HTMLDocumentParser.cpp 74 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248124912501251125212531254125512561257125812591260126112621263126412651266126712681269127012711272127312741275127612771278127912801281128212831284128512861287128812891290129112921293129412951296129712981299130013011302130313041305130613071308130913101311131213131314131513161317131813191320132113221323132413251326132713281329133013311332133313341335133613371338133913401341134213431344134513461347134813491350135113521353135413551356135713581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386138713881389139013911392139313941395139613971398139914001401140214031404140514061407140814091410141114121413141414151416141714181419142014211422142314241425142614271428142914301431143214331434143514361437143814391440144114421443144414451446144714481449145014511452145314541455145614571458145914601461146214631464146514661467146814691470147114721473147414751476147714781479148014811482148314841485148614871488148914901491149214931494149514961497149814991500150115021503150415051506150715081509151015111512151315141515151615171518151915201521152215231524152515261527152815291530153115321533153415351536153715381539154015411542154315441545154615471548154915501551155215531554155515561557155815591560156115621563156415651566156715681569157015711572157315741575157615771578157915801581158215831584158515861587158815891590159115921593159415951596159715981599160016011602160316041605160616071608160916101611161216131614161516161617161816191620162116221623162416251626162716281629163016311632163316341635163616371638163916401641164216431644164516461647164816491650165116521653165416551656165716581659166016611662166316641665166616671668166916701671167216731674167516761677167816791680168116821683168416851686168716881689169016911692169316941695169616971698169917001701170217031704170517061707170817091710171117121713171417151716171717181719172017211722172317241725172617271728172917301731173217331734173517361737173817391740174117421743174417451746174717481749175017511752175317541755175617571758175917601761176217631764176517661767176817691770177117721773177417751776177717781779178017811782178317841785178617871788178917901791179217931794179517961797179817991800180118021803180418051806180718081809181018111812181318141815181618171818181918201821182218231824182518261827182818291830183118321833183418351836183718381839184018411842184318441845184618471848184918501851185218531854185518561857185818591860186118621863186418651866186718681869187018711872187318741875187618771878187918801881188218831884188518861887188818891890189118921893189418951896189718981899190019011902190319041905190619071908190919101911191219131914191519161917191819191920192119221923192419251926192719281929193019311932193319341935193619371938193919401941194219431944194519461947194819491950195119521953195419551956195719581959196019611962196319641965196619671968196919701971197219731974197519761977197819791980198119821983198419851986198719881989199019911992199319941995199619971998199920002001200220032004200520062007200820092010201120122013201420152016201720182019202020212022202320242025202620272028202920302031203220332034203520362037203820392040204120422043204420452046204720482049
  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. case InsertionMode::InCaption:
  137. handle_in_caption(token);
  138. break;
  139. default:
  140. ASSERT_NOT_REACHED();
  141. }
  142. }
  143. void HTMLDocumentParser::handle_initial(HTMLToken& token)
  144. {
  145. if (token.is_character() && token.is_parser_whitespace()) {
  146. return;
  147. }
  148. if (token.is_comment()) {
  149. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  150. document().append_child(move(comment));
  151. return;
  152. }
  153. if (token.is_doctype()) {
  154. auto doctype = adopt(*new DocumentType(document()));
  155. doctype->set_name(token.m_doctype.name.to_string());
  156. document().append_child(move(doctype));
  157. document().set_quirks_mode(token.m_doctype.force_quirks);
  158. m_insertion_mode = InsertionMode::BeforeHTML;
  159. return;
  160. }
  161. PARSE_ERROR();
  162. document().set_quirks_mode(true);
  163. m_insertion_mode = InsertionMode::BeforeHTML;
  164. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  165. }
  166. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  167. {
  168. if (token.is_doctype()) {
  169. PARSE_ERROR();
  170. return;
  171. }
  172. if (token.is_comment()) {
  173. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  174. document().append_child(move(comment));
  175. return;
  176. }
  177. if (token.is_character() && token.is_parser_whitespace()) {
  178. return;
  179. }
  180. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  181. auto element = create_element_for(token);
  182. document().append_child(element);
  183. m_stack_of_open_elements.push(move(element));
  184. m_insertion_mode = InsertionMode::BeforeHead;
  185. return;
  186. }
  187. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  188. goto AnythingElse;
  189. }
  190. if (token.is_end_tag()) {
  191. PARSE_ERROR();
  192. return;
  193. }
  194. AnythingElse:
  195. auto element = create_element(document(), HTML::TagNames::html);
  196. document().append_child(element);
  197. m_stack_of_open_elements.push(element);
  198. // 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.
  199. m_insertion_mode = InsertionMode::BeforeHead;
  200. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  201. return;
  202. }
  203. Element& HTMLDocumentParser::current_node()
  204. {
  205. return m_stack_of_open_elements.current_node();
  206. }
  207. Element& HTMLDocumentParser::node_before_current_node()
  208. {
  209. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  210. }
  211. RefPtr<Node> HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  212. {
  213. auto& target = current_node();
  214. if (m_foster_parenting) {
  215. TODO();
  216. }
  217. return target;
  218. }
  219. NonnullRefPtr<Element> HTMLDocumentParser::create_element_for(HTMLToken& token)
  220. {
  221. auto element = create_element(document(), token.tag_name());
  222. for (auto& attribute : token.m_tag.attributes) {
  223. element->set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  224. }
  225. return element;
  226. }
  227. RefPtr<Element> HTMLDocumentParser::insert_html_element(HTMLToken& token)
  228. {
  229. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  230. auto element = create_element_for(token);
  231. // FIXME: Check if it's possible to insert `element` at `adjusted_insertion_location`
  232. adjusted_insertion_location->append_child(element);
  233. m_stack_of_open_elements.push(element);
  234. return element;
  235. }
  236. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  237. {
  238. if (token.is_character() && token.is_parser_whitespace()) {
  239. return;
  240. }
  241. if (token.is_comment()) {
  242. insert_comment(token);
  243. return;
  244. }
  245. if (token.is_doctype()) {
  246. PARSE_ERROR();
  247. return;
  248. }
  249. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  250. process_using_the_rules_for(InsertionMode::InBody, token);
  251. return;
  252. }
  253. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  254. auto element = insert_html_element(token);
  255. m_head_element = to<HTMLHeadElement>(element);
  256. m_insertion_mode = InsertionMode::InHead;
  257. return;
  258. }
  259. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  260. goto AnythingElse;
  261. }
  262. if (token.is_end_tag()) {
  263. PARSE_ERROR();
  264. return;
  265. }
  266. AnythingElse:
  267. HTMLToken fake_head_token;
  268. fake_head_token.m_type = HTMLToken::Type::StartTag;
  269. fake_head_token.m_tag.tag_name.append(HTML::TagNames::head);
  270. m_head_element = to<HTMLHeadElement>(insert_html_element(fake_head_token));
  271. m_insertion_mode = InsertionMode::InHead;
  272. process_using_the_rules_for(InsertionMode::InHead, token);
  273. return;
  274. }
  275. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  276. {
  277. auto data = token.m_comment_or_character.data.to_string();
  278. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  279. adjusted_insertion_location->append_child(adopt(*new Comment(document(), data)));
  280. }
  281. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  282. {
  283. if (token.is_parser_whitespace()) {
  284. insert_character(token.codepoint());
  285. return;
  286. }
  287. if (token.is_comment()) {
  288. insert_comment(token);
  289. return;
  290. }
  291. if (token.is_doctype()) {
  292. PARSE_ERROR();
  293. return;
  294. }
  295. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  296. process_using_the_rules_for(InsertionMode::InBody, token);
  297. return;
  298. }
  299. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  300. insert_html_element(token);
  301. m_stack_of_open_elements.pop();
  302. token.acknowledge_self_closing_flag_if_set();
  303. return;
  304. }
  305. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  306. auto element = insert_html_element(token);
  307. m_stack_of_open_elements.pop();
  308. token.acknowledge_self_closing_flag_if_set();
  309. return;
  310. }
  311. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  312. insert_html_element(token);
  313. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  314. m_original_insertion_mode = m_insertion_mode;
  315. m_insertion_mode = InsertionMode::Text;
  316. return;
  317. }
  318. 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)) {
  319. parse_generic_raw_text_element(token);
  320. return;
  321. }
  322. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  323. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  324. auto element = create_element_for(token);
  325. auto& script_element = to<HTMLScriptElement>(*element);
  326. script_element.set_parser_document({}, document());
  327. script_element.set_non_blocking({}, false);
  328. if (m_parsing_fragment) {
  329. TODO();
  330. }
  331. if (m_invoked_via_document_write) {
  332. TODO();
  333. }
  334. adjusted_insertion_location->append_child(element, false);
  335. m_stack_of_open_elements.push(element);
  336. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  337. m_original_insertion_mode = m_insertion_mode;
  338. m_insertion_mode = InsertionMode::Text;
  339. return;
  340. }
  341. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  342. m_stack_of_open_elements.pop();
  343. m_insertion_mode = InsertionMode::AfterHead;
  344. return;
  345. }
  346. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  347. TODO();
  348. }
  349. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  350. // FIXME: Support this properly
  351. insert_html_element(token);
  352. return;
  353. }
  354. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  355. // FIXME: Support this properly
  356. ASSERT(current_node().tag_name() == HTML::TagNames::template_);
  357. m_stack_of_open_elements.pop();
  358. return;
  359. }
  360. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  361. PARSE_ERROR();
  362. return;
  363. }
  364. m_stack_of_open_elements.pop();
  365. m_insertion_mode = InsertionMode::AfterHead;
  366. process_using_the_rules_for(m_insertion_mode, token);
  367. }
  368. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken&)
  369. {
  370. TODO();
  371. }
  372. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  373. {
  374. insert_html_element(token);
  375. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  376. m_original_insertion_mode = m_insertion_mode;
  377. m_insertion_mode = InsertionMode::Text;
  378. }
  379. Text* HTMLDocumentParser::find_character_insertion_node()
  380. {
  381. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  382. if (adjusted_insertion_location->is_document())
  383. return nullptr;
  384. if (adjusted_insertion_location->last_child() && adjusted_insertion_location->last_child()->is_text())
  385. return to<Text>(adjusted_insertion_location->last_child());
  386. auto new_text_node = adopt(*new Text(document(), ""));
  387. adjusted_insertion_location->append_child(new_text_node);
  388. return new_text_node;
  389. }
  390. void HTMLDocumentParser::flush_character_insertions()
  391. {
  392. if (m_character_insertion_builder.is_empty())
  393. return;
  394. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  395. m_character_insertion_node->parent()->children_changed();
  396. m_character_insertion_builder.clear();
  397. }
  398. void HTMLDocumentParser::insert_character(u32 data)
  399. {
  400. auto node = find_character_insertion_node();
  401. if (node == m_character_insertion_node) {
  402. m_character_insertion_builder.append(Utf32View { &data, 1 });
  403. return;
  404. }
  405. if (!m_character_insertion_node) {
  406. m_character_insertion_node = node;
  407. m_character_insertion_builder.append(Utf32View { &data, 1 });
  408. return;
  409. }
  410. flush_character_insertions();
  411. m_character_insertion_node = node;
  412. m_character_insertion_builder.append(Utf32View { &data, 1 });
  413. }
  414. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  415. {
  416. if (token.is_character() && token.is_parser_whitespace()) {
  417. insert_character(token.codepoint());
  418. return;
  419. }
  420. if (token.is_comment()) {
  421. insert_comment(token);
  422. return;
  423. }
  424. if (token.is_doctype()) {
  425. PARSE_ERROR();
  426. return;
  427. }
  428. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  429. process_using_the_rules_for(InsertionMode::InBody, token);
  430. return;
  431. }
  432. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  433. insert_html_element(token);
  434. m_frameset_ok = false;
  435. m_insertion_mode = InsertionMode::InBody;
  436. return;
  437. }
  438. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  439. insert_html_element(token);
  440. m_insertion_mode = InsertionMode::InFrameset;
  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, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  444. PARSE_ERROR();
  445. m_stack_of_open_elements.push(*m_head_element);
  446. process_using_the_rules_for(InsertionMode::InHead, token);
  447. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  448. return entry.ptr() == m_head_element;
  449. });
  450. return;
  451. }
  452. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  453. TODO();
  454. }
  455. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  456. goto AnythingElse;
  457. }
  458. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  459. PARSE_ERROR();
  460. return;
  461. }
  462. AnythingElse:
  463. HTMLToken fake_body_token;
  464. fake_body_token.m_type = HTMLToken::Type::StartTag;
  465. fake_body_token.m_tag.tag_name.append(HTML::TagNames::body);
  466. insert_html_element(fake_body_token);
  467. m_insertion_mode = InsertionMode::InBody;
  468. process_using_the_rules_for(m_insertion_mode, token);
  469. }
  470. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  471. {
  472. 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))
  473. m_stack_of_open_elements.pop();
  474. }
  475. void HTMLDocumentParser::close_a_p_element()
  476. {
  477. generate_implied_end_tags(HTML::TagNames::p);
  478. if (current_node().tag_name() != HTML::TagNames::p) {
  479. PARSE_ERROR();
  480. }
  481. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  482. }
  483. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  484. {
  485. if (token.is_character() && token.is_parser_whitespace()) {
  486. process_using_the_rules_for(InsertionMode::InBody, token);
  487. return;
  488. }
  489. if (token.is_comment()) {
  490. TODO();
  491. }
  492. if (token.is_doctype()) {
  493. PARSE_ERROR();
  494. return;
  495. }
  496. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  497. process_using_the_rules_for(InsertionMode::InBody, token);
  498. return;
  499. }
  500. if (token.is_end_of_file()) {
  501. stop_parsing();
  502. return;
  503. }
  504. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  505. if (m_parsing_fragment) {
  506. TODO();
  507. }
  508. m_insertion_mode = InsertionMode::AfterAfterBody;
  509. return;
  510. }
  511. PARSE_ERROR();
  512. m_insertion_mode = InsertionMode::InBody;
  513. process_using_the_rules_for(InsertionMode::InBody, token);
  514. }
  515. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  516. {
  517. if (token.is_comment()) {
  518. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  519. document().append_child(move(comment));
  520. return;
  521. }
  522. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  523. process_using_the_rules_for(InsertionMode::InBody, token);
  524. return;
  525. }
  526. if (token.is_end_of_file()) {
  527. stop_parsing();
  528. return;
  529. }
  530. PARSE_ERROR();
  531. m_insertion_mode = InsertionMode::InBody;
  532. process_using_the_rules_for(m_insertion_mode, token);
  533. }
  534. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  535. {
  536. // FIXME: This needs to care about "markers"
  537. if (m_list_of_active_formatting_elements.is_empty())
  538. return;
  539. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  540. return;
  541. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  542. return;
  543. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  544. RefPtr<Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  545. ASSERT(entry);
  546. Rewind:
  547. if (index == 0) {
  548. goto Create;
  549. }
  550. --index;
  551. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  552. ASSERT(entry);
  553. if (!m_stack_of_open_elements.contains(*entry))
  554. goto Rewind;
  555. Advance:
  556. ++index;
  557. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  558. ASSERT(entry);
  559. Create:
  560. // FIXME: Hold on to the real token!
  561. HTMLToken fake_token;
  562. fake_token.m_type = HTMLToken::Type::StartTag;
  563. fake_token.m_tag.tag_name.append(entry->tag_name());
  564. auto new_element = insert_html_element(fake_token);
  565. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  566. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  567. goto Advance;
  568. }
  569. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  570. {
  571. auto subject = token.tag_name();
  572. // If the current node is an HTML element whose tag name is subject,
  573. // and the current node is not in the list of active formatting elements,
  574. // then pop the current node off the stack of open elements, and return.
  575. if (current_node().tag_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  576. m_stack_of_open_elements.pop();
  577. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  578. }
  579. size_t outer_loop_counter = 0;
  580. //OuterLoop:
  581. if (outer_loop_counter >= 8)
  582. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  583. ++outer_loop_counter;
  584. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  585. if (!formatting_element)
  586. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  587. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  588. PARSE_ERROR();
  589. // FIXME: If formatting element is not in the stack of open elements,
  590. // then this is a parse error; remove the element from the list, and return.
  591. TODO();
  592. }
  593. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  594. PARSE_ERROR();
  595. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  596. }
  597. if (formatting_element != &current_node()) {
  598. PARSE_ERROR();
  599. }
  600. RefPtr<Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  601. if (!furthest_block) {
  602. while (&current_node() != formatting_element)
  603. m_stack_of_open_elements.pop();
  604. m_stack_of_open_elements.pop();
  605. m_list_of_active_formatting_elements.remove(*formatting_element);
  606. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  607. }
  608. // FIXME: Implement the rest of the AAA :^)
  609. TODO();
  610. }
  611. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name)
  612. {
  613. return tag_name.is_one_of(
  614. HTML::TagNames::address,
  615. HTML::TagNames::applet,
  616. HTML::TagNames::area,
  617. HTML::TagNames::article,
  618. HTML::TagNames::aside,
  619. HTML::TagNames::base,
  620. HTML::TagNames::basefont,
  621. HTML::TagNames::bgsound,
  622. HTML::TagNames::blockquote,
  623. HTML::TagNames::body,
  624. HTML::TagNames::br,
  625. HTML::TagNames::button,
  626. HTML::TagNames::caption,
  627. HTML::TagNames::center,
  628. HTML::TagNames::col,
  629. HTML::TagNames::colgroup,
  630. HTML::TagNames::dd,
  631. HTML::TagNames::details,
  632. HTML::TagNames::dir,
  633. HTML::TagNames::div,
  634. HTML::TagNames::dl,
  635. HTML::TagNames::dt,
  636. HTML::TagNames::embed,
  637. HTML::TagNames::fieldset,
  638. HTML::TagNames::figcaption,
  639. HTML::TagNames::figure,
  640. HTML::TagNames::footer,
  641. HTML::TagNames::form,
  642. HTML::TagNames::frame,
  643. HTML::TagNames::frameset,
  644. HTML::TagNames::h1,
  645. HTML::TagNames::h2,
  646. HTML::TagNames::h3,
  647. HTML::TagNames::h4,
  648. HTML::TagNames::h5,
  649. HTML::TagNames::h6,
  650. HTML::TagNames::head,
  651. HTML::TagNames::header,
  652. HTML::TagNames::hgroup,
  653. HTML::TagNames::hr,
  654. HTML::TagNames::html,
  655. HTML::TagNames::iframe,
  656. HTML::TagNames::img,
  657. HTML::TagNames::input,
  658. HTML::TagNames::keygen,
  659. HTML::TagNames::li,
  660. HTML::TagNames::link,
  661. HTML::TagNames::listing,
  662. HTML::TagNames::main,
  663. HTML::TagNames::marquee,
  664. HTML::TagNames::menu,
  665. HTML::TagNames::meta,
  666. HTML::TagNames::nav,
  667. HTML::TagNames::noembed,
  668. HTML::TagNames::noframes,
  669. HTML::TagNames::noscript,
  670. HTML::TagNames::object,
  671. HTML::TagNames::ol,
  672. HTML::TagNames::p,
  673. HTML::TagNames::param,
  674. HTML::TagNames::plaintext,
  675. HTML::TagNames::pre,
  676. HTML::TagNames::script,
  677. HTML::TagNames::section,
  678. HTML::TagNames::select,
  679. HTML::TagNames::source,
  680. HTML::TagNames::style,
  681. HTML::TagNames::summary,
  682. HTML::TagNames::table,
  683. HTML::TagNames::tbody,
  684. HTML::TagNames::td,
  685. HTML::TagNames::template_,
  686. HTML::TagNames::textarea,
  687. HTML::TagNames::tfoot,
  688. HTML::TagNames::th,
  689. HTML::TagNames::thead,
  690. HTML::TagNames::title,
  691. HTML::TagNames::tr,
  692. HTML::TagNames::track,
  693. HTML::TagNames::ul,
  694. HTML::TagNames::wbr,
  695. HTML::TagNames::xmp);
  696. }
  697. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  698. {
  699. if (token.is_character()) {
  700. if (token.codepoint() == 0) {
  701. PARSE_ERROR();
  702. return;
  703. }
  704. if (token.is_parser_whitespace()) {
  705. reconstruct_the_active_formatting_elements();
  706. insert_character(token.codepoint());
  707. return;
  708. }
  709. reconstruct_the_active_formatting_elements();
  710. insert_character(token.codepoint());
  711. m_frameset_ok = false;
  712. return;
  713. }
  714. if (token.is_comment()) {
  715. insert_comment(token);
  716. return;
  717. }
  718. if (token.is_doctype()) {
  719. PARSE_ERROR();
  720. return;
  721. }
  722. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  723. PARSE_ERROR();
  724. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  725. return;
  726. for (auto& attribute : token.m_tag.attributes) {
  727. if (current_node().has_attribute(attribute.name_builder.string_view()))
  728. continue;
  729. current_node().set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  730. }
  731. return;
  732. }
  733. 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)) {
  734. process_using_the_rules_for(InsertionMode::InHead, token);
  735. return;
  736. }
  737. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  738. process_using_the_rules_for(InsertionMode::InHead, token);
  739. return;
  740. }
  741. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  742. PARSE_ERROR();
  743. if (m_stack_of_open_elements.elements().size() == 1
  744. || node_before_current_node().tag_name() != HTML::TagNames::body
  745. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  746. return;
  747. }
  748. m_frameset_ok = false;
  749. for (auto& attribute : token.m_tag.attributes) {
  750. if (node_before_current_node().has_attribute(attribute.name_builder.string_view()))
  751. continue;
  752. node_before_current_node().set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  753. }
  754. return;
  755. }
  756. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  757. TODO();
  758. }
  759. if (token.is_end_of_file()) {
  760. // FIXME: If the stack of template insertion modes is not empty,
  761. // then process the token using the rules for the "in template" insertion mode.
  762. // FIXME: If there is a node in the stack of open elements that is not either
  763. // a dd element, a dt element, an li element, an optgroup element, an option element,
  764. // a p element, an rb element, an rp element, an rt element, an rtc element,
  765. // a tbody element, a td element, a tfoot element, a th element, a thead element,
  766. // a tr element, the body element, or the html element, then this is a parse error.
  767. stop_parsing();
  768. return;
  769. }
  770. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  771. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  772. PARSE_ERROR();
  773. return;
  774. }
  775. for (auto& node : m_stack_of_open_elements.elements()) {
  776. 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)) {
  777. PARSE_ERROR();
  778. break;
  779. }
  780. }
  781. m_insertion_mode = InsertionMode::AfterBody;
  782. return;
  783. }
  784. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  785. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  786. PARSE_ERROR();
  787. return;
  788. }
  789. for (auto& node : m_stack_of_open_elements.elements()) {
  790. 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)) {
  791. PARSE_ERROR();
  792. break;
  793. }
  794. }
  795. m_insertion_mode = InsertionMode::AfterBody;
  796. process_using_the_rules_for(m_insertion_mode, token);
  797. return;
  798. }
  799. 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)) {
  800. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  801. close_a_p_element();
  802. insert_html_element(token);
  803. return;
  804. }
  805. 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)) {
  806. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  807. close_a_p_element();
  808. 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)) {
  809. PARSE_ERROR();
  810. m_stack_of_open_elements.pop();
  811. }
  812. insert_html_element(token);
  813. return;
  814. }
  815. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  816. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  817. close_a_p_element();
  818. insert_html_element(token);
  819. m_frameset_ok = false;
  820. // If the next token is a U+000A LINE FEED (LF) character token,
  821. // then ignore that token and move on to the next one.
  822. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  823. auto next_token = m_tokenizer.next_token();
  824. if (next_token.has_value() && next_token.value().is_character() && next_token.value().codepoint() == '\n') {
  825. // Ignore it.
  826. } else {
  827. process_using_the_rules_for(m_insertion_mode, next_token.value());
  828. }
  829. return;
  830. }
  831. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  832. if (m_form_element && m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  833. PARSE_ERROR();
  834. return;
  835. }
  836. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  837. close_a_p_element();
  838. auto element = insert_html_element(token);
  839. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  840. m_form_element = to<HTMLFormElement>(*element);
  841. return;
  842. }
  843. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  844. m_frameset_ok = false;
  845. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  846. RefPtr<Element> node = m_stack_of_open_elements.elements()[i];
  847. if (node->tag_name() == HTML::TagNames::li) {
  848. generate_implied_end_tags(HTML::TagNames::li);
  849. if (current_node().tag_name() != HTML::TagNames::li) {
  850. PARSE_ERROR();
  851. }
  852. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  853. break;
  854. }
  855. if (is_special_tag(node->tag_name()) && !node->tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  856. break;
  857. }
  858. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  859. close_a_p_element();
  860. insert_html_element(token);
  861. return;
  862. }
  863. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  864. m_frameset_ok = false;
  865. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  866. RefPtr<Element> node = m_stack_of_open_elements.elements()[i];
  867. if (node->tag_name() == HTML::TagNames::dd) {
  868. generate_implied_end_tags(HTML::TagNames::dd);
  869. if (current_node().tag_name() != HTML::TagNames::dd) {
  870. PARSE_ERROR();
  871. }
  872. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  873. break;
  874. }
  875. if (node->tag_name() == HTML::TagNames::dt) {
  876. generate_implied_end_tags(HTML::TagNames::dt);
  877. if (current_node().tag_name() != HTML::TagNames::dt) {
  878. PARSE_ERROR();
  879. }
  880. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  881. break;
  882. }
  883. if (is_special_tag(node->tag_name()) && !node->tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  884. break;
  885. }
  886. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  887. close_a_p_element();
  888. insert_html_element(token);
  889. return;
  890. }
  891. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  892. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  893. close_a_p_element();
  894. insert_html_element(token);
  895. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  896. return;
  897. }
  898. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  899. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  900. PARSE_ERROR();
  901. generate_implied_end_tags();
  902. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  903. }
  904. reconstruct_the_active_formatting_elements();
  905. insert_html_element(token);
  906. m_frameset_ok = false;
  907. return;
  908. }
  909. 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)) {
  910. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  911. PARSE_ERROR();
  912. return;
  913. }
  914. generate_implied_end_tags();
  915. if (current_node().tag_name() != token.tag_name()) {
  916. PARSE_ERROR();
  917. }
  918. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  919. return;
  920. }
  921. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  922. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  923. auto node = m_form_element;
  924. m_form_element = nullptr;
  925. if (!node || m_stack_of_open_elements.has_in_scope(*node)) {
  926. PARSE_ERROR();
  927. return;
  928. }
  929. generate_implied_end_tags();
  930. if (&current_node() != node) {
  931. PARSE_ERROR();
  932. }
  933. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  934. } else {
  935. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  936. PARSE_ERROR();
  937. return;
  938. }
  939. generate_implied_end_tags();
  940. if (current_node().tag_name() != HTML::TagNames::form) {
  941. PARSE_ERROR();
  942. }
  943. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  944. }
  945. return;
  946. }
  947. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  948. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  949. PARSE_ERROR();
  950. HTMLToken fake_p_token;
  951. fake_p_token.m_type = HTMLToken::Type::StartTag;
  952. fake_p_token.m_tag.tag_name.append(HTML::TagNames::p);
  953. insert_html_element(fake_p_token);
  954. }
  955. close_a_p_element();
  956. return;
  957. }
  958. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  959. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  960. PARSE_ERROR();
  961. return;
  962. }
  963. generate_implied_end_tags(HTML::TagNames::li);
  964. if (current_node().tag_name() != HTML::TagNames::li) {
  965. PARSE_ERROR();
  966. dbg() << "Expected <li> current node, but had <" << current_node().tag_name() << ">";
  967. }
  968. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  969. return;
  970. }
  971. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  972. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  973. PARSE_ERROR();
  974. return;
  975. }
  976. generate_implied_end_tags(token.tag_name());
  977. if (current_node().tag_name() != token.tag_name()) {
  978. PARSE_ERROR();
  979. }
  980. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  981. return;
  982. }
  983. 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)) {
  984. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  985. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  986. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  987. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  988. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  989. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  990. PARSE_ERROR();
  991. return;
  992. }
  993. generate_implied_end_tags();
  994. if (current_node().tag_name() != token.tag_name()) {
  995. PARSE_ERROR();
  996. }
  997. for (;;) {
  998. auto popped_element = m_stack_of_open_elements.pop();
  999. 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))
  1000. break;
  1001. }
  1002. return;
  1003. }
  1004. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1005. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1006. PARSE_ERROR();
  1007. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1008. goto AnyOtherEndTag;
  1009. m_list_of_active_formatting_elements.remove(*element);
  1010. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1011. return entry.ptr() == element;
  1012. });
  1013. }
  1014. reconstruct_the_active_formatting_elements();
  1015. auto element = insert_html_element(token);
  1016. m_list_of_active_formatting_elements.add(*element);
  1017. return;
  1018. }
  1019. 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)) {
  1020. reconstruct_the_active_formatting_elements();
  1021. auto element = insert_html_element(token);
  1022. m_list_of_active_formatting_elements.add(*element);
  1023. return;
  1024. }
  1025. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1026. reconstruct_the_active_formatting_elements();
  1027. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1028. PARSE_ERROR();
  1029. run_the_adoption_agency_algorithm(token);
  1030. reconstruct_the_active_formatting_elements();
  1031. }
  1032. auto element = insert_html_element(token);
  1033. m_list_of_active_formatting_elements.add(*element);
  1034. return;
  1035. }
  1036. 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)) {
  1037. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1038. goto AnyOtherEndTag;
  1039. return;
  1040. }
  1041. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1042. reconstruct_the_active_formatting_elements();
  1043. insert_html_element(token);
  1044. m_list_of_active_formatting_elements.add_marker();
  1045. m_frameset_ok = false;
  1046. return;
  1047. }
  1048. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1049. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1050. PARSE_ERROR();
  1051. return;
  1052. }
  1053. generate_implied_end_tags();
  1054. if (current_node().tag_name() != token.tag_name()) {
  1055. PARSE_ERROR();
  1056. }
  1057. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1058. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1059. return;
  1060. }
  1061. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1062. if (!document().in_quirks_mode()) {
  1063. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1064. close_a_p_element();
  1065. }
  1066. insert_html_element(token);
  1067. m_frameset_ok = false;
  1068. m_insertion_mode = InsertionMode::InTable;
  1069. return;
  1070. }
  1071. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1072. token.drop_attributes();
  1073. goto BRStartTag;
  1074. }
  1075. 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)) {
  1076. BRStartTag:
  1077. reconstruct_the_active_formatting_elements();
  1078. insert_html_element(token);
  1079. m_stack_of_open_elements.pop();
  1080. token.acknowledge_self_closing_flag_if_set();
  1081. m_frameset_ok = false;
  1082. return;
  1083. }
  1084. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1085. reconstruct_the_active_formatting_elements();
  1086. insert_html_element(token);
  1087. m_stack_of_open_elements.pop();
  1088. token.acknowledge_self_closing_flag_if_set();
  1089. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1090. if (type_attribute.is_null() || type_attribute != "hidden") {
  1091. m_frameset_ok = false;
  1092. }
  1093. return;
  1094. }
  1095. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1096. insert_html_element(token);
  1097. m_stack_of_open_elements.pop();
  1098. token.acknowledge_self_closing_flag_if_set();
  1099. return;
  1100. }
  1101. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1102. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1103. close_a_p_element();
  1104. insert_html_element(token);
  1105. m_stack_of_open_elements.pop();
  1106. token.acknowledge_self_closing_flag_if_set();
  1107. m_frameset_ok = false;
  1108. return;
  1109. }
  1110. if (token.is_start_tag() && token.tag_name().equals_ignoring_case("image")) {
  1111. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1112. PARSE_ERROR();
  1113. token.m_tag.tag_name.clear();
  1114. token.m_tag.tag_name.append(HTML::TagNames::img);
  1115. process_using_the_rules_for(m_insertion_mode, token);
  1116. return;
  1117. }
  1118. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1119. insert_html_element(token);
  1120. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1121. // If the next token is a U+000A LINE FEED (LF) character token,
  1122. // then ignore that token and move on to the next one.
  1123. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1124. auto next_token = m_tokenizer.next_token();
  1125. m_original_insertion_mode = m_insertion_mode;
  1126. m_frameset_ok = false;
  1127. m_insertion_mode = InsertionMode::Text;
  1128. if (next_token.has_value() && next_token.value().is_character() && next_token.value().codepoint() == '\n') {
  1129. // Ignore it.
  1130. } else {
  1131. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1132. }
  1133. return;
  1134. }
  1135. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1136. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1137. close_a_p_element();
  1138. }
  1139. reconstruct_the_active_formatting_elements();
  1140. m_frameset_ok = false;
  1141. parse_generic_raw_text_element(token);
  1142. return;
  1143. }
  1144. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1145. m_frameset_ok = false;
  1146. parse_generic_raw_text_element(token);
  1147. return;
  1148. }
  1149. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1150. parse_generic_raw_text_element(token);
  1151. return;
  1152. }
  1153. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1154. reconstruct_the_active_formatting_elements();
  1155. insert_html_element(token);
  1156. m_frameset_ok = false;
  1157. switch (m_insertion_mode) {
  1158. case InsertionMode::InTable:
  1159. case InsertionMode::InCaption:
  1160. case InsertionMode::InTableBody:
  1161. case InsertionMode::InRow:
  1162. case InsertionMode::InCell:
  1163. m_insertion_mode = InsertionMode::InSelectInTable;
  1164. break;
  1165. default:
  1166. m_insertion_mode = InsertionMode::InSelect;
  1167. break;
  1168. }
  1169. return;
  1170. }
  1171. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1172. if (current_node().tag_name() == HTML::TagNames::option)
  1173. m_stack_of_open_elements.pop();
  1174. reconstruct_the_active_formatting_elements();
  1175. insert_html_element(token);
  1176. return;
  1177. }
  1178. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1179. TODO();
  1180. }
  1181. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1182. TODO();
  1183. }
  1184. if (token.is_start_tag() && token.tag_name() == "math") {
  1185. dbg() << "<math> element encountered.";
  1186. reconstruct_the_active_formatting_elements();
  1187. insert_html_element(token);
  1188. return;
  1189. }
  1190. if (token.is_start_tag() && token.tag_name() == "svg") {
  1191. dbg() << "<svg> element encountered.";
  1192. reconstruct_the_active_formatting_elements();
  1193. insert_html_element(token);
  1194. return;
  1195. }
  1196. 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))) {
  1197. PARSE_ERROR();
  1198. return;
  1199. }
  1200. // Any other start tag
  1201. if (token.is_start_tag()) {
  1202. reconstruct_the_active_formatting_elements();
  1203. insert_html_element(token);
  1204. return;
  1205. }
  1206. if (token.is_end_tag()) {
  1207. AnyOtherEndTag:
  1208. RefPtr<Element> node;
  1209. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1210. node = m_stack_of_open_elements.elements()[i];
  1211. if (node->tag_name() == token.tag_name()) {
  1212. generate_implied_end_tags(token.tag_name());
  1213. if (node != current_node()) {
  1214. PARSE_ERROR();
  1215. }
  1216. while (&current_node() != node) {
  1217. m_stack_of_open_elements.pop();
  1218. }
  1219. m_stack_of_open_elements.pop();
  1220. break;
  1221. }
  1222. if (is_special_tag(node->tag_name())) {
  1223. PARSE_ERROR();
  1224. return;
  1225. }
  1226. }
  1227. return;
  1228. }
  1229. TODO();
  1230. }
  1231. void HTMLDocumentParser::increment_script_nesting_level()
  1232. {
  1233. ++m_script_nesting_level;
  1234. }
  1235. void HTMLDocumentParser::decrement_script_nesting_level()
  1236. {
  1237. ASSERT(m_script_nesting_level);
  1238. --m_script_nesting_level;
  1239. }
  1240. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1241. {
  1242. if (token.is_character()) {
  1243. insert_character(token.codepoint());
  1244. return;
  1245. }
  1246. if (token.is_end_of_file()) {
  1247. PARSE_ERROR();
  1248. if (current_node().tag_name() == HTML::TagNames::script)
  1249. to<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1250. m_stack_of_open_elements.pop();
  1251. m_insertion_mode = m_original_insertion_mode;
  1252. process_using_the_rules_for(m_insertion_mode, token);
  1253. return;
  1254. }
  1255. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1256. NonnullRefPtr<HTMLScriptElement> script = to<HTMLScriptElement>(current_node());
  1257. m_stack_of_open_elements.pop();
  1258. m_insertion_mode = m_original_insertion_mode;
  1259. // FIXME: Handle tokenizer insertion point stuff here.
  1260. increment_script_nesting_level();
  1261. script->prepare_script({});
  1262. decrement_script_nesting_level();
  1263. if (script_nesting_level() == 0)
  1264. m_parser_pause_flag = false;
  1265. // FIXME: Handle tokenizer insertion point stuff here too.
  1266. while (document().pending_parsing_blocking_script()) {
  1267. if (script_nesting_level() != 0) {
  1268. m_parser_pause_flag = true;
  1269. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1270. // yielding control back to the caller. (Tokenization will resume when
  1271. // the caller returns to the "outer" tree construction stage.)
  1272. TODO();
  1273. } else {
  1274. auto the_script = document().take_pending_parsing_blocking_script({});
  1275. m_tokenizer.set_blocked(true);
  1276. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1277. // or the script's "ready to be parser-executed" flag is not set:
  1278. // spin the event loop until the parser's Document has no style sheet
  1279. // that is blocking scripts and the script's "ready to be parser-executed"
  1280. // flag is set.
  1281. ASSERT(the_script->is_ready_to_be_parser_executed());
  1282. if (m_aborted)
  1283. return;
  1284. m_tokenizer.set_blocked(false);
  1285. // FIXME: Handle tokenizer insertion point stuff here too.
  1286. ASSERT(script_nesting_level() == 0);
  1287. increment_script_nesting_level();
  1288. the_script->execute_script();
  1289. decrement_script_nesting_level();
  1290. ASSERT(script_nesting_level() == 0);
  1291. m_parser_pause_flag = false;
  1292. // FIXME: Handle tokenizer insertion point stuff here too.
  1293. }
  1294. }
  1295. return;
  1296. }
  1297. if (token.is_end_tag()) {
  1298. m_stack_of_open_elements.pop();
  1299. m_insertion_mode = m_original_insertion_mode;
  1300. return;
  1301. }
  1302. TODO();
  1303. }
  1304. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1305. {
  1306. while (!current_node().tag_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1307. m_stack_of_open_elements.pop();
  1308. }
  1309. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1310. {
  1311. while (!current_node().tag_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1312. m_stack_of_open_elements.pop();
  1313. }
  1314. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1315. {
  1316. while (!current_node().tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1317. m_stack_of_open_elements.pop();
  1318. }
  1319. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1320. {
  1321. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1322. clear_the_stack_back_to_a_table_row_context();
  1323. insert_html_element(token);
  1324. m_insertion_mode = InsertionMode::InCell;
  1325. m_list_of_active_formatting_elements.add_marker();
  1326. return;
  1327. }
  1328. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1329. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1330. PARSE_ERROR();
  1331. return;
  1332. }
  1333. clear_the_stack_back_to_a_table_row_context();
  1334. m_stack_of_open_elements.pop();
  1335. m_insertion_mode = InsertionMode::InTableBody;
  1336. return;
  1337. }
  1338. 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)) {
  1339. if (m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1340. PARSE_ERROR();
  1341. return;
  1342. }
  1343. clear_the_stack_back_to_a_table_row_context();
  1344. m_stack_of_open_elements.pop();
  1345. m_insertion_mode = InsertionMode::InTableBody;
  1346. process_using_the_rules_for(m_insertion_mode, token);
  1347. return;
  1348. }
  1349. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1350. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1351. PARSE_ERROR();
  1352. return;
  1353. }
  1354. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1355. return;
  1356. }
  1357. clear_the_stack_back_to_a_table_row_context();
  1358. m_stack_of_open_elements.pop();
  1359. m_insertion_mode = InsertionMode::InTableBody;
  1360. process_using_the_rules_for(m_insertion_mode, token);
  1361. return;
  1362. }
  1363. 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)) {
  1364. PARSE_ERROR();
  1365. return;
  1366. }
  1367. process_using_the_rules_for(InsertionMode::InTable, token);
  1368. }
  1369. void HTMLDocumentParser::close_the_cell()
  1370. {
  1371. generate_implied_end_tags();
  1372. if (!current_node().tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1373. PARSE_ERROR();
  1374. }
  1375. while (!current_node().tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1376. m_stack_of_open_elements.pop();
  1377. m_stack_of_open_elements.pop();
  1378. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1379. m_insertion_mode = InsertionMode::InRow;
  1380. }
  1381. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1382. {
  1383. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1384. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1385. PARSE_ERROR();
  1386. return;
  1387. }
  1388. generate_implied_end_tags();
  1389. if (current_node().tag_name() != token.tag_name()) {
  1390. PARSE_ERROR();
  1391. }
  1392. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1393. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1394. m_insertion_mode = InsertionMode::InRow;
  1395. return;
  1396. }
  1397. 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)) {
  1398. 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)) {
  1399. PARSE_ERROR();
  1400. return;
  1401. }
  1402. close_the_cell();
  1403. process_using_the_rules_for(m_insertion_mode, token);
  1404. return;
  1405. }
  1406. 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)) {
  1407. PARSE_ERROR();
  1408. return;
  1409. }
  1410. 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)) {
  1411. if (m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1412. PARSE_ERROR();
  1413. return;
  1414. }
  1415. close_the_cell();
  1416. // Reprocess the token.
  1417. process_using_the_rules_for(m_insertion_mode, token);
  1418. return;
  1419. }
  1420. process_using_the_rules_for(InsertionMode::InBody, token);
  1421. }
  1422. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1423. {
  1424. if (token.is_character()) {
  1425. if (token.codepoint() == 0) {
  1426. PARSE_ERROR();
  1427. return;
  1428. }
  1429. m_pending_table_character_tokens.append(token);
  1430. return;
  1431. }
  1432. for (auto& pending_token : m_pending_table_character_tokens) {
  1433. ASSERT(pending_token.is_character());
  1434. if (!pending_token.is_parser_whitespace()) {
  1435. // FIXME: If any of the tokens in the pending table character tokens list
  1436. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1437. // reprocess the character tokens in the pending table character tokens list using
  1438. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1439. TODO();
  1440. }
  1441. }
  1442. for (auto& pending_token : m_pending_table_character_tokens) {
  1443. insert_character(pending_token.codepoint());
  1444. }
  1445. m_insertion_mode = m_original_insertion_mode;
  1446. process_using_the_rules_for(m_insertion_mode, token);
  1447. }
  1448. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1449. {
  1450. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1451. clear_the_stack_back_to_a_table_body_context();
  1452. insert_html_element(token);
  1453. m_insertion_mode = InsertionMode::InRow;
  1454. return;
  1455. }
  1456. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1457. PARSE_ERROR();
  1458. clear_the_stack_back_to_a_table_body_context();
  1459. HTMLToken fake_tr_token;
  1460. fake_tr_token.m_type = HTMLToken::Type::StartTag;
  1461. fake_tr_token.m_tag.tag_name.append(HTML::TagNames::tr);
  1462. insert_html_element(fake_tr_token);
  1463. m_insertion_mode = InsertionMode::InRow;
  1464. process_using_the_rules_for(m_insertion_mode, token);
  1465. return;
  1466. }
  1467. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1468. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1469. PARSE_ERROR();
  1470. return;
  1471. }
  1472. clear_the_stack_back_to_a_table_body_context();
  1473. m_stack_of_open_elements.pop();
  1474. m_insertion_mode = InsertionMode::InTable;
  1475. return;
  1476. }
  1477. 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))
  1478. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1479. // 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.
  1480. clear_the_stack_back_to_a_table_body_context();
  1481. m_stack_of_open_elements.pop();
  1482. m_insertion_mode = InsertionMode::InTable;
  1483. process_using_the_rules_for(InsertionMode::InTable, token);
  1484. return;
  1485. }
  1486. 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)) {
  1487. PARSE_ERROR();
  1488. return;
  1489. }
  1490. process_using_the_rules_for(InsertionMode::InTable, token);
  1491. }
  1492. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1493. {
  1494. 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)) {
  1495. m_pending_table_character_tokens.clear();
  1496. m_original_insertion_mode = m_insertion_mode;
  1497. m_insertion_mode = InsertionMode::InTableText;
  1498. process_using_the_rules_for(InsertionMode::InTableText, token);
  1499. return;
  1500. }
  1501. if (token.is_comment()) {
  1502. insert_comment(token);
  1503. return;
  1504. }
  1505. if (token.is_doctype()) {
  1506. PARSE_ERROR();
  1507. return;
  1508. }
  1509. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1510. clear_the_stack_back_to_a_table_context();
  1511. m_list_of_active_formatting_elements.add_marker();
  1512. insert_html_element(token);
  1513. m_insertion_mode = InsertionMode::InCaption;
  1514. return;
  1515. }
  1516. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1517. TODO();
  1518. }
  1519. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1520. TODO();
  1521. }
  1522. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1523. clear_the_stack_back_to_a_table_context();
  1524. insert_html_element(token);
  1525. m_insertion_mode = InsertionMode::InTableBody;
  1526. return;
  1527. }
  1528. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1529. clear_the_stack_back_to_a_table_context();
  1530. HTMLToken fake_tbody_token;
  1531. fake_tbody_token.m_type = HTMLToken::Type::StartTag;
  1532. fake_tbody_token.m_tag.tag_name.append(HTML::TagNames::tbody);
  1533. insert_html_element(fake_tbody_token);
  1534. m_insertion_mode = InsertionMode::InTableBody;
  1535. process_using_the_rules_for(InsertionMode::InTableBody, token);
  1536. return;
  1537. }
  1538. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1539. PARSE_ERROR();
  1540. TODO();
  1541. }
  1542. if (token.is_end_tag()) {
  1543. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1544. PARSE_ERROR();
  1545. return;
  1546. }
  1547. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1548. reset_the_insertion_mode_appropriately();
  1549. return;
  1550. }
  1551. TODO();
  1552. }
  1553. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1554. {
  1555. (void)token;
  1556. TODO();
  1557. }
  1558. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  1559. {
  1560. if (token.is_character()) {
  1561. if (token.codepoint() == 0) {
  1562. PARSE_ERROR();
  1563. return;
  1564. }
  1565. insert_character(token.codepoint());
  1566. return;
  1567. }
  1568. if (token.is_comment()) {
  1569. insert_comment(token);
  1570. return;
  1571. }
  1572. if (token.is_doctype()) {
  1573. PARSE_ERROR();
  1574. return;
  1575. }
  1576. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  1577. process_using_the_rules_for(InsertionMode::InBody, token);
  1578. return;
  1579. }
  1580. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  1581. if (current_node().tag_name() == HTML::TagNames::option) {
  1582. m_stack_of_open_elements.pop();
  1583. }
  1584. insert_html_element(token);
  1585. return;
  1586. }
  1587. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1588. if (current_node().tag_name() == HTML::TagNames::option) {
  1589. m_stack_of_open_elements.pop();
  1590. }
  1591. if (current_node().tag_name() == HTML::TagNames::optgroup) {
  1592. m_stack_of_open_elements.pop();
  1593. }
  1594. insert_html_element(token);
  1595. return;
  1596. }
  1597. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1598. if (current_node().tag_name() == HTML::TagNames::option && node_before_current_node().tag_name() == HTML::TagNames::optgroup)
  1599. m_stack_of_open_elements.pop();
  1600. if (current_node().tag_name() == HTML::TagNames::optgroup) {
  1601. m_stack_of_open_elements.pop();
  1602. } else {
  1603. PARSE_ERROR();
  1604. return;
  1605. }
  1606. return;
  1607. }
  1608. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  1609. if (current_node().tag_name() == HTML::TagNames::option) {
  1610. m_stack_of_open_elements.pop();
  1611. } else {
  1612. PARSE_ERROR();
  1613. return;
  1614. }
  1615. return;
  1616. }
  1617. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  1618. if (m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  1619. PARSE_ERROR();
  1620. return;
  1621. }
  1622. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1623. reset_the_insertion_mode_appropriately();
  1624. return;
  1625. }
  1626. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1627. PARSE_ERROR();
  1628. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select))
  1629. return;
  1630. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1631. reset_the_insertion_mode_appropriately();
  1632. return;
  1633. }
  1634. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  1635. PARSE_ERROR();
  1636. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  1637. return;
  1638. }
  1639. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1640. reset_the_insertion_mode_appropriately();
  1641. process_using_the_rules_for(m_insertion_mode, token);
  1642. return;
  1643. }
  1644. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  1645. process_using_the_rules_for(InsertionMode::InHead, token);
  1646. return;
  1647. }
  1648. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  1649. process_using_the_rules_for(InsertionMode::InHead, token);
  1650. return;
  1651. }
  1652. if (token.is_end_of_file()) {
  1653. process_using_the_rules_for(InsertionMode::InBody, token);
  1654. return;
  1655. }
  1656. PARSE_ERROR();
  1657. }
  1658. void HTMLDocumentParser::handle_in_caption(HTMLToken& token)
  1659. {
  1660. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::caption) {
  1661. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  1662. PARSE_ERROR();
  1663. return;
  1664. }
  1665. generate_implied_end_tags();
  1666. if (current_node().tag_name() != HTML::TagNames::caption)
  1667. PARSE_ERROR();
  1668. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  1669. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1670. m_insertion_mode = InsertionMode::InTable;
  1671. return;
  1672. }
  1673. 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))
  1674. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1675. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  1676. PARSE_ERROR();
  1677. return;
  1678. }
  1679. generate_implied_end_tags();
  1680. if (current_node().tag_name() != HTML::TagNames::caption)
  1681. PARSE_ERROR();
  1682. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  1683. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1684. m_insertion_mode = InsertionMode::InTable;
  1685. process_using_the_rules_for(m_insertion_mode, token);
  1686. return;
  1687. }
  1688. 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)) {
  1689. PARSE_ERROR();
  1690. return;
  1691. }
  1692. process_using_the_rules_for(InsertionMode::InBody, token);
  1693. }
  1694. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  1695. {
  1696. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1697. RefPtr<Element> node = m_stack_of_open_elements.elements().at(i);
  1698. if (node->tag_name() == HTML::TagNames::select) {
  1699. TODO();
  1700. }
  1701. if (node->tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1702. m_insertion_mode = InsertionMode::InCell;
  1703. return;
  1704. }
  1705. if (node->tag_name() == HTML::TagNames::tr) {
  1706. m_insertion_mode = InsertionMode::InRow;
  1707. return;
  1708. }
  1709. if (node->tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  1710. m_insertion_mode = InsertionMode::InTableBody;
  1711. return;
  1712. }
  1713. if (node->tag_name() == HTML::TagNames::caption) {
  1714. m_insertion_mode = InsertionMode::InCaption;
  1715. return;
  1716. }
  1717. if (node->tag_name() == HTML::TagNames::colgroup) {
  1718. m_insertion_mode = InsertionMode::InColumnGroup;
  1719. return;
  1720. }
  1721. if (node->tag_name() == HTML::TagNames::table) {
  1722. m_insertion_mode = InsertionMode::InTable;
  1723. return;
  1724. }
  1725. if (node->tag_name() == HTML::TagNames::template_) {
  1726. TODO();
  1727. }
  1728. if (node->tag_name() == HTML::TagNames::body) {
  1729. m_insertion_mode = InsertionMode::InBody;
  1730. return;
  1731. }
  1732. if (node->tag_name() == HTML::TagNames::frameset) {
  1733. m_insertion_mode = InsertionMode::InFrameset;
  1734. if (m_parsing_fragment) {
  1735. TODO();
  1736. }
  1737. return;
  1738. }
  1739. if (node->tag_name() == HTML::TagNames::html) {
  1740. TODO();
  1741. }
  1742. }
  1743. m_insertion_mode = InsertionMode::InBody;
  1744. if (m_parsing_fragment) {
  1745. TODO();
  1746. }
  1747. }
  1748. const char* HTMLDocumentParser::insertion_mode_name() const
  1749. {
  1750. switch (m_insertion_mode) {
  1751. #define __ENUMERATE_INSERTION_MODE(mode) \
  1752. case InsertionMode::mode: \
  1753. return #mode;
  1754. ENUMERATE_INSERTION_MODES
  1755. #undef __ENUMERATE_INSERTION_MODE
  1756. }
  1757. ASSERT_NOT_REACHED();
  1758. }
  1759. Document& HTMLDocumentParser::document()
  1760. {
  1761. return *m_document;
  1762. }
  1763. }