HTMLDocumentParser.cpp 71 KB

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677787980818283848586878889909192939495969798991001011021031041051061071081091101111121131141151161171181191201211221231241251261271281291301311321331341351361371381391401411421431441451461471481491501511521531541551561571581591601611621631641651661671681691701711721731741751761771781791801811821831841851861871881891901911921931941951961971981992002012022032042052062072082092102112122132142152162172182192202212222232242252262272282292302312322332342352362372382392402412422432442452462472482492502512522532542552562572582592602612622632642652662672682692702712722732742752762772782792802812822832842852862872882892902912922932942952962972982993003013023033043053063073083093103113123133143153163173183193203213223233243253263273283293303313323333343353363373383393403413423433443453463473483493503513523533543553563573583593603613623633643653663673683693703713723733743753763773783793803813823833843853863873883893903913923933943953963973983994004014024034044054064074084094104114124134144154164174184194204214224234244254264274284294304314324334344354364374384394404414424434444454464474484494504514524534544554564574584594604614624634644654664674684694704714724734744754764774784794804814824834844854864874884894904914924934944954964974984995005015025035045055065075085095105115125135145155165175185195205215225235245255265275285295305315325335345355365375385395405415425435445455465475485495505515525535545555565575585595605615625635645655665675685695705715725735745755765775785795805815825835845855865875885895905915925935945955965975985996006016026036046056066076086096106116126136146156166176186196206216226236246256266276286296306316326336346356366376386396406416426436446456466476486496506516526536546556566576586596606616626636646656666676686696706716726736746756766776786796806816826836846856866876886896906916926936946956966976986997007017027037047057067077087097107117127137147157167177187197207217227237247257267277287297307317327337347357367377387397407417427437447457467477487497507517527537547557567577587597607617627637647657667677687697707717727737747757767777787797807817827837847857867877887897907917927937947957967977987998008018028038048058068078088098108118128138148158168178188198208218228238248258268278288298308318328338348358368378388398408418428438448458468478488498508518528538548558568578588598608618628638648658668678688698708718728738748758768778788798808818828838848858868878888898908918928938948958968978988999009019029039049059069079089099109119129139149159169179189199209219229239249259269279289299309319329339349359369379389399409419429439449459469479489499509519529539549559569579589599609619629639649659669679689699709719729739749759769779789799809819829839849859869879889899909919929939949959969979989991000100110021003100410051006100710081009101010111012101310141015101610171018101910201021102210231024102510261027102810291030103110321033103410351036103710381039104010411042104310441045104610471048104910501051105210531054105510561057105810591060106110621063106410651066106710681069107010711072107310741075107610771078107910801081108210831084108510861087108810891090109110921093109410951096109710981099110011011102110311041105110611071108110911101111111211131114111511161117111811191120112111221123112411251126112711281129113011311132113311341135113611371138113911401141114211431144114511461147114811491150115111521153115411551156115711581159116011611162116311641165116611671168116911701171117211731174117511761177117811791180118111821183118411851186118711881189119011911192119311941195119611971198119912001201120212031204120512061207120812091210121112121213121412151216121712181219122012211222122312241225122612271228122912301231123212331234123512361237123812391240124112421243124412451246124712481249125012511252125312541255125612571258125912601261126212631264126512661267126812691270127112721273127412751276127712781279128012811282128312841285128612871288128912901291129212931294129512961297129812991300130113021303130413051306130713081309131013111312131313141315131613171318131913201321132213231324132513261327132813291330133113321333133413351336133713381339134013411342134313441345134613471348134913501351135213531354135513561357135813591360136113621363136413651366136713681369137013711372137313741375137613771378137913801381138213831384138513861387138813891390139113921393139413951396139713981399140014011402140314041405140614071408140914101411141214131414141514161417141814191420142114221423142414251426142714281429143014311432143314341435143614371438143914401441144214431444144514461447144814491450145114521453145414551456145714581459146014611462146314641465146614671468146914701471147214731474147514761477147814791480148114821483148414851486148714881489149014911492149314941495149614971498149915001501150215031504150515061507150815091510151115121513151415151516151715181519152015211522152315241525152615271528152915301531153215331534153515361537153815391540154115421543154415451546154715481549155015511552155315541555155615571558155915601561156215631564156515661567156815691570157115721573157415751576157715781579158015811582158315841585158615871588158915901591159215931594159515961597159815991600160116021603160416051606160716081609161016111612161316141615161616171618161916201621162216231624162516261627162816291630163116321633163416351636163716381639164016411642164316441645164616471648164916501651165216531654165516561657165816591660166116621663166416651666166716681669167016711672167316741675167616771678167916801681168216831684168516861687168816891690169116921693169416951696169716981699170017011702170317041705170617071708170917101711171217131714171517161717171817191720172117221723172417251726172717281729173017311732173317341735173617371738173917401741174217431744174517461747174817491750175117521753175417551756175717581759176017611762176317641765176617671768176917701771177217731774177517761777177817791780178117821783178417851786178717881789179017911792179317941795179617971798179918001801180218031804180518061807180818091810181118121813181418151816181718181819182018211822182318241825182618271828182918301831183218331834183518361837183818391840184118421843184418451846184718481849185018511852185318541855185618571858185918601861186218631864186518661867186818691870187118721873187418751876187718781879188018811882188318841885188618871888188918901891189218931894189518961897189818991900190119021903190419051906190719081909191019111912191319141915191619171918191919201921192219231924192519261927192819291930193119321933193419351936193719381939194019411942194319441945194619471948194919501951195219531954195519561957195819591960196119621963196419651966196719681969197019711972197319741975197619771978197919801981198219831984198519861987198819891990199119921993
  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. m_insertion_mode = InsertionMode::BeforeHTML;
  155. return;
  156. }
  157. PARSE_ERROR();
  158. document().set_quirks_mode(true);
  159. m_insertion_mode = InsertionMode::BeforeHTML;
  160. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  161. }
  162. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  163. {
  164. if (token.is_doctype()) {
  165. PARSE_ERROR();
  166. return;
  167. }
  168. if (token.is_comment()) {
  169. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  170. document().append_child(move(comment));
  171. return;
  172. }
  173. if (token.is_character() && token.is_parser_whitespace()) {
  174. return;
  175. }
  176. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  177. auto element = create_element_for(token);
  178. document().append_child(element);
  179. m_stack_of_open_elements.push(move(element));
  180. m_insertion_mode = InsertionMode::BeforeHead;
  181. return;
  182. }
  183. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  184. goto AnythingElse;
  185. }
  186. if (token.is_end_tag()) {
  187. PARSE_ERROR();
  188. return;
  189. }
  190. AnythingElse:
  191. auto element = create_element(document(), HTML::TagNames::html);
  192. document().append_child(element);
  193. m_stack_of_open_elements.push(element);
  194. // 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.
  195. m_insertion_mode = InsertionMode::BeforeHead;
  196. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  197. return;
  198. }
  199. Element& HTMLDocumentParser::current_node()
  200. {
  201. return m_stack_of_open_elements.current_node();
  202. }
  203. Element& HTMLDocumentParser::node_before_current_node()
  204. {
  205. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  206. }
  207. RefPtr<Node> HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  208. {
  209. auto& target = current_node();
  210. if (m_foster_parenting) {
  211. TODO();
  212. }
  213. return target;
  214. }
  215. NonnullRefPtr<Element> HTMLDocumentParser::create_element_for(HTMLToken& token)
  216. {
  217. auto element = create_element(document(), token.tag_name());
  218. for (auto& attribute : token.m_tag.attributes) {
  219. element->set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  220. }
  221. return element;
  222. }
  223. RefPtr<Element> HTMLDocumentParser::insert_html_element(HTMLToken& token)
  224. {
  225. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  226. auto element = create_element_for(token);
  227. // FIXME: Check if it's possible to insert `element` at `adjusted_insertion_location`
  228. adjusted_insertion_location->append_child(element);
  229. m_stack_of_open_elements.push(element);
  230. return element;
  231. }
  232. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  233. {
  234. if (token.is_character() && token.is_parser_whitespace()) {
  235. return;
  236. }
  237. if (token.is_comment()) {
  238. insert_comment(token);
  239. return;
  240. }
  241. if (token.is_doctype()) {
  242. PARSE_ERROR();
  243. return;
  244. }
  245. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  246. process_using_the_rules_for(InsertionMode::InBody, token);
  247. return;
  248. }
  249. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  250. auto element = insert_html_element(token);
  251. m_head_element = to<HTMLHeadElement>(element);
  252. m_insertion_mode = InsertionMode::InHead;
  253. return;
  254. }
  255. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  256. goto AnythingElse;
  257. }
  258. if (token.is_end_tag()) {
  259. PARSE_ERROR();
  260. return;
  261. }
  262. AnythingElse:
  263. HTMLToken fake_head_token;
  264. fake_head_token.m_type = HTMLToken::Type::StartTag;
  265. fake_head_token.m_tag.tag_name.append(HTML::TagNames::head);
  266. m_head_element = to<HTMLHeadElement>(insert_html_element(fake_head_token));
  267. m_insertion_mode = InsertionMode::InHead;
  268. process_using_the_rules_for(InsertionMode::InHead, token);
  269. return;
  270. }
  271. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  272. {
  273. auto data = token.m_comment_or_character.data.to_string();
  274. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  275. adjusted_insertion_location->append_child(adopt(*new Comment(document(), data)));
  276. }
  277. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  278. {
  279. if (token.is_parser_whitespace()) {
  280. insert_character(token.codepoint());
  281. return;
  282. }
  283. if (token.is_comment()) {
  284. insert_comment(token);
  285. return;
  286. }
  287. if (token.is_doctype()) {
  288. PARSE_ERROR();
  289. return;
  290. }
  291. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  292. process_using_the_rules_for(InsertionMode::InBody, token);
  293. return;
  294. }
  295. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  296. insert_html_element(token);
  297. m_stack_of_open_elements.pop();
  298. token.acknowledge_self_closing_flag_if_set();
  299. return;
  300. }
  301. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  302. auto element = insert_html_element(token);
  303. m_stack_of_open_elements.pop();
  304. token.acknowledge_self_closing_flag_if_set();
  305. return;
  306. }
  307. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  308. insert_html_element(token);
  309. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  310. m_original_insertion_mode = m_insertion_mode;
  311. m_insertion_mode = InsertionMode::Text;
  312. return;
  313. }
  314. 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)) {
  315. parse_generic_raw_text_element(token);
  316. return;
  317. }
  318. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  319. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  320. auto element = create_element_for(token);
  321. auto& script_element = to<HTMLScriptElement>(*element);
  322. script_element.set_parser_document({}, document());
  323. script_element.set_non_blocking({}, false);
  324. if (m_parsing_fragment) {
  325. TODO();
  326. }
  327. if (m_invoked_via_document_write) {
  328. TODO();
  329. }
  330. adjusted_insertion_location->append_child(element, false);
  331. m_stack_of_open_elements.push(element);
  332. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  333. m_original_insertion_mode = m_insertion_mode;
  334. m_insertion_mode = InsertionMode::Text;
  335. return;
  336. }
  337. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  338. m_stack_of_open_elements.pop();
  339. m_insertion_mode = InsertionMode::AfterHead;
  340. return;
  341. }
  342. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  343. TODO();
  344. }
  345. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  346. // FIXME: Support this properly
  347. insert_html_element(token);
  348. return;
  349. }
  350. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  351. // FIXME: Support this properly
  352. ASSERT(current_node().tag_name() == HTML::TagNames::template_);
  353. m_stack_of_open_elements.pop();
  354. return;
  355. }
  356. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  357. PARSE_ERROR();
  358. return;
  359. }
  360. m_stack_of_open_elements.pop();
  361. m_insertion_mode = InsertionMode::AfterHead;
  362. process_using_the_rules_for(m_insertion_mode, token);
  363. }
  364. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken&)
  365. {
  366. TODO();
  367. }
  368. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  369. {
  370. insert_html_element(token);
  371. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  372. m_original_insertion_mode = m_insertion_mode;
  373. m_insertion_mode = InsertionMode::Text;
  374. }
  375. Text* HTMLDocumentParser::find_character_insertion_node()
  376. {
  377. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  378. if (adjusted_insertion_location->is_document())
  379. return nullptr;
  380. if (adjusted_insertion_location->last_child() && adjusted_insertion_location->last_child()->is_text())
  381. return to<Text>(adjusted_insertion_location->last_child());
  382. auto new_text_node = adopt(*new Text(document(), ""));
  383. adjusted_insertion_location->append_child(new_text_node);
  384. return new_text_node;
  385. }
  386. void HTMLDocumentParser::flush_character_insertions()
  387. {
  388. if (m_character_insertion_builder.is_empty())
  389. return;
  390. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  391. m_character_insertion_node->parent()->children_changed();
  392. m_character_insertion_builder.clear();
  393. }
  394. void HTMLDocumentParser::insert_character(u32 data)
  395. {
  396. auto node = find_character_insertion_node();
  397. if (node == m_character_insertion_node) {
  398. m_character_insertion_builder.append(Utf32View { &data, 1 });
  399. return;
  400. }
  401. if (!m_character_insertion_node) {
  402. m_character_insertion_node = node;
  403. m_character_insertion_builder.append(Utf32View { &data, 1 });
  404. return;
  405. }
  406. flush_character_insertions();
  407. m_character_insertion_node = node;
  408. m_character_insertion_builder.append(Utf32View { &data, 1 });
  409. }
  410. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  411. {
  412. if (token.is_character() && token.is_parser_whitespace()) {
  413. insert_character(token.codepoint());
  414. return;
  415. }
  416. if (token.is_comment()) {
  417. insert_comment(token);
  418. return;
  419. }
  420. if (token.is_doctype()) {
  421. PARSE_ERROR();
  422. return;
  423. }
  424. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  425. process_using_the_rules_for(InsertionMode::InBody, token);
  426. return;
  427. }
  428. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  429. insert_html_element(token);
  430. m_frameset_ok = false;
  431. m_insertion_mode = InsertionMode::InBody;
  432. return;
  433. }
  434. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  435. insert_html_element(token);
  436. m_insertion_mode = InsertionMode::InFrameset;
  437. return;
  438. }
  439. 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)) {
  440. PARSE_ERROR();
  441. m_stack_of_open_elements.push(*m_head_element);
  442. process_using_the_rules_for(InsertionMode::InHead, token);
  443. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  444. return entry.ptr() == m_head_element;
  445. });
  446. return;
  447. }
  448. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  449. TODO();
  450. }
  451. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  452. goto AnythingElse;
  453. }
  454. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  455. PARSE_ERROR();
  456. return;
  457. }
  458. AnythingElse:
  459. HTMLToken fake_body_token;
  460. fake_body_token.m_type = HTMLToken::Type::StartTag;
  461. fake_body_token.m_tag.tag_name.append(HTML::TagNames::body);
  462. insert_html_element(fake_body_token);
  463. m_insertion_mode = InsertionMode::InBody;
  464. process_using_the_rules_for(m_insertion_mode, token);
  465. }
  466. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  467. {
  468. 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))
  469. m_stack_of_open_elements.pop();
  470. }
  471. void HTMLDocumentParser::close_a_p_element()
  472. {
  473. generate_implied_end_tags(HTML::TagNames::p);
  474. if (current_node().tag_name() != HTML::TagNames::p) {
  475. PARSE_ERROR();
  476. }
  477. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  478. }
  479. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  480. {
  481. if (token.is_character() && token.is_parser_whitespace()) {
  482. process_using_the_rules_for(InsertionMode::InBody, token);
  483. return;
  484. }
  485. if (token.is_comment()) {
  486. TODO();
  487. }
  488. if (token.is_doctype()) {
  489. PARSE_ERROR();
  490. return;
  491. }
  492. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  493. process_using_the_rules_for(InsertionMode::InBody, token);
  494. return;
  495. }
  496. if (token.is_end_of_file()) {
  497. stop_parsing();
  498. return;
  499. }
  500. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  501. if (m_parsing_fragment) {
  502. TODO();
  503. }
  504. m_insertion_mode = InsertionMode::AfterAfterBody;
  505. return;
  506. }
  507. PARSE_ERROR();
  508. m_insertion_mode = InsertionMode::InBody;
  509. process_using_the_rules_for(InsertionMode::InBody, token);
  510. }
  511. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  512. {
  513. if (token.is_comment()) {
  514. auto comment = adopt(*new Comment(document(), token.m_comment_or_character.data.to_string()));
  515. document().append_child(move(comment));
  516. return;
  517. }
  518. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  519. process_using_the_rules_for(InsertionMode::InBody, token);
  520. return;
  521. }
  522. if (token.is_end_of_file()) {
  523. stop_parsing();
  524. return;
  525. }
  526. PARSE_ERROR();
  527. m_insertion_mode = InsertionMode::InBody;
  528. process_using_the_rules_for(m_insertion_mode, token);
  529. }
  530. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  531. {
  532. // FIXME: This needs to care about "markers"
  533. if (m_list_of_active_formatting_elements.is_empty())
  534. return;
  535. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  536. return;
  537. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  538. return;
  539. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  540. RefPtr<Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  541. ASSERT(entry);
  542. Rewind:
  543. if (index == 0) {
  544. goto Create;
  545. }
  546. --index;
  547. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  548. ASSERT(entry);
  549. if (!m_stack_of_open_elements.contains(*entry))
  550. goto Rewind;
  551. Advance:
  552. ++index;
  553. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  554. ASSERT(entry);
  555. Create:
  556. // FIXME: Hold on to the real token!
  557. HTMLToken fake_token;
  558. fake_token.m_type = HTMLToken::Type::StartTag;
  559. fake_token.m_tag.tag_name.append(entry->tag_name());
  560. auto new_element = insert_html_element(fake_token);
  561. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  562. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  563. goto Advance;
  564. }
  565. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  566. {
  567. auto subject = token.tag_name();
  568. // If the current node is an HTML element whose tag name is subject,
  569. // and the current node is not in the list of active formatting elements,
  570. // then pop the current node off the stack of open elements, and return.
  571. if (current_node().tag_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  572. m_stack_of_open_elements.pop();
  573. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  574. }
  575. size_t outer_loop_counter = 0;
  576. //OuterLoop:
  577. if (outer_loop_counter >= 8)
  578. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  579. ++outer_loop_counter;
  580. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  581. if (!formatting_element)
  582. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  583. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  584. PARSE_ERROR();
  585. // FIXME: If formatting element is not in the stack of open elements,
  586. // then this is a parse error; remove the element from the list, and return.
  587. TODO();
  588. }
  589. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  590. PARSE_ERROR();
  591. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  592. }
  593. if (formatting_element != &current_node()) {
  594. PARSE_ERROR();
  595. }
  596. RefPtr<Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  597. if (!furthest_block) {
  598. while (&current_node() != formatting_element)
  599. m_stack_of_open_elements.pop();
  600. m_stack_of_open_elements.pop();
  601. m_list_of_active_formatting_elements.remove(*formatting_element);
  602. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  603. }
  604. // FIXME: Implement the rest of the AAA :^)
  605. TODO();
  606. }
  607. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name)
  608. {
  609. return tag_name.is_one_of(
  610. HTML::TagNames::address,
  611. HTML::TagNames::applet,
  612. HTML::TagNames::area,
  613. HTML::TagNames::article,
  614. HTML::TagNames::aside,
  615. HTML::TagNames::base,
  616. HTML::TagNames::basefont,
  617. HTML::TagNames::bgsound,
  618. HTML::TagNames::blockquote,
  619. HTML::TagNames::body,
  620. HTML::TagNames::br,
  621. HTML::TagNames::button,
  622. HTML::TagNames::caption,
  623. HTML::TagNames::center,
  624. HTML::TagNames::col,
  625. HTML::TagNames::colgroup,
  626. HTML::TagNames::dd,
  627. HTML::TagNames::details,
  628. HTML::TagNames::dir,
  629. HTML::TagNames::div,
  630. HTML::TagNames::dl,
  631. HTML::TagNames::dt,
  632. HTML::TagNames::embed,
  633. HTML::TagNames::fieldset,
  634. HTML::TagNames::figcaption,
  635. HTML::TagNames::figure,
  636. HTML::TagNames::footer,
  637. HTML::TagNames::form,
  638. HTML::TagNames::frame,
  639. HTML::TagNames::frameset,
  640. HTML::TagNames::h1,
  641. HTML::TagNames::h2,
  642. HTML::TagNames::h3,
  643. HTML::TagNames::h4,
  644. HTML::TagNames::h5,
  645. HTML::TagNames::h6,
  646. HTML::TagNames::head,
  647. HTML::TagNames::header,
  648. HTML::TagNames::hgroup,
  649. HTML::TagNames::hr,
  650. HTML::TagNames::html,
  651. HTML::TagNames::iframe,
  652. HTML::TagNames::img,
  653. HTML::TagNames::input,
  654. HTML::TagNames::keygen,
  655. HTML::TagNames::li,
  656. HTML::TagNames::link,
  657. HTML::TagNames::listing,
  658. HTML::TagNames::main,
  659. HTML::TagNames::marquee,
  660. HTML::TagNames::menu,
  661. HTML::TagNames::meta,
  662. HTML::TagNames::nav,
  663. HTML::TagNames::noembed,
  664. HTML::TagNames::noframes,
  665. HTML::TagNames::noscript,
  666. HTML::TagNames::object,
  667. HTML::TagNames::ol,
  668. HTML::TagNames::p,
  669. HTML::TagNames::param,
  670. HTML::TagNames::plaintext,
  671. HTML::TagNames::pre,
  672. HTML::TagNames::script,
  673. HTML::TagNames::section,
  674. HTML::TagNames::select,
  675. HTML::TagNames::source,
  676. HTML::TagNames::style,
  677. HTML::TagNames::summary,
  678. HTML::TagNames::table,
  679. HTML::TagNames::tbody,
  680. HTML::TagNames::td,
  681. HTML::TagNames::template_,
  682. HTML::TagNames::textarea,
  683. HTML::TagNames::tfoot,
  684. HTML::TagNames::th,
  685. HTML::TagNames::thead,
  686. HTML::TagNames::title,
  687. HTML::TagNames::tr,
  688. HTML::TagNames::track,
  689. HTML::TagNames::ul,
  690. HTML::TagNames::wbr,
  691. HTML::TagNames::xmp);
  692. }
  693. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  694. {
  695. if (token.is_character()) {
  696. if (token.codepoint() == 0) {
  697. PARSE_ERROR();
  698. return;
  699. }
  700. if (token.is_parser_whitespace()) {
  701. reconstruct_the_active_formatting_elements();
  702. insert_character(token.codepoint());
  703. return;
  704. }
  705. reconstruct_the_active_formatting_elements();
  706. insert_character(token.codepoint());
  707. m_frameset_ok = false;
  708. return;
  709. }
  710. if (token.is_comment()) {
  711. insert_comment(token);
  712. return;
  713. }
  714. if (token.is_doctype()) {
  715. PARSE_ERROR();
  716. return;
  717. }
  718. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  719. PARSE_ERROR();
  720. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  721. return;
  722. for (auto& attribute : token.m_tag.attributes) {
  723. if (current_node().has_attribute(attribute.name_builder.string_view()))
  724. continue;
  725. current_node().set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  726. }
  727. return;
  728. }
  729. 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)) {
  730. process_using_the_rules_for(InsertionMode::InHead, token);
  731. return;
  732. }
  733. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  734. process_using_the_rules_for(InsertionMode::InHead, token);
  735. return;
  736. }
  737. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  738. PARSE_ERROR();
  739. if (m_stack_of_open_elements.elements().size() == 1
  740. || node_before_current_node().tag_name() != HTML::TagNames::body
  741. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  742. return;
  743. }
  744. m_frameset_ok = false;
  745. for (auto& attribute : token.m_tag.attributes) {
  746. if (node_before_current_node().has_attribute(attribute.name_builder.string_view()))
  747. continue;
  748. node_before_current_node().set_attribute(attribute.name_builder.to_string(), attribute.value_builder.to_string());
  749. }
  750. return;
  751. }
  752. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  753. TODO();
  754. }
  755. if (token.is_end_of_file()) {
  756. // FIXME: If the stack of template insertion modes is not empty,
  757. // then process the token using the rules for the "in template" insertion mode.
  758. // FIXME: If there is a node in the stack of open elements that is not either
  759. // a dd element, a dt element, an li element, an optgroup element, an option element,
  760. // a p element, an rb element, an rp element, an rt element, an rtc element,
  761. // a tbody element, a td element, a tfoot element, a th element, a thead element,
  762. // a tr element, the body element, or the html element, then this is a parse error.
  763. stop_parsing();
  764. return;
  765. }
  766. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  767. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  768. PARSE_ERROR();
  769. return;
  770. }
  771. for (auto& node : m_stack_of_open_elements.elements()) {
  772. 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)) {
  773. PARSE_ERROR();
  774. break;
  775. }
  776. }
  777. m_insertion_mode = InsertionMode::AfterBody;
  778. return;
  779. }
  780. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  781. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  782. PARSE_ERROR();
  783. return;
  784. }
  785. for (auto& node : m_stack_of_open_elements.elements()) {
  786. 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)) {
  787. PARSE_ERROR();
  788. break;
  789. }
  790. }
  791. m_insertion_mode = InsertionMode::AfterBody;
  792. process_using_the_rules_for(m_insertion_mode, token);
  793. return;
  794. }
  795. 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)) {
  796. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  797. close_a_p_element();
  798. insert_html_element(token);
  799. return;
  800. }
  801. 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)) {
  802. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  803. close_a_p_element();
  804. 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)) {
  805. PARSE_ERROR();
  806. m_stack_of_open_elements.pop();
  807. }
  808. insert_html_element(token);
  809. return;
  810. }
  811. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  812. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  813. close_a_p_element();
  814. insert_html_element(token);
  815. m_frameset_ok = false;
  816. // If the next token is a U+000A LINE FEED (LF) character token,
  817. // then ignore that token and move on to the next one.
  818. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  819. auto next_token = m_tokenizer.next_token();
  820. if (next_token.has_value() && next_token.value().is_character() && next_token.value().codepoint() == '\n') {
  821. // Ignore it.
  822. } else {
  823. process_using_the_rules_for(m_insertion_mode, next_token.value());
  824. }
  825. return;
  826. }
  827. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  828. if (m_form_element && m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  829. PARSE_ERROR();
  830. return;
  831. }
  832. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  833. close_a_p_element();
  834. auto element = insert_html_element(token);
  835. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  836. m_form_element = to<HTMLFormElement>(*element);
  837. return;
  838. }
  839. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  840. m_frameset_ok = false;
  841. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  842. RefPtr<Element> node = m_stack_of_open_elements.elements()[i];
  843. if (node->tag_name() == HTML::TagNames::li) {
  844. generate_implied_end_tags(HTML::TagNames::li);
  845. if (current_node().tag_name() != HTML::TagNames::li) {
  846. PARSE_ERROR();
  847. }
  848. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  849. break;
  850. }
  851. if (is_special_tag(node->tag_name()) && !node->tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  852. break;
  853. }
  854. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  855. close_a_p_element();
  856. insert_html_element(token);
  857. return;
  858. }
  859. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  860. m_frameset_ok = false;
  861. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  862. RefPtr<Element> node = m_stack_of_open_elements.elements()[i];
  863. if (node->tag_name() == HTML::TagNames::dd) {
  864. generate_implied_end_tags(HTML::TagNames::dd);
  865. if (current_node().tag_name() != HTML::TagNames::dd) {
  866. PARSE_ERROR();
  867. }
  868. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  869. break;
  870. }
  871. if (node->tag_name() == HTML::TagNames::dt) {
  872. generate_implied_end_tags(HTML::TagNames::dt);
  873. if (current_node().tag_name() != HTML::TagNames::dt) {
  874. PARSE_ERROR();
  875. }
  876. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  877. break;
  878. }
  879. if (is_special_tag(node->tag_name()) && !node->tag_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  880. break;
  881. }
  882. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  883. close_a_p_element();
  884. insert_html_element(token);
  885. return;
  886. }
  887. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  888. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  889. close_a_p_element();
  890. insert_html_element(token);
  891. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  892. return;
  893. }
  894. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  895. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  896. PARSE_ERROR();
  897. generate_implied_end_tags();
  898. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  899. }
  900. reconstruct_the_active_formatting_elements();
  901. insert_html_element(token);
  902. m_frameset_ok = false;
  903. return;
  904. }
  905. 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)) {
  906. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  907. PARSE_ERROR();
  908. return;
  909. }
  910. generate_implied_end_tags();
  911. if (current_node().tag_name() != token.tag_name()) {
  912. PARSE_ERROR();
  913. }
  914. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  915. return;
  916. }
  917. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  918. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  919. auto node = m_form_element;
  920. m_form_element = nullptr;
  921. if (!node || m_stack_of_open_elements.has_in_scope(*node)) {
  922. PARSE_ERROR();
  923. return;
  924. }
  925. generate_implied_end_tags();
  926. if (&current_node() != node) {
  927. PARSE_ERROR();
  928. }
  929. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  930. } else {
  931. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  932. PARSE_ERROR();
  933. return;
  934. }
  935. generate_implied_end_tags();
  936. if (current_node().tag_name() != HTML::TagNames::form) {
  937. PARSE_ERROR();
  938. }
  939. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  940. }
  941. return;
  942. }
  943. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  944. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  945. PARSE_ERROR();
  946. HTMLToken fake_p_token;
  947. fake_p_token.m_type = HTMLToken::Type::StartTag;
  948. fake_p_token.m_tag.tag_name.append(HTML::TagNames::p);
  949. insert_html_element(fake_p_token);
  950. }
  951. close_a_p_element();
  952. return;
  953. }
  954. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  955. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  956. PARSE_ERROR();
  957. return;
  958. }
  959. generate_implied_end_tags(HTML::TagNames::li);
  960. if (current_node().tag_name() != HTML::TagNames::li) {
  961. PARSE_ERROR();
  962. dbg() << "Expected <li> current node, but had <" << current_node().tag_name() << ">";
  963. }
  964. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  965. return;
  966. }
  967. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  968. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  969. PARSE_ERROR();
  970. return;
  971. }
  972. generate_implied_end_tags(token.tag_name());
  973. if (current_node().tag_name() != token.tag_name()) {
  974. PARSE_ERROR();
  975. }
  976. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  977. return;
  978. }
  979. 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)) {
  980. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  981. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  982. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  983. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  984. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  985. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  986. PARSE_ERROR();
  987. return;
  988. }
  989. generate_implied_end_tags();
  990. if (current_node().tag_name() != token.tag_name()) {
  991. PARSE_ERROR();
  992. }
  993. for (;;) {
  994. auto popped_element = m_stack_of_open_elements.pop();
  995. 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))
  996. break;
  997. }
  998. return;
  999. }
  1000. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1001. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1002. PARSE_ERROR();
  1003. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1004. goto AnyOtherEndTag;
  1005. m_list_of_active_formatting_elements.remove(*element);
  1006. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1007. return entry.ptr() == element;
  1008. });
  1009. }
  1010. reconstruct_the_active_formatting_elements();
  1011. auto element = insert_html_element(token);
  1012. m_list_of_active_formatting_elements.add(*element);
  1013. return;
  1014. }
  1015. 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)) {
  1016. reconstruct_the_active_formatting_elements();
  1017. auto element = insert_html_element(token);
  1018. m_list_of_active_formatting_elements.add(*element);
  1019. return;
  1020. }
  1021. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1022. reconstruct_the_active_formatting_elements();
  1023. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1024. PARSE_ERROR();
  1025. run_the_adoption_agency_algorithm(token);
  1026. reconstruct_the_active_formatting_elements();
  1027. }
  1028. auto element = insert_html_element(token);
  1029. m_list_of_active_formatting_elements.add(*element);
  1030. return;
  1031. }
  1032. 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)) {
  1033. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1034. goto AnyOtherEndTag;
  1035. return;
  1036. }
  1037. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1038. reconstruct_the_active_formatting_elements();
  1039. insert_html_element(token);
  1040. m_list_of_active_formatting_elements.add_marker();
  1041. m_frameset_ok = false;
  1042. return;
  1043. }
  1044. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1045. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1046. PARSE_ERROR();
  1047. return;
  1048. }
  1049. generate_implied_end_tags();
  1050. if (current_node().tag_name() != token.tag_name()) {
  1051. PARSE_ERROR();
  1052. }
  1053. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1054. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1055. return;
  1056. }
  1057. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1058. if (!document().in_quirks_mode()) {
  1059. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1060. close_a_p_element();
  1061. }
  1062. insert_html_element(token);
  1063. m_frameset_ok = false;
  1064. m_insertion_mode = InsertionMode::InTable;
  1065. return;
  1066. }
  1067. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1068. token.drop_attributes();
  1069. goto BRStartTag;
  1070. }
  1071. 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)) {
  1072. BRStartTag:
  1073. reconstruct_the_active_formatting_elements();
  1074. insert_html_element(token);
  1075. m_stack_of_open_elements.pop();
  1076. token.acknowledge_self_closing_flag_if_set();
  1077. m_frameset_ok = false;
  1078. return;
  1079. }
  1080. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1081. reconstruct_the_active_formatting_elements();
  1082. insert_html_element(token);
  1083. m_stack_of_open_elements.pop();
  1084. token.acknowledge_self_closing_flag_if_set();
  1085. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1086. if (type_attribute.is_null() || type_attribute != "hidden") {
  1087. m_frameset_ok = false;
  1088. }
  1089. return;
  1090. }
  1091. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1092. insert_html_element(token);
  1093. m_stack_of_open_elements.pop();
  1094. token.acknowledge_self_closing_flag_if_set();
  1095. return;
  1096. }
  1097. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1098. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1099. close_a_p_element();
  1100. insert_html_element(token);
  1101. m_stack_of_open_elements.pop();
  1102. token.acknowledge_self_closing_flag_if_set();
  1103. m_frameset_ok = false;
  1104. return;
  1105. }
  1106. if (token.is_start_tag() && token.tag_name().equals_ignoring_case("image")) {
  1107. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1108. PARSE_ERROR();
  1109. token.m_tag.tag_name.clear();
  1110. token.m_tag.tag_name.append(HTML::TagNames::img);
  1111. process_using_the_rules_for(m_insertion_mode, token);
  1112. return;
  1113. }
  1114. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1115. insert_html_element(token);
  1116. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1117. // If the next token is a U+000A LINE FEED (LF) character token,
  1118. // then ignore that token and move on to the next one.
  1119. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1120. auto next_token = m_tokenizer.next_token();
  1121. m_original_insertion_mode = m_insertion_mode;
  1122. m_frameset_ok = false;
  1123. m_insertion_mode = InsertionMode::Text;
  1124. if (next_token.has_value() && next_token.value().is_character() && next_token.value().codepoint() == '\n') {
  1125. // Ignore it.
  1126. } else {
  1127. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1128. }
  1129. return;
  1130. }
  1131. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1132. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1133. close_a_p_element();
  1134. }
  1135. reconstruct_the_active_formatting_elements();
  1136. m_frameset_ok = false;
  1137. parse_generic_raw_text_element(token);
  1138. return;
  1139. }
  1140. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1141. m_frameset_ok = false;
  1142. parse_generic_raw_text_element(token);
  1143. return;
  1144. }
  1145. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1146. parse_generic_raw_text_element(token);
  1147. return;
  1148. }
  1149. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1150. reconstruct_the_active_formatting_elements();
  1151. insert_html_element(token);
  1152. m_frameset_ok = false;
  1153. switch (m_insertion_mode) {
  1154. case InsertionMode::InTable:
  1155. case InsertionMode::InCaption:
  1156. case InsertionMode::InTableBody:
  1157. case InsertionMode::InRow:
  1158. case InsertionMode::InCell:
  1159. m_insertion_mode = InsertionMode::InSelectInTable;
  1160. break;
  1161. default:
  1162. m_insertion_mode = InsertionMode::InSelect;
  1163. break;
  1164. }
  1165. return;
  1166. }
  1167. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1168. if (current_node().tag_name() == HTML::TagNames::option)
  1169. m_stack_of_open_elements.pop();
  1170. reconstruct_the_active_formatting_elements();
  1171. insert_html_element(token);
  1172. return;
  1173. }
  1174. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1175. TODO();
  1176. }
  1177. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1178. TODO();
  1179. }
  1180. if (token.is_start_tag() && token.tag_name() == "math") {
  1181. dbg() << "<math> element encountered.";
  1182. reconstruct_the_active_formatting_elements();
  1183. insert_html_element(token);
  1184. return;
  1185. }
  1186. if (token.is_start_tag() && token.tag_name() == "svg") {
  1187. dbg() << "<svg> element encountered.";
  1188. reconstruct_the_active_formatting_elements();
  1189. insert_html_element(token);
  1190. return;
  1191. }
  1192. 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))) {
  1193. PARSE_ERROR();
  1194. return;
  1195. }
  1196. // Any other start tag
  1197. if (token.is_start_tag()) {
  1198. reconstruct_the_active_formatting_elements();
  1199. insert_html_element(token);
  1200. return;
  1201. }
  1202. if (token.is_end_tag()) {
  1203. AnyOtherEndTag:
  1204. RefPtr<Element> node;
  1205. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1206. node = m_stack_of_open_elements.elements()[i];
  1207. if (node->tag_name() == token.tag_name()) {
  1208. generate_implied_end_tags(token.tag_name());
  1209. if (node != current_node()) {
  1210. PARSE_ERROR();
  1211. }
  1212. while (&current_node() != node) {
  1213. m_stack_of_open_elements.pop();
  1214. }
  1215. m_stack_of_open_elements.pop();
  1216. break;
  1217. }
  1218. if (is_special_tag(node->tag_name())) {
  1219. PARSE_ERROR();
  1220. return;
  1221. }
  1222. }
  1223. return;
  1224. }
  1225. TODO();
  1226. }
  1227. void HTMLDocumentParser::increment_script_nesting_level()
  1228. {
  1229. ++m_script_nesting_level;
  1230. }
  1231. void HTMLDocumentParser::decrement_script_nesting_level()
  1232. {
  1233. ASSERT(m_script_nesting_level);
  1234. --m_script_nesting_level;
  1235. }
  1236. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1237. {
  1238. if (token.is_character()) {
  1239. insert_character(token.codepoint());
  1240. return;
  1241. }
  1242. if (token.is_end_of_file()) {
  1243. PARSE_ERROR();
  1244. if (current_node().tag_name() == HTML::TagNames::script)
  1245. to<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1246. m_stack_of_open_elements.pop();
  1247. m_insertion_mode = m_original_insertion_mode;
  1248. process_using_the_rules_for(m_insertion_mode, token);
  1249. return;
  1250. }
  1251. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1252. NonnullRefPtr<HTMLScriptElement> script = to<HTMLScriptElement>(current_node());
  1253. m_stack_of_open_elements.pop();
  1254. m_insertion_mode = m_original_insertion_mode;
  1255. // FIXME: Handle tokenizer insertion point stuff here.
  1256. increment_script_nesting_level();
  1257. script->prepare_script({});
  1258. decrement_script_nesting_level();
  1259. if (script_nesting_level() == 0)
  1260. m_parser_pause_flag = false;
  1261. // FIXME: Handle tokenizer insertion point stuff here too.
  1262. while (document().pending_parsing_blocking_script()) {
  1263. if (script_nesting_level() != 0) {
  1264. m_parser_pause_flag = true;
  1265. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1266. // yielding control back to the caller. (Tokenization will resume when
  1267. // the caller returns to the "outer" tree construction stage.)
  1268. TODO();
  1269. } else {
  1270. auto the_script = document().take_pending_parsing_blocking_script({});
  1271. m_tokenizer.set_blocked(true);
  1272. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1273. // or the script's "ready to be parser-executed" flag is not set:
  1274. // spin the event loop until the parser's Document has no style sheet
  1275. // that is blocking scripts and the script's "ready to be parser-executed"
  1276. // flag is set.
  1277. ASSERT(the_script->is_ready_to_be_parser_executed());
  1278. if (m_aborted)
  1279. return;
  1280. m_tokenizer.set_blocked(false);
  1281. // FIXME: Handle tokenizer insertion point stuff here too.
  1282. ASSERT(script_nesting_level() == 0);
  1283. increment_script_nesting_level();
  1284. the_script->execute_script();
  1285. decrement_script_nesting_level();
  1286. ASSERT(script_nesting_level() == 0);
  1287. m_parser_pause_flag = false;
  1288. // FIXME: Handle tokenizer insertion point stuff here too.
  1289. }
  1290. }
  1291. return;
  1292. }
  1293. if (token.is_end_tag()) {
  1294. m_stack_of_open_elements.pop();
  1295. m_insertion_mode = m_original_insertion_mode;
  1296. return;
  1297. }
  1298. TODO();
  1299. }
  1300. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1301. {
  1302. while (!current_node().tag_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1303. m_stack_of_open_elements.pop();
  1304. }
  1305. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1306. {
  1307. while (!current_node().tag_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1308. m_stack_of_open_elements.pop();
  1309. }
  1310. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1311. {
  1312. while (!current_node().tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1313. m_stack_of_open_elements.pop();
  1314. }
  1315. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1316. {
  1317. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1318. clear_the_stack_back_to_a_table_row_context();
  1319. insert_html_element(token);
  1320. m_insertion_mode = InsertionMode::InCell;
  1321. m_list_of_active_formatting_elements.add_marker();
  1322. return;
  1323. }
  1324. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1325. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1326. PARSE_ERROR();
  1327. return;
  1328. }
  1329. clear_the_stack_back_to_a_table_row_context();
  1330. m_stack_of_open_elements.pop();
  1331. m_insertion_mode = InsertionMode::InTableBody;
  1332. return;
  1333. }
  1334. 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)) {
  1335. if (m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1336. PARSE_ERROR();
  1337. return;
  1338. }
  1339. clear_the_stack_back_to_a_table_row_context();
  1340. m_stack_of_open_elements.pop();
  1341. m_insertion_mode = InsertionMode::InTableBody;
  1342. process_using_the_rules_for(m_insertion_mode, token);
  1343. return;
  1344. }
  1345. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1346. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1347. PARSE_ERROR();
  1348. return;
  1349. }
  1350. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1351. return;
  1352. }
  1353. clear_the_stack_back_to_a_table_row_context();
  1354. m_stack_of_open_elements.pop();
  1355. m_insertion_mode = InsertionMode::InTableBody;
  1356. process_using_the_rules_for(m_insertion_mode, token);
  1357. return;
  1358. }
  1359. 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)) {
  1360. PARSE_ERROR();
  1361. return;
  1362. }
  1363. process_using_the_rules_for(InsertionMode::InTable, token);
  1364. }
  1365. void HTMLDocumentParser::close_the_cell()
  1366. {
  1367. generate_implied_end_tags();
  1368. if (!current_node().tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1369. PARSE_ERROR();
  1370. }
  1371. while (!current_node().tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1372. m_stack_of_open_elements.pop();
  1373. m_stack_of_open_elements.pop();
  1374. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1375. m_insertion_mode = InsertionMode::InRow;
  1376. }
  1377. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1378. {
  1379. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1380. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1381. PARSE_ERROR();
  1382. return;
  1383. }
  1384. generate_implied_end_tags();
  1385. if (current_node().tag_name() != token.tag_name()) {
  1386. PARSE_ERROR();
  1387. }
  1388. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1389. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1390. m_insertion_mode = InsertionMode::InRow;
  1391. return;
  1392. }
  1393. 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)) {
  1394. 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)) {
  1395. PARSE_ERROR();
  1396. return;
  1397. }
  1398. close_the_cell();
  1399. process_using_the_rules_for(m_insertion_mode, token);
  1400. return;
  1401. }
  1402. 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)) {
  1403. PARSE_ERROR();
  1404. return;
  1405. }
  1406. 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)) {
  1407. if (m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1408. PARSE_ERROR();
  1409. return;
  1410. }
  1411. close_the_cell();
  1412. // Reprocess the token.
  1413. process_using_the_rules_for(m_insertion_mode, token);
  1414. return;
  1415. }
  1416. process_using_the_rules_for(InsertionMode::InBody, token);
  1417. }
  1418. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1419. {
  1420. if (token.is_character()) {
  1421. if (token.codepoint() == 0) {
  1422. PARSE_ERROR();
  1423. return;
  1424. }
  1425. m_pending_table_character_tokens.append(token);
  1426. return;
  1427. }
  1428. for (auto& pending_token : m_pending_table_character_tokens) {
  1429. ASSERT(pending_token.is_character());
  1430. if (!pending_token.is_parser_whitespace()) {
  1431. // FIXME: If any of the tokens in the pending table character tokens list
  1432. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1433. // reprocess the character tokens in the pending table character tokens list using
  1434. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1435. TODO();
  1436. }
  1437. }
  1438. for (auto& pending_token : m_pending_table_character_tokens) {
  1439. insert_character(pending_token.codepoint());
  1440. }
  1441. m_insertion_mode = m_original_insertion_mode;
  1442. process_using_the_rules_for(m_insertion_mode, token);
  1443. }
  1444. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1445. {
  1446. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1447. clear_the_stack_back_to_a_table_body_context();
  1448. insert_html_element(token);
  1449. m_insertion_mode = InsertionMode::InRow;
  1450. return;
  1451. }
  1452. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1453. PARSE_ERROR();
  1454. clear_the_stack_back_to_a_table_body_context();
  1455. HTMLToken fake_tr_token;
  1456. fake_tr_token.m_type = HTMLToken::Type::StartTag;
  1457. fake_tr_token.m_tag.tag_name.append(HTML::TagNames::tr);
  1458. insert_html_element(fake_tr_token);
  1459. m_insertion_mode = InsertionMode::InRow;
  1460. process_using_the_rules_for(m_insertion_mode, token);
  1461. return;
  1462. }
  1463. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1464. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1465. PARSE_ERROR();
  1466. return;
  1467. }
  1468. clear_the_stack_back_to_a_table_body_context();
  1469. m_stack_of_open_elements.pop();
  1470. m_insertion_mode = InsertionMode::InTable;
  1471. return;
  1472. }
  1473. 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))
  1474. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1475. // 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.
  1476. clear_the_stack_back_to_a_table_body_context();
  1477. m_stack_of_open_elements.pop();
  1478. m_insertion_mode = InsertionMode::InTable;
  1479. process_using_the_rules_for(InsertionMode::InTable, token);
  1480. return;
  1481. }
  1482. 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)) {
  1483. PARSE_ERROR();
  1484. return;
  1485. }
  1486. process_using_the_rules_for(InsertionMode::InTable, token);
  1487. }
  1488. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1489. {
  1490. 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)) {
  1491. m_pending_table_character_tokens.clear();
  1492. m_original_insertion_mode = m_insertion_mode;
  1493. m_insertion_mode = InsertionMode::InTableText;
  1494. process_using_the_rules_for(InsertionMode::InTableText, token);
  1495. return;
  1496. }
  1497. if (token.is_comment()) {
  1498. insert_comment(token);
  1499. return;
  1500. }
  1501. if (token.is_doctype()) {
  1502. PARSE_ERROR();
  1503. return;
  1504. }
  1505. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1506. TODO();
  1507. }
  1508. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1509. TODO();
  1510. }
  1511. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1512. TODO();
  1513. }
  1514. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1515. clear_the_stack_back_to_a_table_context();
  1516. insert_html_element(token);
  1517. m_insertion_mode = InsertionMode::InTableBody;
  1518. return;
  1519. }
  1520. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1521. clear_the_stack_back_to_a_table_context();
  1522. HTMLToken fake_tbody_token;
  1523. fake_tbody_token.m_type = HTMLToken::Type::StartTag;
  1524. fake_tbody_token.m_tag.tag_name.append(HTML::TagNames::tbody);
  1525. insert_html_element(fake_tbody_token);
  1526. m_insertion_mode = InsertionMode::InTableBody;
  1527. process_using_the_rules_for(InsertionMode::InTableBody, token);
  1528. return;
  1529. }
  1530. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1531. PARSE_ERROR();
  1532. TODO();
  1533. }
  1534. if (token.is_end_tag()) {
  1535. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1536. PARSE_ERROR();
  1537. return;
  1538. }
  1539. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1540. reset_the_insertion_mode_appropriately();
  1541. return;
  1542. }
  1543. TODO();
  1544. }
  1545. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1546. {
  1547. (void)token;
  1548. TODO();
  1549. }
  1550. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  1551. {
  1552. if (token.is_character()) {
  1553. if (token.codepoint() == 0) {
  1554. PARSE_ERROR();
  1555. return;
  1556. }
  1557. insert_character(token.codepoint());
  1558. return;
  1559. }
  1560. if (token.is_comment()) {
  1561. insert_comment(token);
  1562. return;
  1563. }
  1564. if (token.is_doctype()) {
  1565. PARSE_ERROR();
  1566. return;
  1567. }
  1568. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  1569. process_using_the_rules_for(InsertionMode::InBody, token);
  1570. return;
  1571. }
  1572. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  1573. if (current_node().tag_name() == HTML::TagNames::option) {
  1574. m_stack_of_open_elements.pop();
  1575. }
  1576. insert_html_element(token);
  1577. return;
  1578. }
  1579. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1580. if (current_node().tag_name() == HTML::TagNames::option) {
  1581. m_stack_of_open_elements.pop();
  1582. }
  1583. if (current_node().tag_name() == HTML::TagNames::optgroup) {
  1584. m_stack_of_open_elements.pop();
  1585. }
  1586. insert_html_element(token);
  1587. return;
  1588. }
  1589. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  1590. if (current_node().tag_name() == HTML::TagNames::option && node_before_current_node().tag_name() == HTML::TagNames::optgroup)
  1591. m_stack_of_open_elements.pop();
  1592. if (current_node().tag_name() == HTML::TagNames::optgroup) {
  1593. m_stack_of_open_elements.pop();
  1594. } else {
  1595. PARSE_ERROR();
  1596. return;
  1597. }
  1598. return;
  1599. }
  1600. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  1601. if (current_node().tag_name() == HTML::TagNames::option) {
  1602. m_stack_of_open_elements.pop();
  1603. } else {
  1604. PARSE_ERROR();
  1605. return;
  1606. }
  1607. return;
  1608. }
  1609. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  1610. if (m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  1611. PARSE_ERROR();
  1612. return;
  1613. }
  1614. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1615. reset_the_insertion_mode_appropriately();
  1616. return;
  1617. }
  1618. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1619. PARSE_ERROR();
  1620. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select))
  1621. return;
  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().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  1627. PARSE_ERROR();
  1628. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  1629. return;
  1630. }
  1631. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  1632. reset_the_insertion_mode_appropriately();
  1633. process_using_the_rules_for(m_insertion_mode, token);
  1634. return;
  1635. }
  1636. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  1637. process_using_the_rules_for(InsertionMode::InHead, token);
  1638. return;
  1639. }
  1640. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  1641. process_using_the_rules_for(InsertionMode::InHead, token);
  1642. return;
  1643. }
  1644. if (token.is_end_of_file()) {
  1645. process_using_the_rules_for(InsertionMode::InBody, token);
  1646. return;
  1647. }
  1648. PARSE_ERROR();
  1649. }
  1650. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  1651. {
  1652. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1653. RefPtr<Element> node = m_stack_of_open_elements.elements().at(i);
  1654. if (node->tag_name() == HTML::TagNames::select) {
  1655. TODO();
  1656. }
  1657. if (node->tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1658. m_insertion_mode = InsertionMode::InCell;
  1659. return;
  1660. }
  1661. if (node->tag_name() == HTML::TagNames::tr) {
  1662. m_insertion_mode = InsertionMode::InRow;
  1663. return;
  1664. }
  1665. if (node->tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  1666. m_insertion_mode = InsertionMode::InTableBody;
  1667. return;
  1668. }
  1669. if (node->tag_name() == HTML::TagNames::caption) {
  1670. m_insertion_mode = InsertionMode::InCaption;
  1671. return;
  1672. }
  1673. if (node->tag_name() == HTML::TagNames::colgroup) {
  1674. m_insertion_mode = InsertionMode::InColumnGroup;
  1675. return;
  1676. }
  1677. if (node->tag_name() == HTML::TagNames::table) {
  1678. m_insertion_mode = InsertionMode::InTable;
  1679. return;
  1680. }
  1681. if (node->tag_name() == HTML::TagNames::template_) {
  1682. TODO();
  1683. }
  1684. if (node->tag_name() == HTML::TagNames::body) {
  1685. m_insertion_mode = InsertionMode::InBody;
  1686. return;
  1687. }
  1688. if (node->tag_name() == HTML::TagNames::frameset) {
  1689. m_insertion_mode = InsertionMode::InFrameset;
  1690. if (m_parsing_fragment) {
  1691. TODO();
  1692. }
  1693. return;
  1694. }
  1695. if (node->tag_name() == HTML::TagNames::html) {
  1696. TODO();
  1697. }
  1698. }
  1699. m_insertion_mode = InsertionMode::InBody;
  1700. if (m_parsing_fragment) {
  1701. TODO();
  1702. }
  1703. }
  1704. const char* HTMLDocumentParser::insertion_mode_name() const
  1705. {
  1706. switch (m_insertion_mode) {
  1707. #define __ENUMERATE_INSERTION_MODE(mode) \
  1708. case InsertionMode::mode: \
  1709. return #mode;
  1710. ENUMERATE_INSERTION_MODES
  1711. #undef __ENUMERATE_INSERTION_MODE
  1712. }
  1713. ASSERT_NOT_REACHED();
  1714. }
  1715. Document& HTMLDocumentParser::document()
  1716. {
  1717. return *m_document;
  1718. }
  1719. }