_appropriately();
  2211. process_using_the_rules_for(m_insertion_mode, token);
  2212. }
  2213. }
  2214. void HTMLDocumentParser::handle_in_frameset(HTMLToken& token)
  2215. {
  2216. if (token.is_character() && token.is_parser_whitespace()) {
  2217. insert_character(token.code_point());
  2218. return;
  2219. }
  2220. if (token.is_comment()) {
  2221. insert_comment(token);
  2222. return;
  2223. }
  2224. if (token.is_doctype()) {
  2225. PARSE_ERROR();
  2226. return;
  2227. }
  2228. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2229. process_using_the_rules_for(InsertionMode::InBody, token);
  2230. return;
  2231. }
  2232. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2233. insert_html_element(token);
  2234. return;
  2235. }
  2236. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2237. // FIXME: If the current node is the root html element, then this is a parse error; ignore the token. (fragment case)
  2238. m_stack_of_open_elements.pop();
  2239. if (!m_parsing_fragment && current_node().local_name() != HTML::TagNames::frameset) {
  2240. m_insertion_mode = InsertionMode::AfterFrameset;
  2241. }
  2242. return;
  2243. }
  2244. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frame) {
  2245. insert_html_element(token);
  2246. m_stack_of_open_elements.pop();
  2247. token.acknowledge_self_closing_flag_if_set();
  2248. return;
  2249. }
  2250. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2251. process_using_the_rules_for(InsertionMode::InHead, token);
  2252. return;
  2253. }
  2254. if (token.is_end_of_file()) {
  2255. //FIXME: If the current node is not the root html element, then this is a parse error.
  2256. stop_parsing();
  2257. return;
  2258. }
  2259. PARSE_ERROR();
  2260. }
  2261. void HTMLDocumentParser::handle_after_frameset(HTMLToken& token)
  2262. {
  2263. if (token.is_character() && token.is_parser_whitespace()) {
  2264. insert_character(token.code_point());
  2265. return;
  2266. }
  2267. if (token.is_comment()) {
  2268. insert_comment(token);
  2269. return;
  2270. }
  2271. if (token.is_doctype()) {
  2272. PARSE_ERROR();
  2273. return;
  2274. }
  2275. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2276. process_using_the_rules_for(InsertionMode::InBody, token);
  2277. return;
  2278. }
  2279. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  2280. m_insertion_mode = InsertionMode::AfterAfterFrameset;
  2281. return;
  2282. }
  2283. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2284. process_using_the_rules_for(InsertionMode::InHead, token);
  2285. return;
  2286. }
  2287. if (token.is_end_of_file()) {
  2288. stop_parsing();
  2289. return;
  2290. }
  2291. PARSE_ERROR();
  2292. }
  2293. void HTMLDocumentParser::handle_after_after_frameset(HTMLToken& token)
  2294. {
  2295. if (token.is_comment()) {
  2296. auto comment = adopt(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  2297. document().append_child(move(comment));
  2298. return;
  2299. }
  2300. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  2301. process_using_the_rules_for(InsertionMode::InBody, token);
  2302. return;
  2303. }
  2304. if (token.is_end_of_file()) {
  2305. stop_parsing();
  2306. return;
  2307. }
  2308. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2309. process_using_the_rules_for(InsertionMode::InHead, token);
  2310. return;
  2311. }
  2312. PARSE_ERROR();
  2313. }
  2314. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  2315. {
  2316. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2317. bool last = i == 0;
  2318. // NOTE: When parsing fragments, we substitute the context element for the root of the stack of open elements.
  2319. RefPtr<DOM::Element> node;
  2320. if (last && m_parsing_fragment) {
  2321. node = m_context_element;
  2322. } else {
  2323. node = m_stack_of_open_elements.elements().at(i);
  2324. }
  2325. if (node->local_name() == HTML::TagNames::select) {
  2326. TODO();
  2327. }
  2328. if (!last && node->local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2329. m_insertion_mode = InsertionMode::InCell;
  2330. return;
  2331. }
  2332. if (node->local_name() == HTML::TagNames::tr) {
  2333. m_insertion_mode = InsertionMode::InRow;
  2334. return;
  2335. }
  2336. if (node->local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  2337. m_insertion_mode = InsertionMode::InTableBody;
  2338. return;
  2339. }
  2340. if (node->local_name() == HTML::TagNames::caption) {
  2341. m_insertion_mode = InsertionMode::InCaption;
  2342. return;
  2343. }
  2344. if (node->local_name() == HTML::TagNames::colgroup) {
  2345. m_insertion_mode = InsertionMode::InColumnGroup;
  2346. return;
  2347. }
  2348. if (node->local_name() == HTML::TagNames::table) {
  2349. m_insertion_mode = InsertionMode::InTable;
  2350. return;
  2351. }
  2352. if (node->local_name() == HTML::TagNames::template_) {
  2353. m_insertion_mode = m_stack_of_template_insertion_modes.last();
  2354. return;
  2355. }
  2356. if (!last && node->local_name() == HTML::TagNames::head) {
  2357. m_insertion_mode = InsertionMode::InHead;
  2358. return;
  2359. }
  2360. if (node->local_name() == HTML::TagNames::body) {
  2361. m_insertion_mode = InsertionMode::InBody;
  2362. return;
  2363. }
  2364. if (node->local_name() == HTML::TagNames::frameset) {
  2365. ASSERT(m_parsing_fragment);
  2366. m_insertion_mode = InsertionMode::InFrameset;
  2367. return;
  2368. }
  2369. if (node->local_name() == HTML::TagNames::html) {
  2370. if (!m_head_element) {
  2371. ASSERT(m_parsing_fragment);
  2372. m_insertion_mode = InsertionMode::BeforeHead;
  2373. return;
  2374. }
  2375. m_insertion_mode = InsertionMode::AfterHead;
  2376. return;
  2377. }
  2378. }
  2379. ASSERT(m_parsing_fragment);
  2380. m_insertion_mode = InsertionMode::InBody;
  2381. }
  2382. const char* HTMLDocumentParser::insertion_mode_name() const
  2383. {
  2384. switch (m_insertion_mode) {
  2385. #define __ENUMERATE_INSERTION_MODE(mode) \
  2386. case InsertionMode::mode: \
  2387. return #mode;
  2388. ENUMERATE_INSERTION_MODES
  2389. #undef __ENUMERATE_INSERTION_MODE
  2390. }
  2391. ASSERT_NOT_REACHED();
  2392. }
  2393. DOM::Document& HTMLDocumentParser::document()
  2394. {
  2395. return *m_document;
  2396. }
  2397. NonnullRefPtrVector<DOM::Node> HTMLDocumentParser::parse_html_fragment(DOM::Element& context_element, const StringView& markup)
  2398. {
  2399. HTMLDocumentParser parser(markup, "utf-8");
  2400. parser.m_context_element = context_element;
  2401. parser.m_parsing_fragment = true;
  2402. parser.document().set_quirks_mode(context_element.document().mode());
  2403. if (context_element.local_name().is_one_of(HTML::TagNames::title, HTML::TagNames::textarea)) {
  2404. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  2405. } else if (context_element.local_name().is_one_of(HTML::TagNames::style, HTML::TagNames::xmp, HTML::TagNames::iframe, HTML::TagNames::noembed, HTML::TagNames::noframes)) {
  2406. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2407. } else if (context_element.local_name().is_one_of(HTML::TagNames::script)) {
  2408. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  2409. } else if (context_element.local_name().is_one_of(HTML::TagNames::noscript)) {
  2410. if (context_element.document().is_scripting_enabled())
  2411. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2412. } else if (context_element.local_name().is_one_of(HTML::TagNames::plaintext)) {
  2413. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  2414. }
  2415. auto root = create_element(context_element.document(), HTML::TagNames::html);
  2416. parser.document().append_child(root);
  2417. parser.m_stack_of_open_elements.push(root);
  2418. if (context_element.local_name() == HTML::TagNames::template_) {
  2419. parser.m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  2420. }
  2421. // FIXME: Create a start tag token whose name is the local name of context and whose attributes are the attributes of context.
  2422. parser.reset_the_insertion_mode_appropriately();
  2423. for (auto* form_candidate = &context_element; form_candidate; form_candidate = form_candidate->parent_element()) {
  2424. if (is<HTMLFormElement>(*form_candidate)) {
  2425. parser.m_form_element = downcast<HTMLFormElement>(*form_candidate);
  2426. break;
  2427. }
  2428. }
  2429. parser.run(context_element.document().url());
  2430. NonnullRefPtrVector<DOM::Node> children;
  2431. while (RefPtr<DOM::Node> child = root->first_child()) {
  2432. root->remove_child(*child);
  2433. context_element.document().adopt_node(*child);
  2434. children.append(*child);
  2435. }
  2436. return children;
  2437. }
  2438. }