HTMLDocumentParser.cpp 71 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658165916601661166216631664166516661667166816691670167116721673167416751676167716781679168016811682168316841685168616871688168916901691169216931694169516961697169816991700170117021703170417051706170717081709171017111712171317141715171617171718171917201721172217231724172517261727172817291730173117321733173417351736173717381739174017411742174317441745174617471748174917501751175217531754175517561757175817591760176117621763176417651766176717681769177017711772177317741775177617771778177917801781178217831784178517861787178817891790179117921793179417951796179717981799180018011802180318041805180618071808180918101811181218131814181518161817181818191820182118221823182418251826182718281829183018311832183318341835183618371838183918401841184218431844184518461847184818491850185118521853185418551856185718581859186018611862186318641865186618671868186918701871187218731874187518761877187818791880188118821883188418851886188718881889189018911892189318941895189618971898189919001901190219031904190519061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192919301931193219331934193519361937193819391940194119421943194419451946194719481949195019511952195319541955195619571958195919601961196219631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994
  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/HTMLFormElement.h>
  34. #include <LibWeb/DOM/HTMLHeadElement.h>
  35. #include <LibWeb/DOM/HTMLScriptElement.h>
  36. #include <LibWeb/DOM/Text.h>
  37. #include <LibWeb/Parser/HTMLDocumentParser.h>
  38. #include <LibWeb/Parser/HTMLToken.h>
  39. #define PARSE_ERROR() \
  40. do { \
  41. dbg() << "Parse error! " << __PRETTY_FUNCTION__ << " @ " << __LINE__; \
  42. } while (0)
  43. namespace Web {
  44. HTMLDocumentParser::HTMLDocumentParser(const StringView& input, const String& encoding)
  45. : m_tokenizer(input, encoding)
  46. {
  47. }
  48. HTMLDocumentParser::~HTMLDocumentParser()
  49. {
  50. }
  51. void HTMLDocumentParser::run(const URL& url)
  52. {
  53. m_document = adopt(*new Document);
  54. m_document->set_url(url);
  55. m_document->set_source(m_tokenizer.source());
  56. for (;;) {
  57. auto optional_token = m_tokenizer.next_token();
  58. if (!optional_token.has_value())
  59. break;
  60. auto& token = optional_token.value();
  61. #ifdef PARSER_DEBUG
  62. dbg() << "[" << insertion_mode_name() << "] " << token.to_string();
  63. #endif
  64. process_using_the_rules_for(m_insertion_mode, token);
  65. if (m_stop_parsing) {
  66. dbg() << "Stop parsing! :^)";
  67. break;
  68. }
  69. }
  70. flush_character_insertions();
  71. // "The end"
  72. auto scripts_to_execute_when_parsing_has_finished = m_document->take_scripts_to_execute_when_parsing_has_finished({});
  73. for (auto& script : scripts_to_execute_when_parsing_has_finished) {
  74. script.execute_script();
  75. }
  76. m_document->dispatch_event(Event::create("DOMContentLoaded"));
  77. auto scripts_to_execute_as_soon_as_possible = m_document->take_scripts_to_execute_as_soon_as_possible({});
  78. for (auto& script : scripts_to_execute_as_soon_as_possible) {
  79. script.execute_script();
  80. }
  81. }
  82. void HTMLDocumentParser::process_using_the_rules_for(InsertionMode mode, HTMLToken& token)
  83. {
  84. switch (mode) {
  85. case InsertionMode::Initial:
  86. handle_initial(token);
  87. break;
  88. case InsertionMode::BeforeHTML:
  89. handle_before_html(token);
  90. break;
  91. case InsertionMode::BeforeHead:
  92. handle_before_head(token);
  93. break;
  94. case InsertionMode::InHead:
  95. handle_in_head(token);
  96. break;
  97. case InsertionMode::InHeadNoscript:
  98. handle_in_head_noscript(token);
  99. break;
  100. case InsertionMode::AfterHead:
  101. handle_after_head(token);
  102. break;
  103. case InsertionMode::InBody:
  104. handle_in_body(token);
  105. break;
  106. case InsertionMode::AfterBody:
  107. handle_after_body(token);
  108. break;
  109. case InsertionMode::AfterAfterBody:
  110. handle_after_after_body(token);
  111. break;
  112. case InsertionMode::Text:
  113. handle_text(token);
  114. break;
  115. case InsertionMode::InTable:
  116. handle_in_table(token);
  117. break;
  118. case InsertionMode::InTableBody:
  119. handle_in_table_body(token);
  120. break;
  121. case InsertionMode::InRow:
  122. handle_in_row(token);
  123. break;
  124. case InsertionMode::InCell:
  125. handle_in_cell(token);
  126. break;
  127. case InsertionMode::InTableText:
  128. handle_in_table_text(token);
  129. break;
  130. case InsertionMode::InSelectInTable:
  131. handle_in_select_in_table(token);
  132. break;
  133. case InsertionMode::InSelect:
  134. handle_in_select(token);
  135. break;
  136. default:
  137. ASSERT_NOT_REACHED();
  138. }
  139. }
  140. void HTMLDocumentParser::handle_initial(HTMLToken& token)
  141. {
  142. if (token.is_character() && token.is_parser_whitespace()) {
  143. return;
  144. }
  145. if (token.is_comment()) {
  146. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  147. document().append_child(move(comment));
  148. return;
  149. }
  150. if (token.is_doctype()) {
  151. auto doctype = adopt(*new DocumentType(document()));
  152. doctype->set_name(token.m_doctype.name.to_string());
  153. document().append_child(move(doctype));
  154. document().set_quirks_mode(token.m_doctype.force_quirks);
  155. m_insertion_mode = InsertionMode::BeforeHTML;
  156. return;
  157. }
  158. PARSE_ERROR();
  159. document().set_quirks_mode(true);
  160. m_insertion_mode = InsertionMode::BeforeHTML;
  161. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  162. }
  163. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  164. {
  165. if (token.is_doctype()) {
  166. PARSE_ERROR();
  167. return;
  168. }
  169. if (token.is_comment()) {
  170. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  171. document().append_child(move(comment));
  172. return;
  173. }
  174. if (token.is_character() && token.is_parser_whitespace()) {
  175. return;
  176. }
  177. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  178. auto element = create_element_for(token);
  179. document().append_child(element);
  180. m_stack_of_open_elements.push(move(element));
  181. m_insertion_mode = InsertionMode::BeforeHead;
  182. return;
  183. }
  184. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  185. goto AnythingElse;
  186. }
  187. if (token.is_end_tag()) {
  188. PARSE_ERROR();
  189. return;
  190. }
  191. AnythingElse:
  192. auto element = create_element(document(), HTML::TagNames::html);
  193. document().append_child(element);
  194. m_stack_of_open_elements.push(element);
  195. // 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.
  196. m_insertion_mode = InsertionMode::BeforeHead;
  197. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  198. return;
  199. }
  200. Element& HTMLDocumentParser::current_node()
  201. {
  202. return m_stack_of_open_elements.current_node();
  203. }
  204. Element& HTMLDocumentParser::node_before_current_node()
  205. {
  206. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  207. }
  208. RefPtr<Node> HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  209. {
  210. auto& target = current_node();
  211. if (m_foster_parenting) {
  212. TODO();
  213. }
  214. return target;
  215. }
  216. NonnullRefPtr<Element> HTMLDocumentParser::create_element_for(HTMLToken& token)
  217. {
  218. auto element = create_element(document(), token.tag_name());
  219. for (auto& attribute : token.m_tag.attributes) {
  220. element->set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  221. }
  222. return element;
  223. }
  224. RefPtr<Element> HTMLDocumentParser::insert_html_element(HTMLToken& token)
  225. {
  226. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  227. auto element = create_element_for(token);
  228. // FIXME: Check if it's possible to insert `element` at `adjusted_insertion_location`
  229. adjusted_insertion_location->append_child(element);
  230. m_stack_of_open_elements.push(element);
  231. return element;
  232. }
  233. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  234. {
  235. if (token.is_character() && token.is_parser_whitespace()) {
  236. return;
  237. }
  238. if (token.is_comment()) {
  239. insert_comment(token);
  240. return;
  241. }
  242. if (token.is_doctype()) {
  243. PARSE_ERROR();
  244. return;
  245. }
  246. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  247. process_using_the_rules_for(InsertionMode::InBody, token);
  248. return;
  249. }
  250. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  251. auto element = insert_html_element(token);
  252. m_head_element = to<HTMLHeadElement>(element);
  253. m_insertion_mode = InsertionMode::InHead;
  254. return;
  255. }
  256. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  257. goto AnythingElse;
  258. }
  259. if (token.is_end_tag()) {
  260. PARSE_ERROR();
  261. return;
  262. }
  263. AnythingElse:
  264. HTMLToken fake_head_token;
  265. fake_head_token.m_type = HTMLToken::Type::StartTag;
  266. fake_head_token.m_tag.tag_name.append(HTML::TagNames::head);
  267. m_head_element = to<HTMLHeadElement>(insert_html_element(fake_head_token));
  268. m_insertion_mode = InsertionMode::InHead;
  269. process_using_the_rules_for(InsertionMode::InHead, token);
  270. return;
  271. }
  272. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  273. {
  274. auto data = token.m_comment_or_character.data.to_string();
  275. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  276. adjusted_insertion_location->append_child(adopt(*new Comment(document(), data)));
  277. }
  278. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  279. {
  280. if (token.is_parser_whitespace()) {
  281. insert_character(token.codepoint());
  282. return;
  283. }
  284. if (token.is_comment()) {
  285. insert_comment(token);
  286. return;
  287. }
  288. if (token.is_doctype()) {
  289. PARSE_ERROR();
  290. return;
  291. }
  292. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  293. process_using_the_rules_for(InsertionMode::InBody, token);
  294. return;
  295. }
  296. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  297. insert_html_element(token);
  298. m_stack_of_open_elements.pop();
  299. token.acknowledge_self_closing_flag_if_set();
  300. return;
  301. }
  302. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  303. auto element = insert_html_element(token);
  304. m_stack_of_open_elements.pop();
  305. token.acknowledge_self_closing_flag_if_set();
  306. return;
  307. }
  308. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  309. insert_html_element(token);
  310. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  311. m_original_insertion_mode = m_insertion_mode;
  312. m_insertion_mode = InsertionMode::Text;
  313. return;
  314. }
  315. 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)) {
  316. parse_generic_raw_text_element(token);
  317. return;
  318. }
  319. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  320. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  321. auto element = create_element_for(token);
  322. auto& script_element = to<HTMLScriptElement>(*element);
  323. script_element.set_parser_document({}, document());
  324. script_element.set_non_blocking({}, false);
  325. if (m_parsing_fragment) {
  326. TODO();
  327. }
  328. if (m_invoked_via_document_write) {
  329. TODO();
  330. }
  331. adjusted_insertion_location->append_child(element, false);
  332. m_stack_of_open_elements.push(element);
  333. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  334. m_original_insertion_mode = m_insertion_mode;
  335. m_insertion_mode = InsertionMode::Text;
  336. return;
  337. }
  338. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  339. m_stack_of_open_elements.pop();
  340. m_insertion_mode = InsertionMode::AfterHead;
  341. return;
  342. }
  343. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  344. TODO();
  345. }
  346. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  347. // FIXME: Support this properly
  348. insert_html_element(token);
  349. return;
  350. }
  351. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  352. // FIXME: Support this properly
  353. ASSERT(current_node().tag_name() == HTML::TagNames::template_);
  354. m_stack_of_open_elements.pop();
  355. return;
  356. }
  357. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  358. PARSE_ERROR();
  359. return;
  360. }
  361. m_stack_of_open_elements.pop();
  362. m_insertion_mode = InsertionMode::AfterHead;
  363. process_using_the_rules_for(m_insertion_mode, token);
  364. }
  365. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken&)
  366. {
  367. TODO();
  368. }
  369. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  370. {
  371. insert_html_element(token);
  372. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  373. m_original_insertion_mode = m_insertion_mode;
  374. m_insertion_mode = InsertionMode::Text;
  375. }
  376. Text* HTMLDocumentParser::find_character_insertion_node()
  377. {
  378. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  379. if (adjusted_insertion_location->is_document())
  380. return nullptr;
  381. if (adjusted_insertion_location->last_child() && adjusted_insertion_location->last_child()->is_text())
  382. return to<Text>(adjusted_insertion_location->last_child());
  383. auto new_text_node = adopt(*new Text(document(), ""));
  384. adjusted_insertion_location->append_child(new_text_node);
  385. return new_text_node;
  386. }
  387. void HTMLDocumentParser::flush_character_insertions()
  388. {
  389. if (m_character_insertion_builder.is_empty())
  390. return;
  391. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  392. m_character_insertion_node->parent()->children_changed();
  393. m_character_insertion_builder.clear();
  394. }
  395. void HTMLDocumentParser::insert_character(u32 data)
  396. {
  397. auto node = find_character_insertion_node();
  398. if (node == m_character_insertion_node) {
  399. m_character_insertion_builder.append(Utf32View { &data, 1 });
  400. return;
  401. }
  402. if (!m_character_insertion_node) {
  403. m_character_insertion_node = node;
  404. m_character_insertion_builder.append(Utf32View { &data, 1 });
  405. return;
  406. }
  407. flush_character_insertions();
  408. m_character_insertion_node = node;
  409. m_character_insertion_builder.append(Utf32View { &data, 1 });
  410. }
  411. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  412. {
  413. if (token.is_character() && token.is_parser_whitespace()) {
  414. insert_character(token.codepoint());
  415. return;
  416. }
  417. if (token.is_comment()) {
  418. insert_comment(token);
  419. return;
  420. }
  421. if (token.is_doctype()) {
  422. PARSE_ERROR();
  423. return;
  424. }
  425. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  426. process_using_the_rules_for(InsertionMode::InBody, token);
  427. return;
  428. }
  429. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  430. insert_html_element(token);
  431. m_frameset_ok = false;
  432. m_insertion_mode = InsertionMode::InBody;
  433. return;
  434. }
  435. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  436. insert_html_element(token);
  437. m_insertion_mode = InsertionMode::InFrameset;
  438. return;
  439. }
  440. 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)) {
  441. PARSE_ERROR();
  442. m_stack_of_open_elements.push(*m_head_element);
  443. process_using_the_rules_for(InsertionMode::InHead, token);
  444. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  445. return entry.ptr() == m_head_element;
  446. });
  447. return;
  448. }
  449. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  450. TODO();
  451. }
  452. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  453. goto AnythingElse;
  454. }
  455. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  456. PARSE_ERROR();
  457. return;
  458. }
  459. AnythingElse:
  460. HTMLToken fake_body_token;
  461. fake_body_token.m_type = HTMLToken::Type::StartTag;
  462. fake_body_token.m_tag.tag_name.append(HTML::TagNames::body);
  463. insert_html_element(fake_body_token);
  464. m_insertion_mode = InsertionMode::InBody;
  465. process_using_the_rules_for(m_insertion_mode, token);
  466. }
  467. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  468. {
  469. while (current_node().tag_name() != exception && current_node().tag_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))
  470. m_stack_of_open_elements.pop();
  471. }
  472. void HTMLDocumentParser::close_a_p_element()
  473. {
  474. generate_implied_end_tags(HTML::TagNames::p);
  475. if (current_node().tag_name() != HTML::TagNames::p) {
  476. PARSE_ERROR();
  477. }
  478. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  479. }
  480. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  481. {
  482. if (token.is_character() && token.is_parser_whitespace()) {
  483. process_using_the_rules_for(InsertionMode::InBody, token);
  484. return;
  485. }
  486. if (token.is_comment()) {
  487. TODO();
  488. }
  489. if (token.is_doctype()) {
  490. PARSE_ERROR();
  491. return;
  492. }
  493. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  494. process_using_the_rules_for(InsertionMode::InBody, token);
  495. return;
  496. }
  497. if (token.is_end_of_file()) {
  498. stop_parsing();
  499. return;
  500. }
  501. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  502. if (m_parsing_fragment) {
  503. TODO();
  504. }
  505. m_insertion_mode = InsertionMode::AfterAfterBody;
  506. return;
  507. }
  508. PARSE_ERROR();
  509. m_insertion_mode = InsertionMode::InBody;
  510. process_using_the_rules_for(InsertionMode::InBody, token);
  511. }
  512. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  513. {
  514. if (token.is_comment()) {
  515. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  516. document().append_child(move(comment));
  517. return;
  518. }
  519. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  520. process_using_the_rules_for(InsertionMode::InBody, token);
  521. return;
  522. }
  523. if (token.is_end_of_file()) {
  524. stop_parsing();
  525. return;
  526. }
  527. PARSE_ERROR();
  528. m_insertion_mode = InsertionMode::InBody;
  529. process_using_the_rules_for(m_insertion_mode, token);
  530. }
  531. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  532. {
  533. // FIXME: This needs to care about "markers"
  534. if (m_list_of_active_formatting_elements.is_empty())
  535. return;
  536. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  537. return;
  538. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  539. return;
  540. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  541. RefPtr<Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  542. ASSERT(entry);
  543. Rewind:
  544. if (index == 0) {
  545. goto Create;
  546. }
  547. --index;
  548. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  549. ASSERT(entry);
  550. if (!m_stack_of_open_elements.contains(*entry))
  551. goto Rewind;
  552. Advance:
  553. ++index;
  554. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  555. ASSERT(entry);
  556. Create:
  557. // FIXME: Hold on to the real token!
  558. HTMLToken fake_token;
  559. fake_token.m_type = HTMLToken::Type::StartTag;
  560. fake_token.m_tag.tag_name.append(entry->tag_name());
  561. auto new_element = insert_html_element(fake_token);
  562. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  563. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  564. goto Advance;
  565. }
  566. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  567. {
  568. auto subject = token.tag_name();
  569. // If the current node is an HTML element whose tag name is subject,
  570. // and the current node is not in the list of active formatting elements,
  571. // then pop the current node off the stack of open elements, and return.
  572. if (current_node().tag_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  573. m_stack_of_open_elements.pop();
  574. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  575. }
  576. size_t outer_loop_counter = 0;
  577. //OuterLoop:
  578. if (outer_loop_counter >= 8)
  579. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  580. ++outer_loop_counter;
  581. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  582. if (!formatting_element)
  583. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  584. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  585. PARSE_ERROR();
  586. // FIXME: If formatting element is not in the stack of open elements,
  587. // then this is a parse error; remove the element from the list, and return.
  588. TODO();
  589. }
  590. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  591. PARSE_ERROR();
  592. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  593. }
  594. if (formatting_element != &current_node()) {
  595. PARSE_ERROR();
  596. }
  597. RefPtr<Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  598. if (!furthest_block) {
  599. while (&current_node() != formatting_element)
  600. m_stack_of_open_elements.pop();
  601. m_stack_of_open_elements.pop();
  602. m_list_of_active_formatting_elements.remove(*formatting_element);
  603. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  604. }
  605. // FIXME: Implement the rest of the AAA :^)
  606. TODO();
  607. }
  608. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name)
  609. {
  610. return tag_name.is_one_of(
  611. HTML::TagNames::address,
  612. HTML::TagNames::applet,
  613. HTML::TagNames::area,
  614. HTML::TagNames::article,
  615. HTML::TagNames::aside,
  616. HTML::TagNames::base,
  617. HTML::TagNames::basefont,
  618. HTML::TagNames::bgsound,
  619. HTML::TagNames::blockquote,
  620. HTML::TagNames::body,
  621. HTML::TagNames::br,
  622. HTML::TagNames::button,
  623. HTML::TagNames::caption,
  624. HTML::TagNames::center,
  625. HTML::TagNames::col,
  626. HTML::TagNames::colgroup,
  627. HTML::TagNames::dd,
  628. HTML::TagNames::details,
  629. HTML::TagNames::dir,
  630. HTML::TagNames::div,
  631. HTML::TagNames::dl,
  632. HTML::TagNames::dt,
  633. HTML::TagNames::embed,
  634. HTML::TagNames::fieldset,
  635. HTML::TagNames::figcaption,
  636. HTML::TagNames::figure,
  637. HTML::TagNames::footer,
  638. HTML::TagNames::form,
  639. HTML::TagNames::frame,
  640. HTML::TagNames::frameset,
  641. HTML::TagNames::h1,
  642. HTML::TagNames::h2,
  643. HTML::TagNames::h3,
  644. HTML::TagNames::h4,
  645. HTML::TagNames::h5,
  646. HTML::TagNames::h6,
  647. HTML::TagNames::head,
  648. HTML::TagNames::header,
  649. HTML::TagNames::hgroup,
  650. HTML::TagNames::hr,
  651. HTML::TagNames::html,
  652. HTML::TagNames::iframe,
  653. HTML::TagNames::img,
  654. HTML::TagNames::input,
  655. HTML::TagNames::keygen,
  656. HTML::TagNames::li,
  657. HTML::TagNames::link,
  658. HTML::TagNames::listing,
  659. HTML::TagNames::main,
  660. HTML::TagNames::marquee,
  661. HTML::TagNames::menu,
  662. HTML::TagNames::meta,
  663. HTML::TagNames::nav,
  664. HTML::TagNames::noembed,
  665. HTML::TagNames::noframes,
  666. HTML::TagNames::noscript,
  667. HTML::TagNames::object,
  668. HTML::TagNames::ol,
  669. HTML::TagNames::p,
  670. HTML::TagNames::param,
  671. HTML::TagNames::plaintext,
  672. HTML::TagNames::pre,
  673. HTML::TagNames::script,
  674. HTML::TagNames::section,
  675. HTML::TagNames::select,
  676. HTML::TagNames::source,
  677. HTML::TagNames::style,
  678. HTML::TagNames::summary,
  679. HTML::TagNames::table,
  680. HTML::TagNames::tbody,
  681. HTML::TagNames::td,
  682. HTML::TagNames::template_,
  683. HTML::TagNames::textarea,
  684. HTML::TagNames::tfoot,
  685. HTML::TagNames::th,
  686. HTML::TagNames::thead,
  687. HTML::TagNames::title,
  688. HTML::TagNames::tr,
  689. HTML::TagNames::track,
  690. HTML::TagNames::ul,
  691. HTML::TagNames::wbr,
  692. HTML::TagNames::xmp);
  693. }
  694. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  695. {
  696. if (token.is_character()) {
  697. if (token.codepoint() == 0) {
  698. PARSE_ERROR();
  699. return;
  700. }
  701. if (token.is_parser_whitespace()) {
  702. reconstruct_the_active_formatting_elements();
  703. insert_character(token.codepoint());
  704. return;
  705. }
  706. reconstruct_the_active_formatting_elements();
  707. insert_character(token.codepoint());
  708. m_frameset_ok = false;
  709. return;
  710. }
  711. if (token.is_comment()) {
  712. insert_comment(token);
  713. return;
  714. }
  715. if (token.is_doctype()) {
  716. PARSE_ERROR();
  717. return;
  718. }
  719. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  720. PARSE_ERROR();
  721. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  722. return;
  723. for (auto& attribute : token.m_tag.attributes) {
  724. if (current_node().has_attribute(attribute.name_builder.string_view()))
  725. continue;
  726. current_node().set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  727. }
  728. return;
  729. }
  730. 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)) {
  731. process_using_the_rules_for(InsertionMode::InHead, token);
  732. return;
  733. }
  734. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  735. process_using_the_rules_for(InsertionMode::InHead, token);
  736. return;
  737. }
  738. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  739. PARSE_ERROR();
  740. if (m_stack_of_open_elements.elements().size() == 1
  741. || node_before_current_node().tag_name() != HTML::TagNames::body
  742. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  743. return;
  744. }
  745. m_frameset_ok = false;
  746. for (auto& attribute : token.m_tag.attributes) {
  747. if (node_before_current_node().has_attribute(attribute.name_builder.string_view()))
  748. continue;
  749. node_before_current_node().set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  750. }
  751. return;
  752. }
  753. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  754. TODO();
  755. }
  756. if (token.is_end_of_file()) {
  757. // FIXME: If the stack of template insertion modes is not empty,
  758. // then process the token using the rules for the "in template" insertion mode.
  759. // FIXME: If there is a node in the stack of open elements that is not either
  760. // a dd element, a dt element, an li element, an optgroup element, an option element,
  761. // a p element, an rb element, an rp element, an rt element, an rtc element,
  762. // a tbody element, a td element, a tfoot element, a th element, a thead element,
  763. // a tr element, the body element, or the html element, then this is a parse error.
  764. stop_parsing();
  765. return;
  766. }
  767. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  768. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  769. PARSE_ERROR();
  770. return;
  771. }
  772. for (auto& node : m_stack_of_open_elements.elements()) {
  773. if (!node.tag_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)) {
  774. PARSE_ERROR();
  775. break;
  776. }
  777. }
  778. m_insertion_mode = InsertionMode::AfterBody;
  779. return;
  780. }
  781. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  782. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  783. PARSE_ERROR();
  784. return;
  785. }
  786. for (auto& node : m_stack_of_open_elements.elements()) {
  787. if (!node.tag_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)) {
  788. PARSE_ERROR();
  789. break;
  790. }
  791. }
  792. m_insertion_mode = InsertionMode::AfterBody;
  793. process_using_the_rules_for(m_insertion_mode, token);
  794. return;
  795. }
  796. 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)) {
  797. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  798. close_a_p_element();
  799. insert_html_element(token);
  800. return;
  801. }
  802. 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)) {
  803. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  804. close_a_p_element();
  805. if (current_node().tag_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  806. PARSE_ERROR();
  807. m_stack_of_open_elements.pop();
  808. }
  809. insert_html_element(token);
  810. return;
  811. }
  812. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  813. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  814. close_a_p_element();
  815. insert_html_element(token);
  816. m_frameset_ok = false;
  817. // If the next token is a U+000A LINE FEED (LF) character token,
  818. // then ignore that token and move on to the next one.
  819. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  820. auto next_token = m_tokenizer.next_token();
  821. if (next_token.has_value() && next_token.value().is_character() && next_token.value().codepoint() == '\n') {
  822. // Ignore it.
  823. } else {
  824. process_using_the_rules_for(m_insertion_mode, next_token.value());
  825. }
  826. return;
  827. }
  828. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  829. if (m_form_element && m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  830. PARSE_ERROR();
  831. return;
  832. }
  833. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  834. close_a_p_element();
  835. auto element = insert_html_element(token);
  836. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  837. m_form_element = to<HTMLFormElement>(*element);
  838. return;
  839. }
  840. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  841. m_frameset_ok = false;
  842. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  843. RefPtr<Element> node = m_stack_of_open_elements.elements()[i];
  844. if (node->tag_name() == HTML::TagNames::li) {
  845. generate_implied_end_tags(HTML::TagNames::li);
  846. if (current_node().tag_name() != HTML::TagNames::li) {
  847. PARSE_ERROR();
  848. }
  849. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  850. break;
  851. }
  852. if (is_special_tag(node->tag_name()) && !node->tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  853. break;
  854. }
  855. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  856. close_a_p_element();
  857. insert_html_element(token);
  858. return;
  859. }
  860. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  861. m_frameset_ok = false;
  862. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  863. RefPtr<Element> node = m_stack_of_open_elements.elements()[i];
  864. if (node->tag_name() == HTML::TagNames::dd) {
  865. generate_implied_end_tags(HTML::TagNames::dd);
  866. if (current_node().tag_name() != HTML::TagNames::dd) {
  867. PARSE_ERROR();
  868. }
  869. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  870. break;
  871. }
  872. if (node->tag_name() == HTML::TagNames::dt) {
  873. generate_implied_end_tags(HTML::TagNames::dt);
  874. if (current_node().tag_name() != HTML::TagNames::dt) {
  875. PARSE_ERROR();
  876. }
  877. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  878. break;
  879. }
  880. if (is_special_tag(node->tag_name()) && !node->tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  881. break;
  882. }
  883. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  884. close_a_p_element();
  885. insert_html_element(token);
  886. return;
  887. }
  888. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  889. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  890. close_a_p_element();
  891. insert_html_element(token);
  892. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  893. return;
  894. }
  895. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  896. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  897. PARSE_ERROR();
  898. generate_implied_end_tags();
  899. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  900. }
  901. reconstruct_the_active_formatting_elements();
  902. insert_html_element(token);
  903. m_frameset_ok = false;
  904. return;
  905. }
  906. 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)) {
  907. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  908. PARSE_ERROR();
  909. return;
  910. }
  911. generate_implied_end_tags();
  912. if (current_node().tag_name() != token.tag_name()) {
  913. PARSE_ERROR();
  914. }
  915. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  916. return;
  917. }
  918. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  919. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  920. auto node = m_form_element;
  921. m_form_element = nullptr;
  922. if (!node || m_stack_of_open_elements.has_in_scope(*node)) {
  923. PARSE_ERROR();
  924. return;
  925. }
  926. generate_implied_end_tags();
  927. if (&current_node() != node) {
  928. PARSE_ERROR();
  929. }
  930. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  931. } else {
  932. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  933. PARSE_ERROR();
  934. return;
  935. }
  936. generate_implied_end_tags();
  937. if (current_node().tag_name() != HTML::TagNames::form) {
  938. PARSE_ERROR();
  939. }
  940. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  941. }
  942. return;
  943. }
  944. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  945. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  946. PARSE_ERROR();
  947. HTMLToken fake_p_token;
  948. fake_p_token.m_type = HTMLToken::Type::StartTag;
  949. fake_p_token.m_tag.tag_name.append(HTML::TagNames::p);
  950. insert_html_element(fake_p_token);
  951. }
  952. close_a_p_element();
  953. return;
  954. }
  955. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  956. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  957. PARSE_ERROR();
  958. return;
  959. }
  960. generate_implied_end_tags(HTML::TagNames::li);
  961. if (current_node().tag_name() != HTML::TagNames::li) {
  962. PARSE_ERROR();
  963. dbg() << "Expected <li> current node, but had <" << current_node().tag_name() << ">";
  964. }
  965. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  966. return;
  967. }
  968. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  969. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  970. PARSE_ERROR();
  971. return;
  972. }
  973. generate_implied_end_tags(token.tag_name());
  974. if (current_node().tag_name() != token.tag_name()) {
  975. PARSE_ERROR();
  976. }
  977. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  978. return;
  979. }
  980. 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)) {
  981. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  982. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  983. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  984. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  985. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  986. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  987. PARSE_ERROR();
  988. return;
  989. }
  990. generate_implied_end_tags();
  991. if (current_node().tag_name() != token.tag_name()) {
  992. PARSE_ERROR();
  993. }
  994. for (;;) {
  995. auto popped_element = m_stack_of_open_elements.pop();
  996. if (popped_element->tag_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6))
  997. break;
  998. }
  999. return;
  1000. }
  1001. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1002. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1003. PARSE_ERROR();
  1004. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1005. goto AnyOtherEndTag;
  1006. m_list_of_active_formatting_elements.remove(*element);
  1007. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1008. return entry.ptr() == element;
  1009. });
  1010. }
  1011. reconstruct_the_active_formatting_elements();
  1012. auto element = insert_html_element(token);
  1013. m_list_of_active_formatting_elements.add(*element);
  1014. return;
  1015. }
  1016. 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)) {
  1017. reconstruct_the_active_formatting_elements();
  1018. auto element = insert_html_element(token);
  1019. m_list_of_active_formatting_elements.add(*element);
  1020. return;
  1021. }
  1022. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1023. reconstruct_the_active_formatting_elements();
  1024. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1025. PARSE_ERROR();
  1026. run_the_adoption_agency_algorithm(token);
  1027. reconstruct_the_active_formatting_elements();
  1028. }
  1029. auto element = insert_html_element(token);
  1030. m_list_of_active_formatting_elements.add(*element);
  1031. return;
  1032. }
  1033. 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)) {
  1034. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1035. goto AnyOtherEndTag;
  1036. return;
  1037. }
  1038. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1039. reconstruct_the_active_formatting_elements();
  1040. insert_html_element(token);
  1041. m_list_of_active_formatting_elements.add_marker();
  1042. m_frameset_ok = false;
  1043. return;
  1044. }
  1045. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1046. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1047. PARSE_ERROR();
  1048. return;
  1049. }
  1050. generate_implied_end_tags();
  1051. if (current_node().tag_name() != token.tag_name()) {
  1052. PARSE_ERROR();
  1053. }
  1054. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1055. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1056. return;
  1057. }
  1058. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1059. if (!document().in_quirks_mode()) {
  1060. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1061. close_a_p_element();
  1062. }
  1063. insert_html_element(token);
  1064. m_frameset_ok = false;
  1065. m_insertion_mode = InsertionMode::InTable;
  1066. return;
  1067. }
  1068. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1069. token.drop_attributes();
  1070. goto BRStartTag;
  1071. }
  1072. 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)) {
  1073. BRStartTag:
  1074. reconstruct_the_active_formatting_elements();
  1075. insert_html_element(token);
  1076. m_stack_of_open_elements.pop();
  1077. token.acknowledge_self_closing_flag_if_set();
  1078. m_frameset_ok = false;
  1079. return;
  1080. }
  1081. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1082. reconstruct_the_active_formatting_elements();
  1083. insert_html_element(token);
  1084. m_stack_of_open_elements.pop();
  1085. token.acknowledge_self_closing_flag_if_set();
  1086. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1087. if (type_attribute.is_null() || type_attribute != "hidden") {
  1088. m_frameset_ok = false;
  1089. }
  1090. return;
  1091. }
  1092. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1093. insert_html_element(token);
  1094. m_stack_of_open_elements.pop();
  1095. token.acknowledge_self_closing_flag_if_set();
  1096. return;
  1097. }
  1098. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1099. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1100. close_a_p_element();
  1101. insert_html_element(token);
  1102. m_stack_of_open_elements.pop();
  1103. token.acknowledge_self_closing_flag_if_set();
  1104. m_frameset_ok = false;
  1105. return;
  1106. }
  1107. if (token.is_start_tag() && token.tag_name().equals_ignoring_case("image")) {
  1108. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1109. PARSE_ERROR();
  1110. token.m_tag.tag_name.clear();
  1111. token.m_tag.tag_name.append(HTML::TagNames::img);
  1112. process_using_the_rules_for(m_insertion_mode, token);
  1113. return;
  1114. }
  1115. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1116. insert_html_element(token);
  1117. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1118. // If the next token is a U+000A LINE FEED (LF) character token,
  1119. // then ignore that token and move on to the next one.
  1120. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1121. auto next_token = m_tokenizer.next_token();
  1122. m_original_insertion_mode = m_insertion_mode;
  1123. m_frameset_ok = false;
  1124. m_insertion_mode = InsertionMode::Text;
  1125. if (next_token.has_value() && next_token.value().is_character() && next_token.value().codepoint() == '\n') {
  1126. // Ignore it.
  1127. } else {
  1128. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1129. }
  1130. return;
  1131. }
  1132. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1133. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1134. close_a_p_element();
  1135. }
  1136. reconstruct_the_active_formatting_elements();
  1137. m_frameset_ok = false;
  1138. parse_generic_raw_text_element(token);
  1139. return;
  1140. }
  1141. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1142. m_frameset_ok = false;
  1143. parse_generic_raw_text_element(token);
  1144. return;
  1145. }
  1146. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1147. parse_generic_raw_text_element(token);
  1148. return;
  1149. }
  1150. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1151. reconstruct_the_active_formatting_elements();
  1152. insert_html_element(token);
  1153. m_frameset_ok = false;
  1154. switch (m_insertion_mode) {
  1155. case InsertionMode::InTable:
  1156. case InsertionMode::InCaption:
  1157. case InsertionMode::InTableBody:
  1158. case InsertionMode::InRow:
  1159. case InsertionMode::InCell:
  1160. m_insertion_mode = InsertionMode::InSelectInTable;
  1161. break;
  1162. default:
  1163. m_insertion_mode = InsertionMode::InSelect;
  1164. break;
  1165. }
  1166. return;
  1167. }
  1168. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1169. if (current_node().tag_name() == HTML::TagNames::option)
  1170. m_stack_of_open_elements.pop();
  1171. reconstruct_the_active_formatting_elements();
  1172. insert_html_element(token);
  1173. return;
  1174. }
  1175. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1176. TODO();
  1177. }
  1178. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1179. TODO();
  1180. }
  1181. if (token.is_start_tag() && token.tag_name() == "math") {
  1182. dbg() << "<math> element encountered.";
  1183. reconstruct_the_active_formatting_elements();
  1184. insert_html_element(token);
  1185. return;
  1186. }
  1187. if (token.is_start_tag() && token.tag_name() == "svg") {
  1188. dbg() << "<svg> element encountered.";
  1189. reconstruct_the_active_formatting_elements();
  1190. insert_html_element(token);
  1191. return;
  1192. }
  1193. 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))) {
  1194. PARSE_ERROR();
  1195. return;
  1196. }
  1197. // Any other start tag
  1198. if (token.is_start_tag()) {
  1199. reconstruct_the_active_formatting_elements();
  1200. insert_html_element(token);
  1201. return;
  1202. }
  1203. if (token.is_end_tag()) {
  1204. AnyOtherEndTag:
  1205. RefPtr<Element> node;
  1206. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1207. node = m_stack_of_open_elements.elements()[i];
  1208. if (node->tag_name() == token.tag_name()) {
  1209. generate_implied_end_tags(token.tag_name());
  1210. if (node != current_node()) {
  1211. PARSE_ERROR();
  1212. }
  1213. while (&current_node() != node) {
  1214. m_stack_of_open_elements.pop();
  1215. }
  1216. m_stack_of_open_elements.pop();
  1217. break;
  1218. }
  1219. if (is_special_tag(node->tag_name())) {
  1220. PARSE_ERROR();
  1221. return;
  1222. }
  1223. }
  1224. return;
  1225. }
  1226. TODO();
  1227. }
  1228. void HTMLDocumentParser::increment_script_nesting_level()
  1229. {
  1230. ++m_script_nesting_level;
  1231. }
  1232. void HTMLDocumentParser::decrement_script_nesting_level()
  1233. {
  1234. ASSERT(m_script_nesting_level);
  1235. --m_script_nesting_level;
  1236. }
  1237. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1238. {
  1239. if (token.is_character()) {
  1240. insert_character(token.codepoint());
  1241. return;
  1242. }
  1243. if (token.is_end_of_file()) {
  1244. PARSE_ERROR();
  1245. if (current_node().tag_name() == HTML::TagNames::script)
  1246. to<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1247. m_stack_of_open_elements.pop();
  1248. m_insertion_mode = m_original_insertion_mode;
  1249. process_using_the_rules_for(m_insertion_mode, token);
  1250. return;
  1251. }
  1252. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1253. NonnullRefPtr<HTMLScriptElement> script = to<HTMLScriptElement>(current_node());
  1254. m_stack_of_open_elements.pop();
  1255. m_insertion_mode = m_original_insertion_mode;
  1256. // FIXME: Handle tokenizer insertion point stuff here.
  1257. increment_script_nesting_level();
  1258. script->prepare_script({});
  1259. decrement_script_nesting_level();
  1260. if (script_nesting_level() == 0)
  1261. m_parser_pause_flag = false;
  1262. // FIXME: Handle tokenizer insertion point stuff here too.
  1263. while (document().pending_parsing_blocking_script()) {
  1264. if (script_nesting_level() != 0) {
  1265. m_parser_pause_flag = true;
  1266. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1267. // yielding control back to the caller. (Tokenization will resume when
  1268. // the caller returns to the "outer" tree construction stage.)
  1269. TODO();
  1270. } else {
  1271. auto the_script = document().take_pending_parsing_blocking_script({});
  1272. m_tokenizer.set_blocked(true);
  1273. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1274. // or the script's "ready to be parser-executed" flag is not set:
  1275. // spin the event loop until the parser's Document has no style sheet
  1276. // that is blocking scripts and the script's "ready to be parser-executed"
  1277. // flag is set.
  1278. ASSERT(the_script->is_ready_to_be_parser_executed());
  1279. if (m_aborted)
  1280. return;
  1281. m_tokenizer.set_blocked(false);
  1282. // FIXME: Handle tokenizer insertion point stuff here too.
  1283. ASSERT(script_nesting_level() == 0);
  1284. increment_script_nesting_level();
  1285. the_script->execute_script();
  1286. decrement_script_nesting_level();
  1287. ASSERT(script_nesting_level() == 0);
  1288. m_parser_pause_flag = false;
  1289. // FIXME: Handle tokenizer insertion point stuff here too.
  1290. }
  1291. }
  1292. return;
  1293. }
  1294. if (token.is_end_tag()) {
  1295. m_stack_of_open_elements.pop();
  1296. m_insertion_mode = m_original_insertion_mode;
  1297. return;
  1298. }
  1299. TODO();
  1300. }
  1301. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1302. {
  1303. while (!current_node().tag_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1304. m_stack_of_open_elements.pop();
  1305. }
  1306. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1307. {
  1308. while (!current_node().tag_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1309. m_stack_of_open_elements.pop();
  1310. }
  1311. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1312. {
  1313. while (!current_node().tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1314. m_stack_of_open_elements.pop();
  1315. }
  1316. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1317. {
  1318. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1319. clear_the_stack_back_to_a_table_row_context();
  1320. insert_html_element(token);
  1321. m_insertion_mode = InsertionMode::InCell;
  1322. m_list_of_active_formatting_elements.add_marker();
  1323. return;
  1324. }
  1325. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1326. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1327. PARSE_ERROR();
  1328. return;
  1329. }
  1330. clear_the_stack_back_to_a_table_row_context();
  1331. m_stack_of_open_elements.pop();
  1332. m_insertion_mode = InsertionMode::InTableBody;
  1333. return;
  1334. }
  1335. 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)) {
  1336. if (m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1337. PARSE_ERROR();
  1338. return;
  1339. }
  1340. clear_the_stack_back_to_a_table_row_context();
  1341. m_stack_of_open_elements.pop();
  1342. m_insertion_mode = InsertionMode::InTableBody;
  1343. process_using_the_rules_for(m_insertion_mode, token);
  1344. return;
  1345. }
  1346. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1347. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1348. PARSE_ERROR();
  1349. return;
  1350. }
  1351. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1352. return;
  1353. }
  1354. clear_the_stack_back_to_a_table_row_context();
  1355. m_stack_of_open_elements.pop();
  1356. m_insertion_mode = InsertionMode::InTableBody;
  1357. process_using_the_rules_for(m_insertion_mode, token);
  1358. return;
  1359. }
  1360. 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)) {
  1361. PARSE_ERROR();
  1362. return;
  1363. }
  1364. process_using_the_rules_for(InsertionMode::InTable, token);
  1365. }
  1366. void HTMLDocumentParser::close_the_cell()
  1367. {
  1368. generate_implied_end_tags();
  1369. if (!current_node().tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1370. PARSE_ERROR();
  1371. }
  1372. while (!current_node().tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1373. m_stack_of_open_elements.pop();
  1374. m_stack_of_open_elements.pop();
  1375. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1376. m_insertion_mode = InsertionMode::InRow;
  1377. }
  1378. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1379. {
  1380. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1381. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1382. PARSE_ERROR();
  1383. return;
  1384. }
  1385. generate_implied_end_tags();
  1386. if (current_node().tag_name() != token.tag_name()) {
  1387. PARSE_ERROR();
  1388. }
  1389. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1390. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1391. m_insertion_mode = InsertionMode::InRow;
  1392. return;
  1393. }
  1394. 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)) {
  1395. 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)) {
  1396. PARSE_ERROR();
  1397. return;
  1398. }
  1399. close_the_cell();
  1400. process_using_the_rules_for(m_insertion_mode, token);
  1401. return;
  1402. }
  1403. 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)) {
  1404. PARSE_ERROR();
  1405. return;
  1406. }
  1407. 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)) {
  1408. if (m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1409. PARSE_ERROR();
  1410. return;
  1411. }
  1412. close_the_cell();
  1413. // Reprocess the token.
  1414. process_using_the_rules_for(m_insertion_mode, token);
  1415. return;
  1416. }
  1417. process_using_the_rules_for(InsertionMode::InBody, token);
  1418. }
  1419. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1420. {
  1421. if (token.is_character()) {
  1422. if (token.codepoint() == 0) {
  1423. PARSE_ERROR();
  1424. return;
  1425. }
  1426. m_pending_table_character_tokens.append(token);
  1427. return;
  1428. }
  1429. for (auto& pending_token : m_pending_table_character_tokens) {
  1430. ASSERT(pending_token.is_character());
  1431. if (!pending_token.is_parser_whitespace()) {
  1432. // FIXME: If any of the tokens in the pending table character tokens list
  1433. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1434. // reprocess the character tokens in the pending table character tokens list using
  1435. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1436. TODO();
  1437. }
  1438. }
  1439. for (auto& pending_token : m_pending_table_character_tokens) {
  1440. insert_character(pending_token.codepoint());
  1441. }
  1442. m_insertion_mode = m_original_insertion_mode;
  1443. process_using_the_rules_for(m_insertion_mode, token);
  1444. }
  1445. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1446. {
  1447. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1448. clear_the_stack_back_to_a_table_body_context();
  1449. insert_html_element(token);
  1450. m_insertion_mode = InsertionMode::InRow;
  1451. return;
  1452. }
  1453. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1454. PARSE_ERROR();
  1455. clear_the_stack_back_to_a_table_body_context();
  1456. HTMLToken fake_tr_token;
  1457. fake_tr_token.m_type = HTMLToken::Type::StartTag;
  1458. fake_tr_token.m_tag.tag_name.append(HTML::TagNames::tr);
  1459. insert_html_element(fake_tr_token);
  1460. m_insertion_mode = InsertionMode::InRow;
  1461. process_using_the_rules_for(m_insertion_mode, token);
  1462. return;
  1463. }
  1464. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1465. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1466. PARSE_ERROR();
  1467. return;
  1468. }
  1469. clear_the_stack_back_to_a_table_body_context();
  1470. m_stack_of_open_elements.pop();
  1471. m_insertion_mode = InsertionMode::InTable;
  1472. return;
  1473. }
  1474. 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))
  1475. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1476. // FIXME: If the stack of open elements does not have a tbody, thead, or tfoot element in table scope, this is a parse error; ignore the token.
  1477. clear_the_stack_back_to_a_table_body_context();
  1478. m_stack_of_open_elements.pop();
  1479. m_insertion_mode = InsertionMode::InTable;
  1480. process_using_the_rules_for(InsertionMode::InTable, token);
  1481. return;
  1482. }
  1483. 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)) {
  1484. PARSE_ERROR();
  1485. return;
  1486. }
  1487. process_using_the_rules_for(InsertionMode::InTable, token);
  1488. }
  1489. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1490. {
  1491. if (token.is_character() && current_node().tag_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1492. m_pending_table_character_tokens.clear();
  1493. m_original_insertion_mode = m_insertion_mode;
  1494. m_insertion_mode = InsertionMode::InTableText;
  1495. process_using_the_rules_for(InsertionMode::InTableText, token);
  1496. return;
  1497. }
  1498. if (token.is_comment()) {
  1499. insert_comment(token);
  1500. return;
  1501. }
  1502. if (token.is_doctype()) {
  1503. PARSE_ERROR();
  1504. return;
  1505. }
  1506. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1507. TODO();
  1508. }
  1509. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1510. TODO();
  1511. }
  1512. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1513. TODO();
  1514. }
  1515. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1516. clear_the_stack_back_to_a_table_context();
  1517. insert_html_element(token);
  1518. m_insertion_mode = InsertionMode::InTableBody;
  1519. return;
  1520. }
  1521. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1522. clear_the_stack_back_to_a_table_context();
  1523. HTMLToken fake_tbody_token;
  1524. fake_tbody_token.m_type = HTMLToken::Type::StartTag;
  1525. fake_tbody_token.m_tag.tag_name.append(HTML::TagNames::tbody);
  1526. insert_html_element(fake_tbody_token);
  1527. m_insertion_mode = InsertionMode::InTableBody;
  1528. process_using_the_rules_for(InsertionMode::InTableBody, token);
  1529. return;
  1530. }
  1531. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1532. PARSE_ERROR();
  1533. TODO();
  1534. }
  1535. if (token.is_end_tag()) {
  1536. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1537. PARSE_ERROR();
  1538. return;
  1539. }
  1540. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1541. reset_the_insertion_mode_appropriately();
  1542. return;
  1543. }
  1544. TODO();
  1545. }
  1546. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1547. {
  1548. (void)token;
  1549. TODO();
  1550. }
  1551. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  1552. {
  1553. if (token.is_character()) {
  1554. if (token.codepoint() == 0) {
  1555. PARSE_ERROR();
  1556. return;
  1557. }
  1558. insert_character(token.codepoint());
  1559. return;
  1560. }
  1561. if (token.is_comment()) {
  1562. insert_comment(token);
  1563. return;
  1564. }
  1565. if (token.is_doctype()) {
  1566. PARSE_ERROR();
  1567. return;
  1568. }
  1569. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  1570. process_using_the_rules_for(InsertionMode::InBody, token);
  1571. return;
  1572. }
  1573. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  1574. if (current_node().tag_name() == HTML::TagNames::option) {
  1575. m_stack_of_open_elements.pop();
  1576. }
  1577. insert_html_element(token);
  1578. return;
  1579. }
  1580. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1581. if (current_node().tag_name() == HTML::TagNames::option) {
  1582. m_stack_of_open_elements.pop();
  1583. }
  1584. if (current_node().tag_name() == HTML::TagNames::optgroup) {
  1585. m_stack_of_open_elements.pop();
  1586. }
  1587. insert_html_element(token);
  1588. return;
  1589. }
  1590. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1591. if (current_node().tag_name() == HTML::TagNames::option && node_before_current_node().tag_name() == HTML::TagNames::optgroup)
  1592. m_stack_of_open_elements.pop();
  1593. if (current_node().tag_name() == HTML::TagNames::optgroup) {
  1594. m_stack_of_open_elements.pop();
  1595. } else {
  1596. PARSE_ERROR();
  1597. return;
  1598. }
  1599. return;
  1600. }
  1601. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  1602. if (current_node().tag_name() == HTML::TagNames::option) {
  1603. m_stack_of_open_elements.pop();
  1604. } else {
  1605. PARSE_ERROR();
  1606. return;
  1607. }
  1608. return;
  1609. }
  1610. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  1611. if (m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  1612. PARSE_ERROR();
  1613. return;
  1614. }
  1615. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1616. reset_the_insertion_mode_appropriately();
  1617. return;
  1618. }
  1619. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1620. PARSE_ERROR();
  1621. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select))
  1622. return;
  1623. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1624. reset_the_insertion_mode_appropriately();
  1625. return;
  1626. }
  1627. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  1628. PARSE_ERROR();
  1629. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  1630. return;
  1631. }
  1632. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1633. reset_the_insertion_mode_appropriately();
  1634. process_using_the_rules_for(m_insertion_mode, token);
  1635. return;
  1636. }
  1637. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  1638. process_using_the_rules_for(InsertionMode::InHead, token);
  1639. return;
  1640. }
  1641. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  1642. process_using_the_rules_for(InsertionMode::InHead, token);
  1643. return;
  1644. }
  1645. if (token.is_end_of_file()) {
  1646. process_using_the_rules_for(InsertionMode::InBody, token);
  1647. return;
  1648. }
  1649. PARSE_ERROR();
  1650. }
  1651. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  1652. {
  1653. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1654. RefPtr<Element> node = m_stack_of_open_elements.elements().at(i);
  1655. if (node->tag_name() == HTML::TagNames::select) {
  1656. TODO();
  1657. }
  1658. if (node->tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1659. m_insertion_mode = InsertionMode::InCell;
  1660. return;
  1661. }
  1662. if (node->tag_name() == HTML::TagNames::tr) {
  1663. m_insertion_mode = InsertionMode::InRow;
  1664. return;
  1665. }
  1666. if (node->tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  1667. m_insertion_mode = InsertionMode::InTableBody;
  1668. return;
  1669. }
  1670. if (node->tag_name() == HTML::TagNames::caption) {
  1671. m_insertion_mode = InsertionMode::InCaption;
  1672. return;
  1673. }
  1674. if (node->tag_name() == HTML::TagNames::colgroup) {
  1675. m_insertion_mode = InsertionMode::InColumnGroup;
  1676. return;
  1677. }
  1678. if (node->tag_name() == HTML::TagNames::table) {
  1679. m_insertion_mode = InsertionMode::InTable;
  1680. return;
  1681. }
  1682. if (node->tag_name() == HTML::TagNames::template_) {
  1683. TODO();
  1684. }
  1685. if (node->tag_name() == HTML::TagNames::body) {
  1686. m_insertion_mode = InsertionMode::InBody;
  1687. return;
  1688. }
  1689. if (node->tag_name() == HTML::TagNames::frameset) {
  1690. m_insertion_mode = InsertionMode::InFrameset;
  1691. if (m_parsing_fragment) {
  1692. TODO();
  1693. }
  1694. return;
  1695. }
  1696. if (node->tag_name() == HTML::TagNames::html) {
  1697. TODO();
  1698. }
  1699. }
  1700. m_insertion_mode = InsertionMode::InBody;
  1701. if (m_parsing_fragment) {
  1702. TODO();
  1703. }
  1704. }
  1705. const char* HTMLDocumentParser::insertion_mode_name() const
  1706. {
  1707. switch (m_insertion_mode) {
  1708. #define __ENUMERATE_INSERTION_MODE(mode) \
  1709. case InsertionMode::mode: \
  1710. return #mode;
  1711. ENUMERATE_INSERTION_MODES
  1712. #undef __ENUMERATE_INSERTION_MODE
  1713. }
  1714. ASSERT_NOT_REACHED();
  1715. }
  1716. Document& HTMLDocumentParser::document()
  1717. {
  1718. return *m_document;
  1719. }
  1720. }