HTMLDocumentParser.cpp 116 KB

1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677787980818283848586878889909192939495969798991001011021031041051061071081091101111121131141151161171181191201211221231241251261271281291301311321331341351361371381391401411421431441451461471481491501511521531541551561571581591601611621631641651661671681691701711721731741751761771781791801811821831841851861871881891901911921931941951961971981992002012022032042052062072082092102112122132142152162172182192202212222232242252262272282292302312322332342352362372382392402412422432442452462472482492502512522532542552562572582592602612622632642652662672682692702712722732742752762772782792802812822832842852862872882892902912922932942952962972982993003013023033043053063073083093103113123133143153163173183193203213223233243253263273283293303313323333343353363373383393403413423433443453463473483493503513523533543553563573583593603613623633643653663673683693703713723733743753763773783793803813823833843853863873883893903913923933943953963973983994004014024034044054064074084094104114124134144154164174184194204214224234244254264274284294304314324334344354364374384394404414424434444454464474484494504514524534544554564574584594604614624634644654664674684694704714724734744754764774784794804814824834844854864874884894904914924934944954964974984995005015025035045055065075085095105115125135145155165175185195205215225235245255265275285295305315325335345355365375385395405415425435445455465475485495505515525535545555565575585595605615625635645655665675685695705715725735745755765775785795805815825835845855865875885895905915925935945955965975985996006016026036046056066076086096106116126136146156166176186196206216226236246256266276286296306316326336346356366376386396406416426436446456466476486496506516526536546556566576586596606616626636646656666676686696706716726736746756766776786796806816826836846856866876886896906916926936946956966976986997007017027037047057067077087097107117127137147157167177187197207217227237247257267277287297307317327337347357367377387397407417427437447457467477487497507517527537547557567577587597607617627637647657667677687697707717727737747757767777787797807817827837847857867877887897907917927937947957967977987998008018028038048058068078088098108118128138148158168178188198208218228238248258268278288298308318328338348358368378388398408418428438448458468478488498508518528538548558568578588598608618628638648658668678688698708718728738748758768778788798808818828838848858868878888898908918928938948958968978988999009019029039049059069079089099109119129139149159169179189199209219229239249259269279289299309319329339349359369379389399409419429439449459469479489499509519529539549559569579589599609619629639649659669679689699709719729739749759769779789799809819829839849859869879889899909919929939949959969979989991000100110021003100410051006100710081009101010111012101310141015101610171018101910201021102210231024102510261027102810291030103110321033103410351036103710381039104010411042104310441045104610471048104910501051105210531054105510561057105810591060106110621063106410651066106710681069107010711072107310741075107610771078107910801081108210831084108510861087108810891090109110921093109410951096109710981099110011011102110311041105110611071108110911101111111211131114111511161117111811191120112111221123112411251126112711281129113011311132113311341135113611371138113911401141114211431144114511461147114811491150115111521153115411551156115711581159116011611162116311641165116611671168116911701171117211731174117511761177117811791180118111821183118411851186118711881189119011911192119311941195119611971198119912001201120212031204120512061207120812091210121112121213121412151216121712181219122012211222122312241225122612271228122912301231123212331234123512361237123812391240124112421243124412451246124712481249125012511252125312541255125612571258125912601261126212631264126512661267126812691270127112721273127412751276127712781279128012811282128312841285128612871288128912901291129212931294129512961297129812991300130113021303130413051306130713081309131013111312131313141315131613171318131913201321132213231324132513261327132813291330133113321333133413351336133713381339134013411342134313441345134613471348134913501351135213531354135513561357135813591360136113621363136413651366136713681369137013711372137313741375137613771378137913801381138213831384138513861387138813891390139113921393139413951396139713981399140014011402140314041405140614071408140914101411141214131414141514161417141814191420142114221423142414251426142714281429143014311432143314341435143614371438143914401441144214431444144514461447144814491450145114521453145414551456145714581459146014611462146314641465146614671468146914701471147214731474147514761477147814791480148114821483148414851486148714881489149014911492149314941495149614971498149915001501150215031504150515061507150815091510151115121513151415151516151715181519152015211522152315241525152615271528152915301531153215331534153515361537153815391540154115421543154415451546154715481549155015511552155315541555155615571558155915601561156215631564156515661567156815691570157115721573157415751576157715781579158015811582158315841585158615871588158915901591159215931594159515961597159815991600160116021603160416051606160716081609161016111612161316141615161616171618161916201621162216231624162516261627162816291630163116321633163416351636163716381639164016411642164316441645164616471648164916501651165216531654165516561657165816591660166116621663166416651666166716681669167016711672167316741675167616771678167916801681168216831684168516861687168816891690169116921693169416951696169716981699170017011702170317041705170617071708170917101711171217131714171517161717171817191720172117221723172417251726172717281729173017311732173317341735173617371738173917401741174217431744174517461747174817491750175117521753175417551756175717581759176017611762176317641765176617671768176917701771177217731774177517761777177817791780178117821783178417851786178717881789179017911792179317941795179617971798179918001801180218031804180518061807180818091810181118121813181418151816181718181819182018211822182318241825182618271828182918301831183218331834183518361837183818391840184118421843184418451846184718481849185018511852185318541855185618571858185918601861186218631864186518661867186818691870187118721873187418751876187718781879188018811882188318841885188618871888188918901891189218931894189518961897189818991900190119021903190419051906190719081909191019111912191319141915191619171918191919201921192219231924192519261927192819291930193119321933193419351936193719381939194019411942194319441945194619471948194919501951195219531954195519561957195819591960196119621963196419651966196719681969197019711972197319741975197619771978197919801981198219831984198519861987198819891990199119921993199419951996199719981999200020012002200320042005200620072008200920102011201220132014201520162017201820192020202120222023202420252026202720282029203020312032203320342035203620372038203920402041204220432044204520462047204820492050205120522053205420552056205720582059206020612062206320642065206620672068206920702071207220732074207520762077207820792080208120822083208420852086208720882089209020912092209320942095209620972098209921002101210221032104210521062107210821092110211121122113211421152116211721182119212021212122212321242125212621272128212921302131213221332134213521362137213821392140214121422143214421452146214721482149215021512152215321542155215621572158215921602161216221632164216521662167216821692170217121722173217421752176217721782179218021812182218321842185218621872188218921902191219221932194219521962197219821992200220122022203220422052206220722082209221022112212221322142215221622172218221922202221222222232224222522262227222822292230223122322233223422352236223722382239224022412242224322442245224622472248224922502251225222532254225522562257225822592260226122622263226422652266226722682269227022712272227322742275227622772278227922802281228222832284228522862287228822892290229122922293229422952296229722982299230023012302230323042305230623072308230923102311231223132314231523162317231823192320232123222323232423252326232723282329233023312332233323342335233623372338233923402341234223432344234523462347234823492350235123522353235423552356235723582359236023612362236323642365236623672368236923702371237223732374237523762377237823792380238123822383238423852386238723882389239023912392239323942395239623972398239924002401240224032404240524062407240824092410241124122413241424152416241724182419242024212422242324242425242624272428242924302431243224332434243524362437243824392440244124422443244424452446244724482449245024512452245324542455245624572458245924602461246224632464246524662467246824692470247124722473247424752476247724782479248024812482248324842485248624872488248924902491249224932494249524962497249824992500250125022503250425052506250725082509251025112512251325142515251625172518251925202521252225232524252525262527252825292530253125322533253425352536253725382539254025412542254325442545254625472548254925502551255225532554255525562557255825592560256125622563256425652566256725682569257025712572257325742575257625772578257925802581258225832584258525862587258825892590259125922593259425952596259725982599260026012602260326042605260626072608260926102611261226132614261526162617261826192620262126222623262426252626262726282629263026312632263326342635263626372638263926402641264226432644264526462647264826492650265126522653265426552656265726582659266026612662266326642665266626672668266926702671267226732674267526762677267826792680268126822683268426852686268726882689269026912692269326942695269626972698269927002701270227032704270527062707270827092710271127122713271427152716271727182719272027212722272327242725272627272728272927302731273227332734273527362737273827392740274127422743274427452746274727482749275027512752275327542755275627572758275927602761276227632764276527662767276827692770277127722773277427752776277727782779278027812782278327842785278627872788278927902791279227932794279527962797279827992800280128022803280428052806280728082809281028112812281328142815281628172818281928202821282228232824282528262827282828292830283128322833283428352836283728382839284028412842284328442845284628472848284928502851285228532854285528562857285828592860286128622863286428652866286728682869287028712872287328742875287628772878287928802881288228832884288528862887288828892890289128922893289428952896289728982899290029012902290329042905290629072908290929102911291229132914291529162917291829192920292129222923292429252926292729282929293029312932293329342935293629372938293929402941294229432944294529462947294829492950295129522953295429552956295729582959296029612962296329642965296629672968296929702971297229732974297529762977297829792980298129822983298429852986298729882989299029912992299329942995299629972998299930003001300230033004300530063007300830093010301130123013301430153016301730183019302030213022302330243025302630273028302930303031303230333034303530363037303830393040
  1. /*
  2. * Copyright (c) 2020, Andreas Kling <kling@serenityos.org>
  3. *
  4. * SPDX-License-Identifier: BSD-2-Clause
  5. */
  6. #include <AK/Debug.h>
  7. #include <AK/SourceLocation.h>
  8. #include <AK/Utf32View.h>
  9. #include <LibTextCodec/Decoder.h>
  10. #include <LibWeb/DOM/Comment.h>
  11. #include <LibWeb/DOM/Document.h>
  12. #include <LibWeb/DOM/DocumentType.h>
  13. #include <LibWeb/DOM/ElementFactory.h>
  14. #include <LibWeb/DOM/Event.h>
  15. #include <LibWeb/DOM/Text.h>
  16. #include <LibWeb/DOM/Window.h>
  17. #include <LibWeb/HTML/EventNames.h>
  18. #include <LibWeb/HTML/HTMLFormElement.h>
  19. #include <LibWeb/HTML/HTMLHeadElement.h>
  20. #include <LibWeb/HTML/HTMLScriptElement.h>
  21. #include <LibWeb/HTML/HTMLTableElement.h>
  22. #include <LibWeb/HTML/HTMLTemplateElement.h>
  23. #include <LibWeb/HTML/Parser/HTMLDocumentParser.h>
  24. #include <LibWeb/HTML/Parser/HTMLToken.h>
  25. #include <LibWeb/Namespace.h>
  26. #include <LibWeb/SVG/TagNames.h>
  27. namespace Web::HTML {
  28. static inline void log_parse_error(const SourceLocation& location = SourceLocation::current())
  29. {
  30. dbgln("Parse error! {}", location);
  31. }
  32. static Vector<FlyString> s_quirks_public_ids = {
  33. "+//Silmaril//dtd html Pro v0r11 19970101//",
  34. "-//AS//DTD HTML 3.0 asWedit + extensions//",
  35. "-//AdvaSoft Ltd//DTD HTML 3.0 asWedit + extensions//",
  36. "-//IETF//DTD HTML 2.0 Level 1//",
  37. "-//IETF//DTD HTML 2.0 Level 2//",
  38. "-//IETF//DTD HTML 2.0 Strict Level 1//",
  39. "-//IETF//DTD HTML 2.0 Strict Level 2//",
  40. "-//IETF//DTD HTML 2.0 Strict//",
  41. "-//IETF//DTD HTML 2.0//",
  42. "-//IETF//DTD HTML 2.1E//",
  43. "-//IETF//DTD HTML 3.0//",
  44. "-//IETF//DTD HTML 3.2 Final//",
  45. "-//IETF//DTD HTML 3.2//",
  46. "-//IETF//DTD HTML 3//",
  47. "-//IETF//DTD HTML Level 0//",
  48. "-//IETF//DTD HTML Level 1//",
  49. "-//IETF//DTD HTML Level 2//",
  50. "-//IETF//DTD HTML Level 3//",
  51. "-//IETF//DTD HTML Strict Level 0//",
  52. "-//IETF//DTD HTML Strict Level 1//",
  53. "-//IETF//DTD HTML Strict Level 2//",
  54. "-//IETF//DTD HTML Strict Level 3//",
  55. "-//IETF//DTD HTML Strict//",
  56. "-//IETF//DTD HTML//",
  57. "-//Metrius//DTD Metrius Presentational//",
  58. "-//Microsoft//DTD Internet Explorer 2.0 HTML Strict//",
  59. "-//Microsoft//DTD Internet Explorer 2.0 HTML//",
  60. "-//Microsoft//DTD Internet Explorer 2.0 Tables//",
  61. "-//Microsoft//DTD Internet Explorer 3.0 HTML Strict//",
  62. "-//Microsoft//DTD Internet Explorer 3.0 HTML//",
  63. "-//Microsoft//DTD Internet Explorer 3.0 Tables//",
  64. "-//Netscape Comm. Corp.//DTD HTML//",
  65. "-//Netscape Comm. Corp.//DTD Strict HTML//",
  66. "-//O'Reilly and Associates//DTD HTML 2.0//",
  67. "-//O'Reilly and Associates//DTD HTML Extended 1.0//",
  68. "-//O'Reilly and Associates//DTD HTML Extended Relaxed 1.0//",
  69. "-//SQ//DTD HTML 2.0 HoTMetaL + extensions//",
  70. "-//SoftQuad Software//DTD HoTMetaL PRO 6.0::19990601::extensions to HTML 4.0//",
  71. "-//SoftQuad//DTD HoTMetaL PRO 4.0::19971010::extensions to HTML 4.0//",
  72. "-//Spyglass//DTD HTML 2.0 Extended//",
  73. "-//Sun Microsystems Corp.//DTD HotJava HTML//",
  74. "-//Sun Microsystems Corp.//DTD HotJava Strict HTML//",
  75. "-//W3C//DTD HTML 3 1995-03-24//",
  76. "-//W3C//DTD HTML 3.2 Draft//",
  77. "-//W3C//DTD HTML 3.2 Final//",
  78. "-//W3C//DTD HTML 3.2//",
  79. "-//W3C//DTD HTML 3.2S Draft//",
  80. "-//W3C//DTD HTML 4.0 Frameset//",
  81. "-//W3C//DTD HTML 4.0 Transitional//",
  82. "-//W3C//DTD HTML Experimental 19960712//",
  83. "-//W3C//DTD HTML Experimental 970421//",
  84. "-//W3C//DTD W3 HTML//",
  85. "-//W3O//DTD W3 HTML 3.0//",
  86. "-//WebTechs//DTD Mozilla HTML 2.0//",
  87. "-//WebTechs//DTD Mozilla HTML//"
  88. };
  89. RefPtr<DOM::Document> parse_html_document(const StringView& data, const URL& url, const String& encoding)
  90. {
  91. auto document = DOM::Document::create(url);
  92. HTMLDocumentParser parser(document, data, encoding);
  93. parser.run(url);
  94. return document;
  95. }
  96. HTMLDocumentParser::HTMLDocumentParser(DOM::Document& document, const StringView& input, const String& encoding)
  97. : m_tokenizer(input, encoding)
  98. , m_document(document)
  99. {
  100. m_document->set_should_invalidate_styles_on_attribute_changes(false);
  101. m_document->set_encoding(TextCodec::get_standardized_encoding(encoding));
  102. }
  103. HTMLDocumentParser::~HTMLDocumentParser()
  104. {
  105. m_document->set_should_invalidate_styles_on_attribute_changes(true);
  106. }
  107. void HTMLDocumentParser::run(const URL& url)
  108. {
  109. m_document->set_url(url);
  110. m_document->set_source(m_tokenizer.source());
  111. for (;;) {
  112. auto optional_token = m_tokenizer.next_token();
  113. if (!optional_token.has_value())
  114. break;
  115. auto& token = optional_token.value();
  116. dbgln_if(PARSER_DEBUG, "[{}] {}", insertion_mode_name(), token.to_string());
  117. // FIXME: If the adjusted current node is a MathML text integration point and the token is a start tag whose tag name is neither "mglyph" nor "malignmark"
  118. // FIXME: If the adjusted current node is a MathML text integration point and the token is a character token
  119. // FIXME: If the adjusted current node is a MathML annotation-xml element and the token is a start tag whose tag name is "svg"
  120. // FIXME: If the adjusted current node is an HTML integration point and the token is a start tag
  121. // FIXME: If the adjusted current node is an HTML integration point and the token is a character token
  122. if (m_stack_of_open_elements.is_empty()
  123. || adjusted_current_node().namespace_() == Namespace::HTML
  124. || token.is_end_of_file()) {
  125. process_using_the_rules_for(m_insertion_mode, token);
  126. } else {
  127. process_using_the_rules_for_foreign_content(token);
  128. }
  129. if (m_stop_parsing) {
  130. dbgln_if(PARSER_DEBUG, "Stop parsing{}! :^)", m_parsing_fragment ? " fragment" : "");
  131. break;
  132. }
  133. }
  134. flush_character_insertions();
  135. // "The end"
  136. m_document->set_ready_state("interactive");
  137. auto scripts_to_execute_when_parsing_has_finished = m_document->take_scripts_to_execute_when_parsing_has_finished({});
  138. for (auto& script : scripts_to_execute_when_parsing_has_finished) {
  139. // FIXME: Spin the event loop until the script is ready to be parser executed and there's no style sheets blocking scripts.
  140. script.execute_script();
  141. }
  142. auto content_loaded_event = DOM::Event::create(HTML::EventNames::DOMContentLoaded);
  143. content_loaded_event->set_bubbles(true);
  144. m_document->dispatch_event(content_loaded_event);
  145. // FIXME: The document parser shouldn't execute these, it should just spin the event loop until the list becomes empty.
  146. // FIXME: Once the set has been added, also spin the event loop until the set becomes empty.
  147. auto scripts_to_execute_as_soon_as_possible = m_document->take_scripts_to_execute_as_soon_as_possible({});
  148. for (auto& script : scripts_to_execute_as_soon_as_possible) {
  149. script.execute_script();
  150. }
  151. // FIXME: Spin the event loop until there is nothing that delays the load event in the Document.
  152. m_document->set_ready_state("complete");
  153. m_document->window().dispatch_event(DOM::Event::create(HTML::EventNames::load));
  154. m_document->set_ready_for_post_load_tasks(true);
  155. m_document->completely_finish_loading();
  156. }
  157. void HTMLDocumentParser::process_using_the_rules_for(InsertionMode mode, HTMLToken& token)
  158. {
  159. switch (mode) {
  160. case InsertionMode::Initial:
  161. handle_initial(token);
  162. break;
  163. case InsertionMode::BeforeHTML:
  164. handle_before_html(token);
  165. break;
  166. case InsertionMode::BeforeHead:
  167. handle_before_head(token);
  168. break;
  169. case InsertionMode::InHead:
  170. handle_in_head(token);
  171. break;
  172. case InsertionMode::InHeadNoscript:
  173. handle_in_head_noscript(token);
  174. break;
  175. case InsertionMode::AfterHead:
  176. handle_after_head(token);
  177. break;
  178. case InsertionMode::InBody:
  179. handle_in_body(token);
  180. break;
  181. case InsertionMode::AfterBody:
  182. handle_after_body(token);
  183. break;
  184. case InsertionMode::AfterAfterBody:
  185. handle_after_after_body(token);
  186. break;
  187. case InsertionMode::Text:
  188. handle_text(token);
  189. break;
  190. case InsertionMode::InTable:
  191. handle_in_table(token);
  192. break;
  193. case InsertionMode::InTableBody:
  194. handle_in_table_body(token);
  195. break;
  196. case InsertionMode::InRow:
  197. handle_in_row(token);
  198. break;
  199. case InsertionMode::InCell:
  200. handle_in_cell(token);
  201. break;
  202. case InsertionMode::InTableText:
  203. handle_in_table_text(token);
  204. break;
  205. case InsertionMode::InSelectInTable:
  206. handle_in_select_in_table(token);
  207. break;
  208. case InsertionMode::InSelect:
  209. handle_in_select(token);
  210. break;
  211. case InsertionMode::InCaption:
  212. handle_in_caption(token);
  213. break;
  214. case InsertionMode::InColumnGroup:
  215. handle_in_column_group(token);
  216. break;
  217. case InsertionMode::InTemplate:
  218. handle_in_template(token);
  219. break;
  220. case InsertionMode::InFrameset:
  221. handle_in_frameset(token);
  222. break;
  223. case InsertionMode::AfterFrameset:
  224. handle_after_frameset(token);
  225. break;
  226. case InsertionMode::AfterAfterFrameset:
  227. handle_after_after_frameset(token);
  228. break;
  229. default:
  230. VERIFY_NOT_REACHED();
  231. }
  232. }
  233. DOM::QuirksMode HTMLDocumentParser::which_quirks_mode(const HTMLToken& doctype_token) const
  234. {
  235. if (doctype_token.m_doctype.force_quirks)
  236. return DOM::QuirksMode::Yes;
  237. // NOTE: The tokenizer puts the name into lower case for us.
  238. if (doctype_token.m_doctype.name.to_string() != "html")
  239. return DOM::QuirksMode::Yes;
  240. auto public_identifier = doctype_token.m_doctype.public_identifier.to_string();
  241. auto system_identifier = doctype_token.m_doctype.system_identifier.to_string();
  242. if (public_identifier.equals_ignoring_case("-//W3O//DTD W3 HTML Strict 3.0//EN//"))
  243. return DOM::QuirksMode::Yes;
  244. if (public_identifier.equals_ignoring_case("-/W3C/DTD HTML 4.0 Transitional/EN"))
  245. return DOM::QuirksMode::Yes;
  246. if (public_identifier.equals_ignoring_case("HTML"))
  247. return DOM::QuirksMode::Yes;
  248. if (system_identifier.equals_ignoring_case("http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd"))
  249. return DOM::QuirksMode::Yes;
  250. for (auto& public_id : s_quirks_public_ids) {
  251. if (public_identifier.starts_with(public_id, CaseSensitivity::CaseInsensitive))
  252. return DOM::QuirksMode::Yes;
  253. }
  254. if (doctype_token.m_doctype.missing_system_identifier) {
  255. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  256. return DOM::QuirksMode::Yes;
  257. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  258. return DOM::QuirksMode::Yes;
  259. }
  260. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Frameset//", CaseSensitivity::CaseInsensitive))
  261. return DOM::QuirksMode::Limited;
  262. if (public_identifier.starts_with("-//W3C//DTD XHTML 1.0 Transitional//", CaseSensitivity::CaseInsensitive))
  263. return DOM::QuirksMode::Limited;
  264. if (!doctype_token.m_doctype.missing_system_identifier) {
  265. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Frameset//", CaseSensitivity::CaseInsensitive))
  266. return DOM::QuirksMode::Limited;
  267. if (public_identifier.starts_with("-//W3C//DTD HTML 4.01 Transitional//", CaseSensitivity::CaseInsensitive))
  268. return DOM::QuirksMode::Limited;
  269. }
  270. return DOM::QuirksMode::No;
  271. }
  272. void HTMLDocumentParser::handle_initial(HTMLToken& token)
  273. {
  274. if (token.is_character() && token.is_parser_whitespace()) {
  275. return;
  276. }
  277. if (token.is_comment()) {
  278. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  279. document().append_child(move(comment));
  280. return;
  281. }
  282. if (token.is_doctype()) {
  283. auto doctype = adopt_ref(*new DOM::DocumentType(document()));
  284. doctype->set_name(token.m_doctype.name.to_string());
  285. doctype->set_public_id(token.m_doctype.public_identifier.to_string());
  286. doctype->set_system_id(token.m_doctype.system_identifier.to_string());
  287. document().append_child(move(doctype));
  288. document().set_quirks_mode(which_quirks_mode(token));
  289. m_insertion_mode = InsertionMode::BeforeHTML;
  290. return;
  291. }
  292. log_parse_error();
  293. document().set_quirks_mode(DOM::QuirksMode::Yes);
  294. m_insertion_mode = InsertionMode::BeforeHTML;
  295. process_using_the_rules_for(InsertionMode::BeforeHTML, token);
  296. }
  297. void HTMLDocumentParser::handle_before_html(HTMLToken& token)
  298. {
  299. if (token.is_doctype()) {
  300. log_parse_error();
  301. return;
  302. }
  303. if (token.is_comment()) {
  304. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  305. document().append_child(move(comment));
  306. return;
  307. }
  308. if (token.is_character() && token.is_parser_whitespace()) {
  309. return;
  310. }
  311. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  312. auto element = create_element_for(token, Namespace::HTML);
  313. document().append_child(element);
  314. m_stack_of_open_elements.push(move(element));
  315. m_insertion_mode = InsertionMode::BeforeHead;
  316. return;
  317. }
  318. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  319. goto AnythingElse;
  320. }
  321. if (token.is_end_tag()) {
  322. log_parse_error();
  323. return;
  324. }
  325. AnythingElse:
  326. auto element = create_element(document(), HTML::TagNames::html, Namespace::HTML);
  327. document().append_child(element);
  328. m_stack_of_open_elements.push(element);
  329. // 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.
  330. m_insertion_mode = InsertionMode::BeforeHead;
  331. process_using_the_rules_for(InsertionMode::BeforeHead, token);
  332. return;
  333. }
  334. DOM::Element& HTMLDocumentParser::current_node()
  335. {
  336. return m_stack_of_open_elements.current_node();
  337. }
  338. DOM::Element& HTMLDocumentParser::adjusted_current_node()
  339. {
  340. if (m_parsing_fragment && m_stack_of_open_elements.elements().size() == 1)
  341. return *m_context_element;
  342. return current_node();
  343. }
  344. DOM::Element& HTMLDocumentParser::node_before_current_node()
  345. {
  346. return m_stack_of_open_elements.elements().at(m_stack_of_open_elements.elements().size() - 2);
  347. }
  348. HTMLDocumentParser::AdjustedInsertionLocation HTMLDocumentParser::find_appropriate_place_for_inserting_node()
  349. {
  350. auto& target = current_node();
  351. HTMLDocumentParser::AdjustedInsertionLocation adjusted_insertion_location;
  352. if (m_foster_parenting && target.local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  353. auto last_template = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::template_);
  354. auto last_table = m_stack_of_open_elements.last_element_with_tag_name(HTML::TagNames::table);
  355. if (last_template.element && (!last_table.element || last_template.index > last_table.index)) {
  356. // This returns the template content, so no need to check the parent is a template.
  357. return { downcast<HTMLTemplateElement>(last_template.element)->content(), nullptr };
  358. }
  359. if (!last_table.element) {
  360. VERIFY(m_parsing_fragment);
  361. // Guaranteed not to be a template element (it will be the html element),
  362. // so no need to check the parent is a template.
  363. return { m_stack_of_open_elements.elements().first(), nullptr };
  364. }
  365. if (last_table.element->parent_node())
  366. adjusted_insertion_location = { last_table.element->parent_node(), last_table.element };
  367. else
  368. adjusted_insertion_location = { m_stack_of_open_elements.element_before(*last_table.element), nullptr };
  369. } else {
  370. adjusted_insertion_location = { target, nullptr };
  371. }
  372. if (is<HTMLTemplateElement>(*adjusted_insertion_location.parent))
  373. return { downcast<HTMLTemplateElement>(*adjusted_insertion_location.parent).content(), nullptr };
  374. return adjusted_insertion_location;
  375. }
  376. NonnullRefPtr<DOM::Element> HTMLDocumentParser::create_element_for(const HTMLToken& token, const FlyString& namespace_)
  377. {
  378. auto element = create_element(document(), token.tag_name(), namespace_);
  379. for (auto& attribute : token.m_tag.attributes) {
  380. element->set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  381. }
  382. return element;
  383. }
  384. // https://html.spec.whatwg.org/multipage/parsing.html#insert-a-foreign-element
  385. NonnullRefPtr<DOM::Element> HTMLDocumentParser::insert_foreign_element(const HTMLToken& token, const FlyString& namespace_)
  386. {
  387. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  388. // FIXME: Pass in adjusted_insertion_location.parent as the intended parent.
  389. auto element = create_element_for(token, namespace_);
  390. auto pre_insertion_validity = adjusted_insertion_location.parent->ensure_pre_insertion_validity(element, adjusted_insertion_location.insert_before_sibling);
  391. // NOTE: If it's not possible to insert the element at the adjusted insertion location, the element is simply dropped.
  392. if (!pre_insertion_validity.is_exception()) {
  393. if (!m_parsing_fragment) {
  394. // FIXME: push a new element queue onto element's relevant agent's custom element reactions stack.
  395. }
  396. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling);
  397. if (!m_parsing_fragment) {
  398. // FIXME: pop the element queue from element's relevant agent's custom element reactions stack, and invoke custom element reactions in that queue.
  399. }
  400. }
  401. m_stack_of_open_elements.push(element);
  402. return element;
  403. }
  404. NonnullRefPtr<DOM::Element> HTMLDocumentParser::insert_html_element(const HTMLToken& token)
  405. {
  406. return insert_foreign_element(token, Namespace::HTML);
  407. }
  408. void HTMLDocumentParser::handle_before_head(HTMLToken& token)
  409. {
  410. if (token.is_character() && token.is_parser_whitespace()) {
  411. return;
  412. }
  413. if (token.is_comment()) {
  414. insert_comment(token);
  415. return;
  416. }
  417. if (token.is_doctype()) {
  418. log_parse_error();
  419. return;
  420. }
  421. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  422. process_using_the_rules_for(InsertionMode::InBody, token);
  423. return;
  424. }
  425. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::head) {
  426. auto element = insert_html_element(token);
  427. m_head_element = downcast<HTMLHeadElement>(*element);
  428. m_insertion_mode = InsertionMode::InHead;
  429. return;
  430. }
  431. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  432. goto AnythingElse;
  433. }
  434. if (token.is_end_tag()) {
  435. log_parse_error();
  436. return;
  437. }
  438. AnythingElse:
  439. m_head_element = downcast<HTMLHeadElement>(*insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::head)));
  440. m_insertion_mode = InsertionMode::InHead;
  441. process_using_the_rules_for(InsertionMode::InHead, token);
  442. return;
  443. }
  444. void HTMLDocumentParser::insert_comment(HTMLToken& token)
  445. {
  446. auto data = token.m_comment_or_character.data.to_string();
  447. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  448. adjusted_insertion_location.parent->insert_before(adopt_ref(*new DOM::Comment(document(), data)), adjusted_insertion_location.insert_before_sibling);
  449. }
  450. void HTMLDocumentParser::handle_in_head(HTMLToken& token)
  451. {
  452. if (token.is_parser_whitespace()) {
  453. insert_character(token.code_point());
  454. return;
  455. }
  456. if (token.is_comment()) {
  457. insert_comment(token);
  458. return;
  459. }
  460. if (token.is_doctype()) {
  461. log_parse_error();
  462. return;
  463. }
  464. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  465. process_using_the_rules_for(InsertionMode::InBody, token);
  466. return;
  467. }
  468. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link)) {
  469. insert_html_element(token);
  470. m_stack_of_open_elements.pop();
  471. token.acknowledge_self_closing_flag_if_set();
  472. return;
  473. }
  474. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::meta) {
  475. auto element = insert_html_element(token);
  476. m_stack_of_open_elements.pop();
  477. token.acknowledge_self_closing_flag_if_set();
  478. return;
  479. }
  480. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::title) {
  481. insert_html_element(token);
  482. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  483. m_original_insertion_mode = m_insertion_mode;
  484. m_insertion_mode = InsertionMode::Text;
  485. return;
  486. }
  487. 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)) {
  488. parse_generic_raw_text_element(token);
  489. return;
  490. }
  491. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noscript && !m_scripting_enabled) {
  492. insert_html_element(token);
  493. m_insertion_mode = InsertionMode::InHeadNoscript;
  494. return;
  495. }
  496. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::script) {
  497. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  498. auto element = create_element_for(token, Namespace::HTML);
  499. auto& script_element = downcast<HTMLScriptElement>(*element);
  500. script_element.set_parser_document({}, document());
  501. script_element.set_non_blocking({}, false);
  502. if (m_parsing_fragment) {
  503. script_element.set_already_started({}, true);
  504. }
  505. if (m_invoked_via_document_write) {
  506. TODO();
  507. }
  508. adjusted_insertion_location.parent->insert_before(element, adjusted_insertion_location.insert_before_sibling, false);
  509. m_stack_of_open_elements.push(element);
  510. m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  511. m_original_insertion_mode = m_insertion_mode;
  512. m_insertion_mode = InsertionMode::Text;
  513. return;
  514. }
  515. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::head) {
  516. m_stack_of_open_elements.pop();
  517. m_insertion_mode = InsertionMode::AfterHead;
  518. return;
  519. }
  520. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  521. goto AnythingElse;
  522. }
  523. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::template_) {
  524. insert_html_element(token);
  525. m_list_of_active_formatting_elements.add_marker();
  526. m_frameset_ok = false;
  527. m_insertion_mode = InsertionMode::InTemplate;
  528. m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  529. return;
  530. }
  531. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  532. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  533. log_parse_error();
  534. return;
  535. }
  536. generate_all_implied_end_tags_thoroughly();
  537. if (current_node().local_name() != HTML::TagNames::template_)
  538. log_parse_error();
  539. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  540. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  541. m_stack_of_template_insertion_modes.take_last();
  542. reset_the_insertion_mode_appropriately();
  543. return;
  544. }
  545. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  546. log_parse_error();
  547. return;
  548. }
  549. AnythingElse:
  550. m_stack_of_open_elements.pop();
  551. m_insertion_mode = InsertionMode::AfterHead;
  552. process_using_the_rules_for(m_insertion_mode, token);
  553. }
  554. void HTMLDocumentParser::handle_in_head_noscript(HTMLToken& token)
  555. {
  556. if (token.is_doctype()) {
  557. log_parse_error();
  558. return;
  559. }
  560. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  561. process_using_the_rules_for(InsertionMode::InBody, token);
  562. return;
  563. }
  564. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::noscript) {
  565. m_stack_of_open_elements.pop();
  566. m_insertion_mode = InsertionMode::InHead;
  567. return;
  568. }
  569. if (token.is_parser_whitespace() || token.is_comment() || (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::style))) {
  570. process_using_the_rules_for(InsertionMode::InHead, token);
  571. return;
  572. }
  573. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  574. goto AnythingElse;
  575. }
  576. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::head, HTML::TagNames::noscript)) {
  577. log_parse_error();
  578. return;
  579. }
  580. AnythingElse:
  581. log_parse_error();
  582. m_stack_of_open_elements.pop();
  583. m_insertion_mode = InsertionMode::InHead;
  584. process_using_the_rules_for(m_insertion_mode, token);
  585. }
  586. void HTMLDocumentParser::parse_generic_raw_text_element(HTMLToken& token)
  587. {
  588. insert_html_element(token);
  589. m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  590. m_original_insertion_mode = m_insertion_mode;
  591. m_insertion_mode = InsertionMode::Text;
  592. }
  593. DOM::Text* HTMLDocumentParser::find_character_insertion_node()
  594. {
  595. auto adjusted_insertion_location = find_appropriate_place_for_inserting_node();
  596. if (adjusted_insertion_location.insert_before_sibling) {
  597. TODO();
  598. }
  599. if (adjusted_insertion_location.parent->is_document())
  600. return nullptr;
  601. if (adjusted_insertion_location.parent->last_child() && adjusted_insertion_location.parent->last_child()->is_text())
  602. return downcast<DOM::Text>(adjusted_insertion_location.parent->last_child());
  603. auto new_text_node = adopt_ref(*new DOM::Text(document(), ""));
  604. adjusted_insertion_location.parent->append_child(new_text_node);
  605. return new_text_node;
  606. }
  607. void HTMLDocumentParser::flush_character_insertions()
  608. {
  609. if (m_character_insertion_builder.is_empty())
  610. return;
  611. m_character_insertion_node->set_data(m_character_insertion_builder.to_string());
  612. m_character_insertion_node->parent()->children_changed();
  613. m_character_insertion_builder.clear();
  614. }
  615. void HTMLDocumentParser::insert_character(u32 data)
  616. {
  617. auto node = find_character_insertion_node();
  618. if (node == m_character_insertion_node) {
  619. m_character_insertion_builder.append(Utf32View { &data, 1 });
  620. return;
  621. }
  622. if (!m_character_insertion_node) {
  623. m_character_insertion_node = node;
  624. m_character_insertion_builder.append(Utf32View { &data, 1 });
  625. return;
  626. }
  627. flush_character_insertions();
  628. m_character_insertion_node = node;
  629. m_character_insertion_builder.append(Utf32View { &data, 1 });
  630. }
  631. void HTMLDocumentParser::handle_after_head(HTMLToken& token)
  632. {
  633. if (token.is_character() && token.is_parser_whitespace()) {
  634. insert_character(token.code_point());
  635. return;
  636. }
  637. if (token.is_comment()) {
  638. insert_comment(token);
  639. return;
  640. }
  641. if (token.is_doctype()) {
  642. log_parse_error();
  643. return;
  644. }
  645. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  646. process_using_the_rules_for(InsertionMode::InBody, token);
  647. return;
  648. }
  649. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  650. insert_html_element(token);
  651. m_frameset_ok = false;
  652. m_insertion_mode = InsertionMode::InBody;
  653. return;
  654. }
  655. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  656. insert_html_element(token);
  657. m_insertion_mode = InsertionMode::InFrameset;
  658. return;
  659. }
  660. 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)) {
  661. log_parse_error();
  662. m_stack_of_open_elements.push(*m_head_element);
  663. process_using_the_rules_for(InsertionMode::InHead, token);
  664. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  665. return entry.ptr() == m_head_element;
  666. });
  667. return;
  668. }
  669. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  670. process_using_the_rules_for(InsertionMode::InHead, token);
  671. return;
  672. }
  673. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::html, HTML::TagNames::br)) {
  674. goto AnythingElse;
  675. }
  676. if ((token.is_start_tag() && token.tag_name() == HTML::TagNames::head) || token.is_end_tag()) {
  677. log_parse_error();
  678. return;
  679. }
  680. AnythingElse:
  681. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::body));
  682. m_insertion_mode = InsertionMode::InBody;
  683. process_using_the_rules_for(m_insertion_mode, token);
  684. }
  685. void HTMLDocumentParser::generate_implied_end_tags(const FlyString& exception)
  686. {
  687. while (current_node().local_name() != exception && current_node().local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc))
  688. m_stack_of_open_elements.pop();
  689. }
  690. void HTMLDocumentParser::generate_all_implied_end_tags_thoroughly()
  691. {
  692. while (current_node().local_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::colgroup, HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))
  693. m_stack_of_open_elements.pop();
  694. }
  695. void HTMLDocumentParser::close_a_p_element()
  696. {
  697. generate_implied_end_tags(HTML::TagNames::p);
  698. if (current_node().local_name() != HTML::TagNames::p) {
  699. log_parse_error();
  700. }
  701. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::p);
  702. }
  703. void HTMLDocumentParser::handle_after_body(HTMLToken& token)
  704. {
  705. if (token.is_character() && token.is_parser_whitespace()) {
  706. process_using_the_rules_for(InsertionMode::InBody, token);
  707. return;
  708. }
  709. if (token.is_comment()) {
  710. auto data = token.m_comment_or_character.data.to_string();
  711. auto& insertion_location = m_stack_of_open_elements.first();
  712. insertion_location.append_child(adopt_ref(*new DOM::Comment(document(), data)));
  713. return;
  714. }
  715. if (token.is_doctype()) {
  716. log_parse_error();
  717. return;
  718. }
  719. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  720. process_using_the_rules_for(InsertionMode::InBody, token);
  721. return;
  722. }
  723. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  724. if (m_parsing_fragment) {
  725. log_parse_error();
  726. return;
  727. }
  728. m_insertion_mode = InsertionMode::AfterAfterBody;
  729. return;
  730. }
  731. if (token.is_end_of_file()) {
  732. stop_parsing();
  733. return;
  734. }
  735. log_parse_error();
  736. m_insertion_mode = InsertionMode::InBody;
  737. process_using_the_rules_for(InsertionMode::InBody, token);
  738. }
  739. void HTMLDocumentParser::handle_after_after_body(HTMLToken& token)
  740. {
  741. if (token.is_comment()) {
  742. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  743. document().append_child(move(comment));
  744. return;
  745. }
  746. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  747. process_using_the_rules_for(InsertionMode::InBody, token);
  748. return;
  749. }
  750. if (token.is_end_of_file()) {
  751. stop_parsing();
  752. return;
  753. }
  754. log_parse_error();
  755. m_insertion_mode = InsertionMode::InBody;
  756. process_using_the_rules_for(m_insertion_mode, token);
  757. }
  758. void HTMLDocumentParser::reconstruct_the_active_formatting_elements()
  759. {
  760. // FIXME: This needs to care about "markers"
  761. if (m_list_of_active_formatting_elements.is_empty())
  762. return;
  763. if (m_list_of_active_formatting_elements.entries().last().is_marker())
  764. return;
  765. if (m_stack_of_open_elements.contains(*m_list_of_active_formatting_elements.entries().last().element))
  766. return;
  767. ssize_t index = m_list_of_active_formatting_elements.entries().size() - 1;
  768. RefPtr<DOM::Element> entry = m_list_of_active_formatting_elements.entries().at(index).element;
  769. VERIFY(entry);
  770. Rewind:
  771. if (index == 0) {
  772. goto Create;
  773. }
  774. --index;
  775. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  776. VERIFY(entry);
  777. if (!m_stack_of_open_elements.contains(*entry))
  778. goto Rewind;
  779. Advance:
  780. ++index;
  781. entry = m_list_of_active_formatting_elements.entries().at(index).element;
  782. VERIFY(entry);
  783. Create:
  784. // FIXME: Hold on to the real token!
  785. auto new_element = insert_html_element(HTMLToken::make_start_tag(entry->local_name()));
  786. m_list_of_active_formatting_elements.entries().at(index).element = *new_element;
  787. if (index != (ssize_t)m_list_of_active_formatting_elements.entries().size() - 1)
  788. goto Advance;
  789. }
  790. HTMLDocumentParser::AdoptionAgencyAlgorithmOutcome HTMLDocumentParser::run_the_adoption_agency_algorithm(HTMLToken& token)
  791. {
  792. auto subject = token.tag_name();
  793. // If the current node is an HTML element whose tag name is subject,
  794. // and the current node is not in the list of active formatting elements,
  795. // then pop the current node off the stack of open elements, and return.
  796. if (current_node().local_name() == subject && !m_list_of_active_formatting_elements.contains(current_node())) {
  797. m_stack_of_open_elements.pop();
  798. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  799. }
  800. size_t outer_loop_counter = 0;
  801. //OuterLoop:
  802. if (outer_loop_counter >= 8)
  803. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  804. ++outer_loop_counter;
  805. auto formatting_element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(subject);
  806. if (!formatting_element)
  807. return AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps;
  808. if (!m_stack_of_open_elements.contains(*formatting_element)) {
  809. log_parse_error();
  810. m_list_of_active_formatting_elements.remove(*formatting_element);
  811. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  812. }
  813. if (!m_stack_of_open_elements.has_in_scope(*formatting_element)) {
  814. log_parse_error();
  815. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  816. }
  817. if (formatting_element != &current_node()) {
  818. log_parse_error();
  819. }
  820. RefPtr<DOM::Element> furthest_block = m_stack_of_open_elements.topmost_special_node_below(*formatting_element);
  821. if (!furthest_block) {
  822. while (&current_node() != formatting_element)
  823. m_stack_of_open_elements.pop();
  824. m_stack_of_open_elements.pop();
  825. m_list_of_active_formatting_elements.remove(*formatting_element);
  826. return AdoptionAgencyAlgorithmOutcome::DoNothing;
  827. }
  828. // FIXME: Implement the rest of the AAA :^)
  829. TODO();
  830. }
  831. bool HTMLDocumentParser::is_special_tag(const FlyString& tag_name, const FlyString& namespace_)
  832. {
  833. if (namespace_ == Namespace::HTML) {
  834. return tag_name.is_one_of(
  835. HTML::TagNames::address,
  836. HTML::TagNames::applet,
  837. HTML::TagNames::area,
  838. HTML::TagNames::article,
  839. HTML::TagNames::aside,
  840. HTML::TagNames::base,
  841. HTML::TagNames::basefont,
  842. HTML::TagNames::bgsound,
  843. HTML::TagNames::blockquote,
  844. HTML::TagNames::body,
  845. HTML::TagNames::br,
  846. HTML::TagNames::button,
  847. HTML::TagNames::caption,
  848. HTML::TagNames::center,
  849. HTML::TagNames::col,
  850. HTML::TagNames::colgroup,
  851. HTML::TagNames::dd,
  852. HTML::TagNames::details,
  853. HTML::TagNames::dir,
  854. HTML::TagNames::div,
  855. HTML::TagNames::dl,
  856. HTML::TagNames::dt,
  857. HTML::TagNames::embed,
  858. HTML::TagNames::fieldset,
  859. HTML::TagNames::figcaption,
  860. HTML::TagNames::figure,
  861. HTML::TagNames::footer,
  862. HTML::TagNames::form,
  863. HTML::TagNames::frame,
  864. HTML::TagNames::frameset,
  865. HTML::TagNames::h1,
  866. HTML::TagNames::h2,
  867. HTML::TagNames::h3,
  868. HTML::TagNames::h4,
  869. HTML::TagNames::h5,
  870. HTML::TagNames::h6,
  871. HTML::TagNames::head,
  872. HTML::TagNames::header,
  873. HTML::TagNames::hgroup,
  874. HTML::TagNames::hr,
  875. HTML::TagNames::html,
  876. HTML::TagNames::iframe,
  877. HTML::TagNames::img,
  878. HTML::TagNames::input,
  879. HTML::TagNames::keygen,
  880. HTML::TagNames::li,
  881. HTML::TagNames::link,
  882. HTML::TagNames::listing,
  883. HTML::TagNames::main,
  884. HTML::TagNames::marquee,
  885. HTML::TagNames::menu,
  886. HTML::TagNames::meta,
  887. HTML::TagNames::nav,
  888. HTML::TagNames::noembed,
  889. HTML::TagNames::noframes,
  890. HTML::TagNames::noscript,
  891. HTML::TagNames::object,
  892. HTML::TagNames::ol,
  893. HTML::TagNames::p,
  894. HTML::TagNames::param,
  895. HTML::TagNames::plaintext,
  896. HTML::TagNames::pre,
  897. HTML::TagNames::script,
  898. HTML::TagNames::section,
  899. HTML::TagNames::select,
  900. HTML::TagNames::source,
  901. HTML::TagNames::style,
  902. HTML::TagNames::summary,
  903. HTML::TagNames::table,
  904. HTML::TagNames::tbody,
  905. HTML::TagNames::td,
  906. HTML::TagNames::template_,
  907. HTML::TagNames::textarea,
  908. HTML::TagNames::tfoot,
  909. HTML::TagNames::th,
  910. HTML::TagNames::thead,
  911. HTML::TagNames::title,
  912. HTML::TagNames::tr,
  913. HTML::TagNames::track,
  914. HTML::TagNames::ul,
  915. HTML::TagNames::wbr,
  916. HTML::TagNames::xmp);
  917. } else if (namespace_ == Namespace::SVG) {
  918. return tag_name.is_one_of(
  919. SVG::TagNames::desc,
  920. SVG::TagNames::foreignObject,
  921. SVG::TagNames::title);
  922. } else if (namespace_ == Namespace::MathML) {
  923. TODO();
  924. }
  925. return false;
  926. }
  927. void HTMLDocumentParser::handle_in_body(HTMLToken& token)
  928. {
  929. if (token.is_character()) {
  930. if (token.code_point() == 0) {
  931. log_parse_error();
  932. return;
  933. }
  934. if (token.is_parser_whitespace()) {
  935. reconstruct_the_active_formatting_elements();
  936. insert_character(token.code_point());
  937. return;
  938. }
  939. reconstruct_the_active_formatting_elements();
  940. insert_character(token.code_point());
  941. m_frameset_ok = false;
  942. return;
  943. }
  944. if (token.is_comment()) {
  945. insert_comment(token);
  946. return;
  947. }
  948. if (token.is_doctype()) {
  949. log_parse_error();
  950. return;
  951. }
  952. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  953. log_parse_error();
  954. if (m_stack_of_open_elements.contains(HTML::TagNames::template_))
  955. return;
  956. for (auto& attribute : token.m_tag.attributes) {
  957. if (current_node().has_attribute(attribute.local_name_builder.string_view()))
  958. continue;
  959. current_node().set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  960. }
  961. return;
  962. }
  963. 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)) {
  964. process_using_the_rules_for(InsertionMode::InHead, token);
  965. return;
  966. }
  967. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  968. process_using_the_rules_for(InsertionMode::InHead, token);
  969. return;
  970. }
  971. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::body) {
  972. log_parse_error();
  973. if (m_stack_of_open_elements.elements().size() == 1
  974. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body
  975. || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  976. VERIFY(m_parsing_fragment);
  977. return;
  978. }
  979. m_frameset_ok = false;
  980. auto& body_element = m_stack_of_open_elements.elements().at(1);
  981. for (auto& attribute : token.m_tag.attributes) {
  982. if (body_element.has_attribute(attribute.local_name_builder.string_view()))
  983. continue;
  984. body_element.set_attribute(attribute.local_name_builder.to_string(), attribute.value_builder.to_string());
  985. }
  986. return;
  987. }
  988. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  989. log_parse_error();
  990. if (m_stack_of_open_elements.elements().size() == 1
  991. || m_stack_of_open_elements.elements().at(1).local_name() != HTML::TagNames::body) {
  992. VERIFY(m_parsing_fragment);
  993. return;
  994. }
  995. if (!m_frameset_ok)
  996. return;
  997. TODO();
  998. }
  999. if (token.is_end_of_file()) {
  1000. if (!m_stack_of_template_insertion_modes.is_empty()) {
  1001. process_using_the_rules_for(InsertionMode::InTemplate, token);
  1002. return;
  1003. }
  1004. for (auto& node : m_stack_of_open_elements.elements()) {
  1005. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1006. log_parse_error();
  1007. break;
  1008. }
  1009. }
  1010. stop_parsing();
  1011. return;
  1012. }
  1013. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::body) {
  1014. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1015. log_parse_error();
  1016. return;
  1017. }
  1018. for (auto& node : m_stack_of_open_elements.elements()) {
  1019. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1020. log_parse_error();
  1021. break;
  1022. }
  1023. }
  1024. m_insertion_mode = InsertionMode::AfterBody;
  1025. return;
  1026. }
  1027. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  1028. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::body)) {
  1029. log_parse_error();
  1030. return;
  1031. }
  1032. for (auto& node : m_stack_of_open_elements.elements()) {
  1033. if (!node.local_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt, HTML::TagNames::li, HTML::TagNames::optgroup, HTML::TagNames::option, HTML::TagNames::p, HTML::TagNames::rb, HTML::TagNames::rp, HTML::TagNames::rt, HTML::TagNames::rtc, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::body, HTML::TagNames::html)) {
  1034. log_parse_error();
  1035. break;
  1036. }
  1037. }
  1038. m_insertion_mode = InsertionMode::AfterBody;
  1039. process_using_the_rules_for(m_insertion_mode, token);
  1040. return;
  1041. }
  1042. 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)) {
  1043. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1044. close_a_p_element();
  1045. insert_html_element(token);
  1046. return;
  1047. }
  1048. 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)) {
  1049. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1050. close_a_p_element();
  1051. if (current_node().local_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6)) {
  1052. log_parse_error();
  1053. m_stack_of_open_elements.pop();
  1054. }
  1055. insert_html_element(token);
  1056. return;
  1057. }
  1058. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::pre, HTML::TagNames::listing)) {
  1059. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1060. close_a_p_element();
  1061. insert_html_element(token);
  1062. m_frameset_ok = false;
  1063. // If the next token is a U+000A LINE FEED (LF) character token,
  1064. // then ignore that token and move on to the next one.
  1065. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1066. auto next_token = m_tokenizer.next_token();
  1067. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1068. // Ignore it.
  1069. } else {
  1070. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1071. }
  1072. return;
  1073. }
  1074. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1075. if (m_form_element && !m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1076. log_parse_error();
  1077. return;
  1078. }
  1079. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1080. close_a_p_element();
  1081. auto element = insert_html_element(token);
  1082. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_))
  1083. m_form_element = downcast<HTMLFormElement>(*element);
  1084. return;
  1085. }
  1086. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::li) {
  1087. m_frameset_ok = false;
  1088. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1089. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1090. if (node->local_name() == HTML::TagNames::li) {
  1091. generate_implied_end_tags(HTML::TagNames::li);
  1092. if (current_node().local_name() != HTML::TagNames::li) {
  1093. log_parse_error();
  1094. }
  1095. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1096. break;
  1097. }
  1098. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1099. break;
  1100. }
  1101. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1102. close_a_p_element();
  1103. insert_html_element(token);
  1104. return;
  1105. }
  1106. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1107. m_frameset_ok = false;
  1108. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1109. RefPtr<DOM::Element> node = m_stack_of_open_elements.elements()[i];
  1110. if (node->local_name() == HTML::TagNames::dd) {
  1111. generate_implied_end_tags(HTML::TagNames::dd);
  1112. if (current_node().local_name() != HTML::TagNames::dd) {
  1113. log_parse_error();
  1114. }
  1115. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dd);
  1116. break;
  1117. }
  1118. if (node->local_name() == HTML::TagNames::dt) {
  1119. generate_implied_end_tags(HTML::TagNames::dt);
  1120. if (current_node().local_name() != HTML::TagNames::dt) {
  1121. log_parse_error();
  1122. }
  1123. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::dt);
  1124. break;
  1125. }
  1126. if (is_special_tag(node->local_name(), node->namespace_()) && !node->local_name().is_one_of(HTML::TagNames::address, HTML::TagNames::div, HTML::TagNames::p))
  1127. break;
  1128. }
  1129. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1130. close_a_p_element();
  1131. insert_html_element(token);
  1132. return;
  1133. }
  1134. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::plaintext) {
  1135. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1136. close_a_p_element();
  1137. insert_html_element(token);
  1138. m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  1139. return;
  1140. }
  1141. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::button) {
  1142. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::button)) {
  1143. log_parse_error();
  1144. generate_implied_end_tags();
  1145. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::button);
  1146. }
  1147. reconstruct_the_active_formatting_elements();
  1148. insert_html_element(token);
  1149. m_frameset_ok = false;
  1150. return;
  1151. }
  1152. 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)) {
  1153. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1154. log_parse_error();
  1155. return;
  1156. }
  1157. generate_implied_end_tags();
  1158. if (current_node().local_name() != token.tag_name()) {
  1159. log_parse_error();
  1160. }
  1161. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1162. return;
  1163. }
  1164. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::form) {
  1165. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1166. auto node = m_form_element;
  1167. m_form_element = nullptr;
  1168. if (!node || !m_stack_of_open_elements.has_in_scope(*node)) {
  1169. log_parse_error();
  1170. return;
  1171. }
  1172. generate_implied_end_tags();
  1173. if (&current_node() != node) {
  1174. log_parse_error();
  1175. }
  1176. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) { return entry.ptr() == node.ptr(); });
  1177. } else {
  1178. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::form)) {
  1179. log_parse_error();
  1180. return;
  1181. }
  1182. generate_implied_end_tags();
  1183. if (current_node().local_name() != HTML::TagNames::form) {
  1184. log_parse_error();
  1185. }
  1186. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::form);
  1187. }
  1188. return;
  1189. }
  1190. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::p) {
  1191. if (!m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1192. log_parse_error();
  1193. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::p));
  1194. }
  1195. close_a_p_element();
  1196. return;
  1197. }
  1198. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::li) {
  1199. if (!m_stack_of_open_elements.has_in_list_item_scope(HTML::TagNames::li)) {
  1200. log_parse_error();
  1201. return;
  1202. }
  1203. generate_implied_end_tags(HTML::TagNames::li);
  1204. if (current_node().local_name() != HTML::TagNames::li) {
  1205. log_parse_error();
  1206. dbgln("Expected <li> current node, but had <{}>", current_node().local_name());
  1207. }
  1208. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::li);
  1209. return;
  1210. }
  1211. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::dd, HTML::TagNames::dt)) {
  1212. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1213. log_parse_error();
  1214. return;
  1215. }
  1216. generate_implied_end_tags(token.tag_name());
  1217. if (current_node().local_name() != token.tag_name()) {
  1218. log_parse_error();
  1219. }
  1220. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1221. return;
  1222. }
  1223. 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)) {
  1224. if (!m_stack_of_open_elements.has_in_scope(HTML::TagNames::h1)
  1225. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h2)
  1226. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h3)
  1227. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h4)
  1228. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h5)
  1229. && !m_stack_of_open_elements.has_in_scope(HTML::TagNames::h6)) {
  1230. log_parse_error();
  1231. return;
  1232. }
  1233. generate_implied_end_tags();
  1234. if (current_node().local_name() != token.tag_name()) {
  1235. log_parse_error();
  1236. }
  1237. for (;;) {
  1238. auto popped_element = m_stack_of_open_elements.pop();
  1239. if (popped_element->local_name().is_one_of(HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6))
  1240. break;
  1241. }
  1242. return;
  1243. }
  1244. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::a) {
  1245. if (auto* element = m_list_of_active_formatting_elements.last_element_with_tag_name_before_marker(HTML::TagNames::a)) {
  1246. log_parse_error();
  1247. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1248. goto AnyOtherEndTag;
  1249. m_list_of_active_formatting_elements.remove(*element);
  1250. m_stack_of_open_elements.elements().remove_first_matching([&](auto& entry) {
  1251. return entry.ptr() == element;
  1252. });
  1253. }
  1254. reconstruct_the_active_formatting_elements();
  1255. auto element = insert_html_element(token);
  1256. m_list_of_active_formatting_elements.add(*element);
  1257. return;
  1258. }
  1259. 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)) {
  1260. reconstruct_the_active_formatting_elements();
  1261. auto element = insert_html_element(token);
  1262. m_list_of_active_formatting_elements.add(*element);
  1263. return;
  1264. }
  1265. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::nobr) {
  1266. reconstruct_the_active_formatting_elements();
  1267. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::nobr)) {
  1268. log_parse_error();
  1269. run_the_adoption_agency_algorithm(token);
  1270. reconstruct_the_active_formatting_elements();
  1271. }
  1272. auto element = insert_html_element(token);
  1273. m_list_of_active_formatting_elements.add(*element);
  1274. return;
  1275. }
  1276. 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)) {
  1277. if (run_the_adoption_agency_algorithm(token) == AdoptionAgencyAlgorithmOutcome::RunAnyOtherEndTagSteps)
  1278. goto AnyOtherEndTag;
  1279. return;
  1280. }
  1281. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1282. reconstruct_the_active_formatting_elements();
  1283. insert_html_element(token);
  1284. m_list_of_active_formatting_elements.add_marker();
  1285. m_frameset_ok = false;
  1286. return;
  1287. }
  1288. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::applet, HTML::TagNames::marquee, HTML::TagNames::object)) {
  1289. if (!m_stack_of_open_elements.has_in_scope(token.tag_name())) {
  1290. log_parse_error();
  1291. return;
  1292. }
  1293. generate_implied_end_tags();
  1294. if (current_node().local_name() != token.tag_name()) {
  1295. log_parse_error();
  1296. }
  1297. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1298. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1299. return;
  1300. }
  1301. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1302. if (!document().in_quirks_mode()) {
  1303. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1304. close_a_p_element();
  1305. }
  1306. insert_html_element(token);
  1307. m_frameset_ok = false;
  1308. m_insertion_mode = InsertionMode::InTable;
  1309. return;
  1310. }
  1311. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::br) {
  1312. token.drop_attributes();
  1313. goto BRStartTag;
  1314. }
  1315. 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)) {
  1316. BRStartTag:
  1317. reconstruct_the_active_formatting_elements();
  1318. insert_html_element(token);
  1319. m_stack_of_open_elements.pop();
  1320. token.acknowledge_self_closing_flag_if_set();
  1321. m_frameset_ok = false;
  1322. return;
  1323. }
  1324. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1325. reconstruct_the_active_formatting_elements();
  1326. insert_html_element(token);
  1327. m_stack_of_open_elements.pop();
  1328. token.acknowledge_self_closing_flag_if_set();
  1329. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1330. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1331. m_frameset_ok = false;
  1332. }
  1333. return;
  1334. }
  1335. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::param, HTML::TagNames::source, HTML::TagNames::track)) {
  1336. insert_html_element(token);
  1337. m_stack_of_open_elements.pop();
  1338. token.acknowledge_self_closing_flag_if_set();
  1339. return;
  1340. }
  1341. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::hr) {
  1342. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p))
  1343. close_a_p_element();
  1344. insert_html_element(token);
  1345. m_stack_of_open_elements.pop();
  1346. token.acknowledge_self_closing_flag_if_set();
  1347. m_frameset_ok = false;
  1348. return;
  1349. }
  1350. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::image) {
  1351. // Parse error. Change the token's tag name to HTML::TagNames::img and reprocess it. (Don't ask.)
  1352. log_parse_error();
  1353. token.m_tag.tag_name.clear();
  1354. token.m_tag.tag_name.append(HTML::TagNames::img);
  1355. process_using_the_rules_for(m_insertion_mode, token);
  1356. return;
  1357. }
  1358. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::textarea) {
  1359. insert_html_element(token);
  1360. m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  1361. // If the next token is a U+000A LINE FEED (LF) character token,
  1362. // then ignore that token and move on to the next one.
  1363. // (Newlines at the start of pre blocks are ignored as an authoring convenience.)
  1364. auto next_token = m_tokenizer.next_token();
  1365. m_original_insertion_mode = m_insertion_mode;
  1366. m_frameset_ok = false;
  1367. m_insertion_mode = InsertionMode::Text;
  1368. if (next_token.has_value() && next_token.value().is_character() && next_token.value().code_point() == '\n') {
  1369. // Ignore it.
  1370. } else {
  1371. process_using_the_rules_for(m_insertion_mode, next_token.value());
  1372. }
  1373. return;
  1374. }
  1375. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::xmp) {
  1376. if (m_stack_of_open_elements.has_in_button_scope(HTML::TagNames::p)) {
  1377. close_a_p_element();
  1378. }
  1379. reconstruct_the_active_formatting_elements();
  1380. m_frameset_ok = false;
  1381. parse_generic_raw_text_element(token);
  1382. return;
  1383. }
  1384. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::iframe) {
  1385. m_frameset_ok = false;
  1386. parse_generic_raw_text_element(token);
  1387. return;
  1388. }
  1389. if (token.is_start_tag() && ((token.tag_name() == HTML::TagNames::noembed) || (token.tag_name() == HTML::TagNames::noscript && m_scripting_enabled))) {
  1390. parse_generic_raw_text_element(token);
  1391. return;
  1392. }
  1393. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  1394. reconstruct_the_active_formatting_elements();
  1395. insert_html_element(token);
  1396. m_frameset_ok = false;
  1397. switch (m_insertion_mode) {
  1398. case InsertionMode::InTable:
  1399. case InsertionMode::InCaption:
  1400. case InsertionMode::InTableBody:
  1401. case InsertionMode::InRow:
  1402. case InsertionMode::InCell:
  1403. m_insertion_mode = InsertionMode::InSelectInTable;
  1404. break;
  1405. default:
  1406. m_insertion_mode = InsertionMode::InSelect;
  1407. break;
  1408. }
  1409. return;
  1410. }
  1411. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::optgroup, HTML::TagNames::option)) {
  1412. if (current_node().local_name() == HTML::TagNames::option)
  1413. m_stack_of_open_elements.pop();
  1414. reconstruct_the_active_formatting_elements();
  1415. insert_html_element(token);
  1416. return;
  1417. }
  1418. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rb, HTML::TagNames::rtc)) {
  1419. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1420. generate_implied_end_tags();
  1421. if (current_node().local_name() != HTML::TagNames::ruby)
  1422. log_parse_error();
  1423. insert_html_element(token);
  1424. return;
  1425. }
  1426. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::rp, HTML::TagNames::rt)) {
  1427. if (m_stack_of_open_elements.has_in_scope(HTML::TagNames::ruby))
  1428. generate_implied_end_tags(HTML::TagNames::rtc);
  1429. if (current_node().local_name() != HTML::TagNames::rtc || current_node().local_name() != HTML::TagNames::ruby)
  1430. log_parse_error();
  1431. insert_html_element(token);
  1432. return;
  1433. }
  1434. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::math) {
  1435. reconstruct_the_active_formatting_elements();
  1436. adjust_mathml_attributes(token);
  1437. adjust_foreign_attributes(token);
  1438. insert_foreign_element(token, Namespace::MathML);
  1439. if (token.is_self_closing()) {
  1440. m_stack_of_open_elements.pop();
  1441. token.acknowledge_self_closing_flag_if_set();
  1442. }
  1443. return;
  1444. }
  1445. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::svg) {
  1446. reconstruct_the_active_formatting_elements();
  1447. adjust_svg_attributes(token);
  1448. adjust_foreign_attributes(token);
  1449. insert_foreign_element(token, Namespace::SVG);
  1450. if (token.is_self_closing()) {
  1451. m_stack_of_open_elements.pop();
  1452. token.acknowledge_self_closing_flag_if_set();
  1453. }
  1454. return;
  1455. }
  1456. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::frame, HTML::TagNames::head, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))) {
  1457. log_parse_error();
  1458. return;
  1459. }
  1460. // Any other start tag
  1461. if (token.is_start_tag()) {
  1462. reconstruct_the_active_formatting_elements();
  1463. insert_html_element(token);
  1464. return;
  1465. }
  1466. if (token.is_end_tag()) {
  1467. AnyOtherEndTag:
  1468. RefPtr<DOM::Element> node;
  1469. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  1470. node = m_stack_of_open_elements.elements()[i];
  1471. if (node->local_name() == token.tag_name()) {
  1472. generate_implied_end_tags(token.tag_name());
  1473. if (node != current_node()) {
  1474. log_parse_error();
  1475. }
  1476. while (&current_node() != node) {
  1477. m_stack_of_open_elements.pop();
  1478. }
  1479. m_stack_of_open_elements.pop();
  1480. break;
  1481. }
  1482. if (is_special_tag(node->local_name(), node->namespace_())) {
  1483. log_parse_error();
  1484. return;
  1485. }
  1486. }
  1487. return;
  1488. }
  1489. }
  1490. void HTMLDocumentParser::adjust_mathml_attributes(HTMLToken& token)
  1491. {
  1492. token.adjust_attribute_name("definitionurl", "definitionURL");
  1493. }
  1494. void HTMLDocumentParser::adjust_svg_tag_names(HTMLToken& token)
  1495. {
  1496. token.adjust_tag_name("altglyph", "altGlyph");
  1497. token.adjust_tag_name("altglyphdef", "altGlyphDef");
  1498. token.adjust_tag_name("altglyphitem", "altGlyphItem");
  1499. token.adjust_tag_name("animatecolor", "animateColor");
  1500. token.adjust_tag_name("animatemotion", "animateMotion");
  1501. token.adjust_tag_name("animatetransform", "animateTransform");
  1502. token.adjust_tag_name("clippath", "clipPath");
  1503. token.adjust_tag_name("feblend", "feBlend");
  1504. token.adjust_tag_name("fecolormatrix", "feColorMatrix");
  1505. token.adjust_tag_name("fecomponenttransfer", "feComponentTransfer");
  1506. token.adjust_tag_name("fecomposite", "feComposite");
  1507. token.adjust_tag_name("feconvolvematrix", "feConvolveMatrix");
  1508. token.adjust_tag_name("fediffuselighting", "feDiffuseLighting");
  1509. token.adjust_tag_name("fedisplacementmap", "feDisplacementMap");
  1510. token.adjust_tag_name("fedistantlight", "feDistantLight");
  1511. token.adjust_tag_name("fedropshadow", "feDropShadow");
  1512. token.adjust_tag_name("feflood", "feFlood");
  1513. token.adjust_tag_name("fefunca", "feFuncA");
  1514. token.adjust_tag_name("fefuncb", "feFuncB");
  1515. token.adjust_tag_name("fefuncg", "feFuncG");
  1516. token.adjust_tag_name("fefuncr", "feFuncR");
  1517. token.adjust_tag_name("fegaussianblur", "feGaussianBlur");
  1518. token.adjust_tag_name("feimage", "feImage");
  1519. token.adjust_tag_name("femerge", "feMerge");
  1520. token.adjust_tag_name("femergenode", "feMergeNode");
  1521. token.adjust_tag_name("femorphology", "feMorphology");
  1522. token.adjust_tag_name("feoffset", "feOffset");
  1523. token.adjust_tag_name("fepointlight", "fePointLight");
  1524. token.adjust_tag_name("fespecularlighting", "feSpecularLighting");
  1525. token.adjust_tag_name("fespotlight", "feSpotlight");
  1526. token.adjust_tag_name("glyphref", "glyphRef");
  1527. token.adjust_tag_name("lineargradient", "linearGradient");
  1528. token.adjust_tag_name("radialgradient", "radialGradient");
  1529. token.adjust_tag_name("textpath", "textPath");
  1530. }
  1531. void HTMLDocumentParser::adjust_svg_attributes(HTMLToken& token)
  1532. {
  1533. token.adjust_attribute_name("attributename", "attributeName");
  1534. token.adjust_attribute_name("attributetype", "attributeType");
  1535. token.adjust_attribute_name("basefrequency", "baseFrequency");
  1536. token.adjust_attribute_name("baseprofile", "baseProfile");
  1537. token.adjust_attribute_name("calcmode", "calcMode");
  1538. token.adjust_attribute_name("clippathunits", "clipPathUnits");
  1539. token.adjust_attribute_name("diffuseconstant", "diffuseConstant");
  1540. token.adjust_attribute_name("edgemode", "edgeMode");
  1541. token.adjust_attribute_name("filterunits", "filterUnits");
  1542. token.adjust_attribute_name("glyphref", "glyphRef");
  1543. token.adjust_attribute_name("gradienttransform", "gradientTransform");
  1544. token.adjust_attribute_name("gradientunits", "gradientUnits");
  1545. token.adjust_attribute_name("kernelmatrix", "kernelMatrix");
  1546. token.adjust_attribute_name("kernelunitlength", "kernelUnitLength");
  1547. token.adjust_attribute_name("keypoints", "keyPoints");
  1548. token.adjust_attribute_name("keysplines", "keySplines");
  1549. token.adjust_attribute_name("keytimes", "keyTimes");
  1550. token.adjust_attribute_name("lengthadjust", "lengthAdjust");
  1551. token.adjust_attribute_name("limitingconeangle", "limitingConeAngle");
  1552. token.adjust_attribute_name("markerheight", "markerHeight");
  1553. token.adjust_attribute_name("markerunits", "markerUnits");
  1554. token.adjust_attribute_name("markerwidth", "markerWidth");
  1555. token.adjust_attribute_name("maskcontentunits", "maskContentUnits");
  1556. token.adjust_attribute_name("maskunits", "maskUnits");
  1557. token.adjust_attribute_name("numoctaves", "numOctaves");
  1558. token.adjust_attribute_name("pathlength", "pathLength");
  1559. token.adjust_attribute_name("patterncontentunits", "patternContentUnits");
  1560. token.adjust_attribute_name("patterntransform", "patternTransform");
  1561. token.adjust_attribute_name("patternunits", "patternUnits");
  1562. token.adjust_attribute_name("pointsatx", "pointsAtX");
  1563. token.adjust_attribute_name("pointsaty", "pointsAtY");
  1564. token.adjust_attribute_name("pointsatz", "pointsAtZ");
  1565. token.adjust_attribute_name("preservealpha", "preserveAlpha");
  1566. token.adjust_attribute_name("preserveaspectratio", "preserveAspectRatio");
  1567. token.adjust_attribute_name("primitiveunits", "primitiveUnits");
  1568. token.adjust_attribute_name("refx", "refX");
  1569. token.adjust_attribute_name("refy", "refY");
  1570. token.adjust_attribute_name("repeatcount", "repeatCount");
  1571. token.adjust_attribute_name("repeatdur", "repeatDur");
  1572. token.adjust_attribute_name("requiredextensions", "requiredExtensions");
  1573. token.adjust_attribute_name("requiredfeatures", "requiredFeatures");
  1574. token.adjust_attribute_name("specularconstant", "specularConstant");
  1575. token.adjust_attribute_name("specularexponent", "specularExponent");
  1576. token.adjust_attribute_name("spreadmethod", "spreadMethod");
  1577. token.adjust_attribute_name("startoffset", "startOffset");
  1578. token.adjust_attribute_name("stddeviation", "stdDeviation");
  1579. token.adjust_attribute_name("stitchtiles", "stitchTiles");
  1580. token.adjust_attribute_name("surfacescale", "surfaceScale");
  1581. token.adjust_attribute_name("systemlanguage", "systemLanguage");
  1582. token.adjust_attribute_name("tablevalues", "tableValues");
  1583. token.adjust_attribute_name("targetx", "targetX");
  1584. token.adjust_attribute_name("targety", "targetY");
  1585. token.adjust_attribute_name("textlength", "textLength");
  1586. token.adjust_attribute_name("viewbox", "viewBox");
  1587. token.adjust_attribute_name("viewtarget", "viewTarget");
  1588. token.adjust_attribute_name("xchannelselector", "xChannelSelector");
  1589. token.adjust_attribute_name("ychannelselector", "yChannelSelector");
  1590. token.adjust_attribute_name("zoomandpan", "zoomAndPan");
  1591. }
  1592. void HTMLDocumentParser::adjust_foreign_attributes(HTMLToken& token)
  1593. {
  1594. token.adjust_foreign_attribute("xlink:actuate", "xlink", "actuate", Namespace::XLink);
  1595. token.adjust_foreign_attribute("xlink:arcrole", "xlink", "arcrole", Namespace::XLink);
  1596. token.adjust_foreign_attribute("xlink:href", "xlink", "href", Namespace::XLink);
  1597. token.adjust_foreign_attribute("xlink:role", "xlink", "role", Namespace::XLink);
  1598. token.adjust_foreign_attribute("xlink:show", "xlink", "show", Namespace::XLink);
  1599. token.adjust_foreign_attribute("xlink:title", "xlink", "title", Namespace::XLink);
  1600. token.adjust_foreign_attribute("xlink:type", "xlink", "type", Namespace::XLink);
  1601. token.adjust_foreign_attribute("xml:lang", "xml", "lang", Namespace::XML);
  1602. token.adjust_foreign_attribute("xml:space", "xml", "space", Namespace::XML);
  1603. token.adjust_foreign_attribute("xmlns", "", "xmlns", Namespace::XMLNS);
  1604. token.adjust_foreign_attribute("xmlns:xlink", "xmlns", "xlink", Namespace::XMLNS);
  1605. }
  1606. void HTMLDocumentParser::increment_script_nesting_level()
  1607. {
  1608. ++m_script_nesting_level;
  1609. }
  1610. void HTMLDocumentParser::decrement_script_nesting_level()
  1611. {
  1612. VERIFY(m_script_nesting_level);
  1613. --m_script_nesting_level;
  1614. }
  1615. void HTMLDocumentParser::handle_text(HTMLToken& token)
  1616. {
  1617. if (token.is_character()) {
  1618. insert_character(token.code_point());
  1619. return;
  1620. }
  1621. if (token.is_end_of_file()) {
  1622. log_parse_error();
  1623. if (current_node().local_name() == HTML::TagNames::script)
  1624. downcast<HTMLScriptElement>(current_node()).set_already_started({}, true);
  1625. m_stack_of_open_elements.pop();
  1626. m_insertion_mode = m_original_insertion_mode;
  1627. process_using_the_rules_for(m_insertion_mode, token);
  1628. return;
  1629. }
  1630. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::script) {
  1631. // Make sure the <script> element has up-to-date text content before preparing the script.
  1632. flush_character_insertions();
  1633. NonnullRefPtr<HTMLScriptElement> script = downcast<HTMLScriptElement>(current_node());
  1634. m_stack_of_open_elements.pop();
  1635. m_insertion_mode = m_original_insertion_mode;
  1636. // FIXME: Handle tokenizer insertion point stuff here.
  1637. increment_script_nesting_level();
  1638. script->prepare_script({});
  1639. decrement_script_nesting_level();
  1640. if (script_nesting_level() == 0)
  1641. m_parser_pause_flag = false;
  1642. // FIXME: Handle tokenizer insertion point stuff here too.
  1643. while (document().pending_parsing_blocking_script()) {
  1644. if (script_nesting_level() != 0) {
  1645. m_parser_pause_flag = true;
  1646. // FIXME: Abort the processing of any nested invocations of the tokenizer,
  1647. // yielding control back to the caller. (Tokenization will resume when
  1648. // the caller returns to the "outer" tree construction stage.)
  1649. TODO();
  1650. } else {
  1651. auto the_script = document().take_pending_parsing_blocking_script({});
  1652. m_tokenizer.set_blocked(true);
  1653. // FIXME: If the parser's Document has a style sheet that is blocking scripts
  1654. // or the script's "ready to be parser-executed" flag is not set:
  1655. // spin the event loop until the parser's Document has no style sheet
  1656. // that is blocking scripts and the script's "ready to be parser-executed"
  1657. // flag is set.
  1658. if (the_script->failed_to_load())
  1659. return;
  1660. VERIFY(the_script->is_ready_to_be_parser_executed());
  1661. if (m_aborted)
  1662. return;
  1663. m_tokenizer.set_blocked(false);
  1664. // FIXME: Handle tokenizer insertion point stuff here too.
  1665. VERIFY(script_nesting_level() == 0);
  1666. increment_script_nesting_level();
  1667. the_script->execute_script();
  1668. decrement_script_nesting_level();
  1669. VERIFY(script_nesting_level() == 0);
  1670. m_parser_pause_flag = false;
  1671. // FIXME: Handle tokenizer insertion point stuff here too.
  1672. }
  1673. }
  1674. return;
  1675. }
  1676. if (token.is_end_tag()) {
  1677. m_stack_of_open_elements.pop();
  1678. m_insertion_mode = m_original_insertion_mode;
  1679. return;
  1680. }
  1681. TODO();
  1682. }
  1683. void HTMLDocumentParser::clear_the_stack_back_to_a_table_context()
  1684. {
  1685. while (!current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::template_, HTML::TagNames::html))
  1686. m_stack_of_open_elements.pop();
  1687. if (current_node().local_name() == HTML::TagNames::html)
  1688. VERIFY(m_parsing_fragment);
  1689. }
  1690. void HTMLDocumentParser::clear_the_stack_back_to_a_table_row_context()
  1691. {
  1692. while (!current_node().local_name().is_one_of(HTML::TagNames::tr, HTML::TagNames::template_, HTML::TagNames::html))
  1693. m_stack_of_open_elements.pop();
  1694. if (current_node().local_name() == HTML::TagNames::html)
  1695. VERIFY(m_parsing_fragment);
  1696. }
  1697. void HTMLDocumentParser::clear_the_stack_back_to_a_table_body_context()
  1698. {
  1699. while (!current_node().local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::template_, HTML::TagNames::html))
  1700. m_stack_of_open_elements.pop();
  1701. if (current_node().local_name() == HTML::TagNames::html)
  1702. VERIFY(m_parsing_fragment);
  1703. }
  1704. void HTMLDocumentParser::handle_in_row(HTMLToken& token)
  1705. {
  1706. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1707. clear_the_stack_back_to_a_table_row_context();
  1708. insert_html_element(token);
  1709. m_insertion_mode = InsertionMode::InCell;
  1710. m_list_of_active_formatting_elements.add_marker();
  1711. return;
  1712. }
  1713. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::tr) {
  1714. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1715. log_parse_error();
  1716. return;
  1717. }
  1718. clear_the_stack_back_to_a_table_row_context();
  1719. m_stack_of_open_elements.pop();
  1720. m_insertion_mode = InsertionMode::InTableBody;
  1721. return;
  1722. }
  1723. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr))
  1724. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1725. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1726. log_parse_error();
  1727. return;
  1728. }
  1729. clear_the_stack_back_to_a_table_row_context();
  1730. m_stack_of_open_elements.pop();
  1731. m_insertion_mode = InsertionMode::InTableBody;
  1732. process_using_the_rules_for(m_insertion_mode, token);
  1733. return;
  1734. }
  1735. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1736. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1737. log_parse_error();
  1738. return;
  1739. }
  1740. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tr)) {
  1741. return;
  1742. }
  1743. clear_the_stack_back_to_a_table_row_context();
  1744. m_stack_of_open_elements.pop();
  1745. m_insertion_mode = InsertionMode::InTableBody;
  1746. process_using_the_rules_for(m_insertion_mode, token);
  1747. return;
  1748. }
  1749. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::td, HTML::TagNames::th)) {
  1750. log_parse_error();
  1751. return;
  1752. }
  1753. process_using_the_rules_for(InsertionMode::InTable, token);
  1754. }
  1755. void HTMLDocumentParser::close_the_cell()
  1756. {
  1757. generate_implied_end_tags();
  1758. if (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1759. log_parse_error();
  1760. }
  1761. while (!current_node().local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th))
  1762. m_stack_of_open_elements.pop();
  1763. m_stack_of_open_elements.pop();
  1764. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1765. m_insertion_mode = InsertionMode::InRow;
  1766. }
  1767. void HTMLDocumentParser::handle_in_cell(HTMLToken& token)
  1768. {
  1769. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  1770. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1771. log_parse_error();
  1772. return;
  1773. }
  1774. generate_implied_end_tags();
  1775. if (current_node().local_name() != token.tag_name()) {
  1776. log_parse_error();
  1777. }
  1778. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(token.tag_name());
  1779. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  1780. m_insertion_mode = InsertionMode::InRow;
  1781. return;
  1782. }
  1783. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1784. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::td) && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::th)) {
  1785. VERIFY(m_parsing_fragment);
  1786. log_parse_error();
  1787. return;
  1788. }
  1789. close_the_cell();
  1790. process_using_the_rules_for(m_insertion_mode, token);
  1791. return;
  1792. }
  1793. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html)) {
  1794. log_parse_error();
  1795. return;
  1796. }
  1797. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1798. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1799. log_parse_error();
  1800. return;
  1801. }
  1802. close_the_cell();
  1803. // Reprocess the token.
  1804. process_using_the_rules_for(m_insertion_mode, token);
  1805. return;
  1806. }
  1807. process_using_the_rules_for(InsertionMode::InBody, token);
  1808. }
  1809. void HTMLDocumentParser::handle_in_table_text(HTMLToken& token)
  1810. {
  1811. if (token.is_character()) {
  1812. if (token.code_point() == 0) {
  1813. log_parse_error();
  1814. return;
  1815. }
  1816. m_pending_table_character_tokens.append(token);
  1817. return;
  1818. }
  1819. for (auto& pending_token : m_pending_table_character_tokens) {
  1820. VERIFY(pending_token.is_character());
  1821. if (!pending_token.is_parser_whitespace()) {
  1822. // If any of the tokens in the pending table character tokens list
  1823. // are character tokens that are not ASCII whitespace, then this is a parse error:
  1824. // reprocess the character tokens in the pending table character tokens list using
  1825. // the rules given in the "anything else" entry in the "in table" insertion mode.
  1826. log_parse_error();
  1827. m_foster_parenting = true;
  1828. process_using_the_rules_for(InsertionMode::InBody, token);
  1829. m_foster_parenting = false;
  1830. return;
  1831. }
  1832. }
  1833. for (auto& pending_token : m_pending_table_character_tokens) {
  1834. insert_character(pending_token.code_point());
  1835. }
  1836. m_insertion_mode = m_original_insertion_mode;
  1837. process_using_the_rules_for(m_insertion_mode, token);
  1838. }
  1839. void HTMLDocumentParser::handle_in_table_body(HTMLToken& token)
  1840. {
  1841. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  1842. clear_the_stack_back_to_a_table_body_context();
  1843. insert_html_element(token);
  1844. m_insertion_mode = InsertionMode::InRow;
  1845. return;
  1846. }
  1847. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::th, HTML::TagNames::td)) {
  1848. log_parse_error();
  1849. clear_the_stack_back_to_a_table_body_context();
  1850. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tr));
  1851. m_insertion_mode = InsertionMode::InRow;
  1852. process_using_the_rules_for(m_insertion_mode, token);
  1853. return;
  1854. }
  1855. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1856. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name())) {
  1857. log_parse_error();
  1858. return;
  1859. }
  1860. clear_the_stack_back_to_a_table_body_context();
  1861. m_stack_of_open_elements.pop();
  1862. m_insertion_mode = InsertionMode::InTable;
  1863. return;
  1864. }
  1865. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead))
  1866. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  1867. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tbody)
  1868. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::thead)
  1869. && !m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::tfoot)) {
  1870. log_parse_error();
  1871. return;
  1872. }
  1873. clear_the_stack_back_to_a_table_body_context();
  1874. m_stack_of_open_elements.pop();
  1875. m_insertion_mode = InsertionMode::InTable;
  1876. process_using_the_rules_for(InsertionMode::InTable, token);
  1877. return;
  1878. }
  1879. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1880. log_parse_error();
  1881. return;
  1882. }
  1883. process_using_the_rules_for(InsertionMode::InTable, token);
  1884. }
  1885. void HTMLDocumentParser::handle_in_table(HTMLToken& token)
  1886. {
  1887. if (token.is_character() && current_node().local_name().is_one_of(HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1888. m_pending_table_character_tokens.clear();
  1889. m_original_insertion_mode = m_insertion_mode;
  1890. m_insertion_mode = InsertionMode::InTableText;
  1891. process_using_the_rules_for(InsertionMode::InTableText, token);
  1892. return;
  1893. }
  1894. if (token.is_comment()) {
  1895. insert_comment(token);
  1896. return;
  1897. }
  1898. if (token.is_doctype()) {
  1899. log_parse_error();
  1900. return;
  1901. }
  1902. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::caption) {
  1903. clear_the_stack_back_to_a_table_context();
  1904. m_list_of_active_formatting_elements.add_marker();
  1905. insert_html_element(token);
  1906. m_insertion_mode = InsertionMode::InCaption;
  1907. return;
  1908. }
  1909. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  1910. clear_the_stack_back_to_a_table_context();
  1911. insert_html_element(token);
  1912. m_insertion_mode = InsertionMode::InColumnGroup;
  1913. return;
  1914. }
  1915. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  1916. clear_the_stack_back_to_a_table_context();
  1917. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::colgroup));
  1918. m_insertion_mode = InsertionMode::InColumnGroup;
  1919. process_using_the_rules_for(m_insertion_mode, token);
  1920. return;
  1921. }
  1922. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  1923. clear_the_stack_back_to_a_table_context();
  1924. insert_html_element(token);
  1925. m_insertion_mode = InsertionMode::InTableBody;
  1926. return;
  1927. }
  1928. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th, HTML::TagNames::tr)) {
  1929. clear_the_stack_back_to_a_table_context();
  1930. insert_html_element(HTMLToken::make_start_tag(HTML::TagNames::tbody));
  1931. m_insertion_mode = InsertionMode::InTableBody;
  1932. process_using_the_rules_for(m_insertion_mode, token);
  1933. return;
  1934. }
  1935. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::table) {
  1936. log_parse_error();
  1937. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table))
  1938. return;
  1939. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1940. reset_the_insertion_mode_appropriately();
  1941. process_using_the_rules_for(m_insertion_mode, token);
  1942. return;
  1943. }
  1944. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::table) {
  1945. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::table)) {
  1946. log_parse_error();
  1947. return;
  1948. }
  1949. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::table);
  1950. reset_the_insertion_mode_appropriately();
  1951. return;
  1952. }
  1953. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  1954. log_parse_error();
  1955. return;
  1956. }
  1957. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::style, HTML::TagNames::script, HTML::TagNames::template_))
  1958. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_)) {
  1959. process_using_the_rules_for(InsertionMode::InHead, token);
  1960. return;
  1961. }
  1962. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::input) {
  1963. auto type_attribute = token.attribute(HTML::AttributeNames::type);
  1964. if (type_attribute.is_null() || !type_attribute.equals_ignoring_case("hidden")) {
  1965. goto AnythingElse;
  1966. }
  1967. log_parse_error();
  1968. insert_html_element(token);
  1969. // FIXME: Is this the correct interpretation of "Pop that input element off the stack of open elements."?
  1970. // Because this wording is the first time it's seen in the spec.
  1971. // Other times it's worded as: "Immediately pop the current node off the stack of open elements."
  1972. m_stack_of_open_elements.pop();
  1973. token.acknowledge_self_closing_flag_if_set();
  1974. return;
  1975. }
  1976. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::form) {
  1977. log_parse_error();
  1978. if (m_form_element || m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  1979. return;
  1980. }
  1981. m_form_element = downcast<HTMLFormElement>(*insert_html_element(token));
  1982. // FIXME: See previous FIXME, as this is the same situation but for form.
  1983. m_stack_of_open_elements.pop();
  1984. return;
  1985. }
  1986. if (token.is_end_of_file()) {
  1987. process_using_the_rules_for(InsertionMode::InBody, token);
  1988. return;
  1989. }
  1990. AnythingElse:
  1991. log_parse_error();
  1992. m_foster_parenting = true;
  1993. process_using_the_rules_for(InsertionMode::InBody, token);
  1994. m_foster_parenting = false;
  1995. }
  1996. void HTMLDocumentParser::handle_in_select_in_table(HTMLToken& token)
  1997. {
  1998. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::td, HTML::TagNames::th)) {
  1999. log_parse_error();
  2000. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2001. reset_the_insertion_mode_appropriately();
  2002. process_using_the_rules_for(m_insertion_mode, token);
  2003. return;
  2004. }
  2005. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::table, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead, HTML::TagNames::tr, HTML::TagNames::td, HTML::TagNames::th)) {
  2006. log_parse_error();
  2007. if (!m_stack_of_open_elements.has_in_table_scope(token.tag_name()))
  2008. return;
  2009. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2010. reset_the_insertion_mode_appropriately();
  2011. process_using_the_rules_for(m_insertion_mode, token);
  2012. return;
  2013. }
  2014. process_using_the_rules_for(InsertionMode::InSelect, token);
  2015. }
  2016. void HTMLDocumentParser::handle_in_select(HTMLToken& token)
  2017. {
  2018. if (token.is_character()) {
  2019. if (token.code_point() == 0) {
  2020. log_parse_error();
  2021. return;
  2022. }
  2023. insert_character(token.code_point());
  2024. return;
  2025. }
  2026. if (token.is_comment()) {
  2027. insert_comment(token);
  2028. return;
  2029. }
  2030. if (token.is_doctype()) {
  2031. log_parse_error();
  2032. return;
  2033. }
  2034. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2035. process_using_the_rules_for(InsertionMode::InBody, token);
  2036. return;
  2037. }
  2038. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::option) {
  2039. if (current_node().local_name() == HTML::TagNames::option) {
  2040. m_stack_of_open_elements.pop();
  2041. }
  2042. insert_html_element(token);
  2043. return;
  2044. }
  2045. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2046. if (current_node().local_name() == HTML::TagNames::option) {
  2047. m_stack_of_open_elements.pop();
  2048. }
  2049. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2050. m_stack_of_open_elements.pop();
  2051. }
  2052. insert_html_element(token);
  2053. return;
  2054. }
  2055. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::optgroup) {
  2056. if (current_node().local_name() == HTML::TagNames::option && node_before_current_node().local_name() == HTML::TagNames::optgroup)
  2057. m_stack_of_open_elements.pop();
  2058. if (current_node().local_name() == HTML::TagNames::optgroup) {
  2059. m_stack_of_open_elements.pop();
  2060. } else {
  2061. log_parse_error();
  2062. return;
  2063. }
  2064. return;
  2065. }
  2066. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::option) {
  2067. if (current_node().local_name() == HTML::TagNames::option) {
  2068. m_stack_of_open_elements.pop();
  2069. } else {
  2070. log_parse_error();
  2071. return;
  2072. }
  2073. return;
  2074. }
  2075. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::select) {
  2076. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2077. VERIFY(m_parsing_fragment);
  2078. log_parse_error();
  2079. return;
  2080. }
  2081. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2082. reset_the_insertion_mode_appropriately();
  2083. return;
  2084. }
  2085. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::select) {
  2086. log_parse_error();
  2087. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2088. VERIFY(m_parsing_fragment);
  2089. return;
  2090. }
  2091. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2092. reset_the_insertion_mode_appropriately();
  2093. return;
  2094. }
  2095. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::input, HTML::TagNames::keygen, HTML::TagNames::textarea)) {
  2096. log_parse_error();
  2097. if (!m_stack_of_open_elements.has_in_select_scope(HTML::TagNames::select)) {
  2098. VERIFY(m_parsing_fragment);
  2099. return;
  2100. }
  2101. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::select);
  2102. reset_the_insertion_mode_appropriately();
  2103. process_using_the_rules_for(m_insertion_mode, token);
  2104. return;
  2105. }
  2106. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::script, HTML::TagNames::template_)) {
  2107. process_using_the_rules_for(InsertionMode::InHead, token);
  2108. return;
  2109. }
  2110. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2111. process_using_the_rules_for(InsertionMode::InHead, token);
  2112. return;
  2113. }
  2114. if (token.is_end_of_file()) {
  2115. process_using_the_rules_for(InsertionMode::InBody, token);
  2116. return;
  2117. }
  2118. log_parse_error();
  2119. }
  2120. void HTMLDocumentParser::handle_in_caption(HTMLToken& token)
  2121. {
  2122. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::caption) {
  2123. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2124. VERIFY(m_parsing_fragment);
  2125. log_parse_error();
  2126. return;
  2127. }
  2128. generate_implied_end_tags();
  2129. if (current_node().local_name() != HTML::TagNames::caption)
  2130. log_parse_error();
  2131. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2132. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2133. m_insertion_mode = InsertionMode::InTable;
  2134. return;
  2135. }
  2136. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr))
  2137. || (token.is_end_tag() && token.tag_name() == HTML::TagNames::table)) {
  2138. if (!m_stack_of_open_elements.has_in_table_scope(HTML::TagNames::caption)) {
  2139. VERIFY(m_parsing_fragment);
  2140. log_parse_error();
  2141. return;
  2142. }
  2143. generate_implied_end_tags();
  2144. if (current_node().local_name() != HTML::TagNames::caption)
  2145. log_parse_error();
  2146. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::caption);
  2147. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2148. m_insertion_mode = InsertionMode::InTable;
  2149. process_using_the_rules_for(m_insertion_mode, token);
  2150. return;
  2151. }
  2152. if (token.is_end_tag() && token.tag_name().is_one_of(HTML::TagNames::body, HTML::TagNames::col, HTML::TagNames::colgroup, HTML::TagNames::html, HTML::TagNames::tbody, HTML::TagNames::td, HTML::TagNames::tfoot, HTML::TagNames::th, HTML::TagNames::thead, HTML::TagNames::tr)) {
  2153. log_parse_error();
  2154. return;
  2155. }
  2156. process_using_the_rules_for(InsertionMode::InBody, token);
  2157. }
  2158. void HTMLDocumentParser::handle_in_column_group(HTMLToken& token)
  2159. {
  2160. if (token.is_character() && token.is_parser_whitespace()) {
  2161. insert_character(token.code_point());
  2162. return;
  2163. }
  2164. if (token.is_comment()) {
  2165. insert_comment(token);
  2166. return;
  2167. }
  2168. if (token.is_doctype()) {
  2169. log_parse_error();
  2170. return;
  2171. }
  2172. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2173. process_using_the_rules_for(InsertionMode::InBody, token);
  2174. return;
  2175. }
  2176. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2177. insert_html_element(token);
  2178. m_stack_of_open_elements.pop();
  2179. token.acknowledge_self_closing_flag_if_set();
  2180. return;
  2181. }
  2182. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::colgroup) {
  2183. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2184. log_parse_error();
  2185. return;
  2186. }
  2187. m_stack_of_open_elements.pop();
  2188. m_insertion_mode = InsertionMode::InTable;
  2189. return;
  2190. }
  2191. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::col) {
  2192. log_parse_error();
  2193. return;
  2194. }
  2195. if ((token.is_start_tag() || token.is_end_tag()) && token.tag_name() == HTML::TagNames::template_) {
  2196. process_using_the_rules_for(InsertionMode::InHead, token);
  2197. return;
  2198. }
  2199. if (token.is_end_of_file()) {
  2200. process_using_the_rules_for(InsertionMode::InBody, token);
  2201. return;
  2202. }
  2203. if (current_node().local_name() != HTML::TagNames::colgroup) {
  2204. log_parse_error();
  2205. return;
  2206. }
  2207. m_stack_of_open_elements.pop();
  2208. m_insertion_mode = InsertionMode::InTable;
  2209. process_using_the_rules_for(m_insertion_mode, token);
  2210. }
  2211. void HTMLDocumentParser::handle_in_template(HTMLToken& token)
  2212. {
  2213. if (token.is_character() || token.is_comment() || token.is_doctype()) {
  2214. process_using_the_rules_for(InsertionMode::InBody, token);
  2215. return;
  2216. }
  2217. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::base, HTML::TagNames::basefont, HTML::TagNames::bgsound, HTML::TagNames::link, HTML::TagNames::meta, HTML::TagNames::noframes, HTML::TagNames::script, HTML::TagNames::style, HTML::TagNames::template_, HTML::TagNames::title)) {
  2218. process_using_the_rules_for(InsertionMode::InHead, token);
  2219. return;
  2220. }
  2221. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::template_) {
  2222. process_using_the_rules_for(InsertionMode::InHead, token);
  2223. return;
  2224. }
  2225. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::caption, HTML::TagNames::colgroup, HTML::TagNames::tbody, HTML::TagNames::tfoot, HTML::TagNames::thead)) {
  2226. m_stack_of_template_insertion_modes.take_last();
  2227. m_stack_of_template_insertion_modes.append(InsertionMode::InTable);
  2228. m_insertion_mode = InsertionMode::InTable;
  2229. process_using_the_rules_for(m_insertion_mode, token);
  2230. return;
  2231. }
  2232. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::col) {
  2233. m_stack_of_template_insertion_modes.take_last();
  2234. m_stack_of_template_insertion_modes.append(InsertionMode::InColumnGroup);
  2235. m_insertion_mode = InsertionMode::InColumnGroup;
  2236. process_using_the_rules_for(m_insertion_mode, token);
  2237. return;
  2238. }
  2239. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::tr) {
  2240. m_stack_of_template_insertion_modes.take_last();
  2241. m_stack_of_template_insertion_modes.append(InsertionMode::InTableBody);
  2242. m_insertion_mode = InsertionMode::InTableBody;
  2243. process_using_the_rules_for(m_insertion_mode, token);
  2244. return;
  2245. }
  2246. if (token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2247. m_stack_of_template_insertion_modes.take_last();
  2248. m_stack_of_template_insertion_modes.append(InsertionMode::InRow);
  2249. m_insertion_mode = InsertionMode::InRow;
  2250. process_using_the_rules_for(m_insertion_mode, token);
  2251. return;
  2252. }
  2253. if (token.is_start_tag()) {
  2254. m_stack_of_template_insertion_modes.take_last();
  2255. m_stack_of_template_insertion_modes.append(InsertionMode::InBody);
  2256. m_insertion_mode = InsertionMode::InBody;
  2257. process_using_the_rules_for(m_insertion_mode, token);
  2258. return;
  2259. }
  2260. if (token.is_end_tag()) {
  2261. log_parse_error();
  2262. return;
  2263. }
  2264. if (token.is_end_of_file()) {
  2265. if (!m_stack_of_open_elements.contains(HTML::TagNames::template_)) {
  2266. VERIFY(m_parsing_fragment);
  2267. stop_parsing();
  2268. return;
  2269. }
  2270. log_parse_error();
  2271. m_stack_of_open_elements.pop_until_an_element_with_tag_name_has_been_popped(HTML::TagNames::template_);
  2272. m_list_of_active_formatting_elements.clear_up_to_the_last_marker();
  2273. m_stack_of_template_insertion_modes.take_last();
  2274. reset_the_insertion_mode_appropriately();
  2275. process_using_the_rules_for(m_insertion_mode, token);
  2276. }
  2277. }
  2278. void HTMLDocumentParser::handle_in_frameset(HTMLToken& token)
  2279. {
  2280. if (token.is_character() && token.is_parser_whitespace()) {
  2281. insert_character(token.code_point());
  2282. return;
  2283. }
  2284. if (token.is_comment()) {
  2285. insert_comment(token);
  2286. return;
  2287. }
  2288. if (token.is_doctype()) {
  2289. log_parse_error();
  2290. return;
  2291. }
  2292. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2293. process_using_the_rules_for(InsertionMode::InBody, token);
  2294. return;
  2295. }
  2296. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2297. insert_html_element(token);
  2298. return;
  2299. }
  2300. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::frameset) {
  2301. // FIXME: If the current node is the root html element, then this is a parse error; ignore the token. (fragment case)
  2302. m_stack_of_open_elements.pop();
  2303. if (!m_parsing_fragment && current_node().local_name() != HTML::TagNames::frameset) {
  2304. m_insertion_mode = InsertionMode::AfterFrameset;
  2305. }
  2306. return;
  2307. }
  2308. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::frame) {
  2309. insert_html_element(token);
  2310. m_stack_of_open_elements.pop();
  2311. token.acknowledge_self_closing_flag_if_set();
  2312. return;
  2313. }
  2314. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2315. process_using_the_rules_for(InsertionMode::InHead, token);
  2316. return;
  2317. }
  2318. if (token.is_end_of_file()) {
  2319. //FIXME: If the current node is not the root html element, then this is a parse error.
  2320. stop_parsing();
  2321. return;
  2322. }
  2323. log_parse_error();
  2324. }
  2325. void HTMLDocumentParser::handle_after_frameset(HTMLToken& token)
  2326. {
  2327. if (token.is_character() && token.is_parser_whitespace()) {
  2328. insert_character(token.code_point());
  2329. return;
  2330. }
  2331. if (token.is_comment()) {
  2332. insert_comment(token);
  2333. return;
  2334. }
  2335. if (token.is_doctype()) {
  2336. log_parse_error();
  2337. return;
  2338. }
  2339. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::html) {
  2340. process_using_the_rules_for(InsertionMode::InBody, token);
  2341. return;
  2342. }
  2343. if (token.is_end_tag() && token.tag_name() == HTML::TagNames::html) {
  2344. m_insertion_mode = InsertionMode::AfterAfterFrameset;
  2345. return;
  2346. }
  2347. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2348. process_using_the_rules_for(InsertionMode::InHead, token);
  2349. return;
  2350. }
  2351. if (token.is_end_of_file()) {
  2352. stop_parsing();
  2353. return;
  2354. }
  2355. log_parse_error();
  2356. }
  2357. void HTMLDocumentParser::handle_after_after_frameset(HTMLToken& token)
  2358. {
  2359. if (token.is_comment()) {
  2360. auto comment = adopt_ref(*new DOM::Comment(document(), token.m_comment_or_character.data.to_string()));
  2361. document().append_child(move(comment));
  2362. return;
  2363. }
  2364. if (token.is_doctype() || token.is_parser_whitespace() || (token.is_start_tag() && token.tag_name() == HTML::TagNames::html)) {
  2365. process_using_the_rules_for(InsertionMode::InBody, token);
  2366. return;
  2367. }
  2368. if (token.is_end_of_file()) {
  2369. stop_parsing();
  2370. return;
  2371. }
  2372. if (token.is_start_tag() && token.tag_name() == HTML::TagNames::noframes) {
  2373. process_using_the_rules_for(InsertionMode::InHead, token);
  2374. return;
  2375. }
  2376. log_parse_error();
  2377. }
  2378. void HTMLDocumentParser::process_using_the_rules_for_foreign_content(HTMLToken& token)
  2379. {
  2380. if (token.is_character()) {
  2381. if (token.code_point() == 0) {
  2382. log_parse_error();
  2383. insert_character(0xFFFD);
  2384. return;
  2385. }
  2386. if (token.is_parser_whitespace()) {
  2387. insert_character(token.code_point());
  2388. return;
  2389. }
  2390. insert_character(token.code_point());
  2391. m_frameset_ok = false;
  2392. return;
  2393. }
  2394. if (token.is_comment()) {
  2395. insert_comment(token);
  2396. return;
  2397. }
  2398. if (token.is_doctype()) {
  2399. log_parse_error();
  2400. return;
  2401. }
  2402. if ((token.is_start_tag() && token.tag_name().is_one_of(HTML::TagNames::b, HTML::TagNames::big, HTML::TagNames::blockquote, HTML::TagNames::body, HTML::TagNames::br, HTML::TagNames::center, HTML::TagNames::code, HTML::TagNames::dd, HTML::TagNames::div, HTML::TagNames::dl, HTML::TagNames::dt, HTML::TagNames::em, HTML::TagNames::embed, HTML::TagNames::h1, HTML::TagNames::h2, HTML::TagNames::h3, HTML::TagNames::h4, HTML::TagNames::h5, HTML::TagNames::h6, HTML::TagNames::head, HTML::TagNames::hr, HTML::TagNames::i, HTML::TagNames::img, HTML::TagNames::li, HTML::TagNames::listing, HTML::TagNames::menu, HTML::TagNames::meta, HTML::TagNames::nobr, HTML::TagNames::ol, HTML::TagNames::p, HTML::TagNames::pre, HTML::TagNames::ruby, HTML::TagNames::s, HTML::TagNames::small, HTML::TagNames::span, HTML::TagNames::strong, HTML::TagNames::strike, HTML::TagNames::sub, HTML::TagNames::sup, HTML::TagNames::table, HTML::TagNames::tt, HTML::TagNames::u, HTML::TagNames::ul, HTML::TagNames::var))
  2403. || (token.is_start_tag() && token.tag_name() == HTML::TagNames::font && (token.has_attribute(HTML::AttributeNames::color) || token.has_attribute(HTML::AttributeNames::face) || token.has_attribute(HTML::AttributeNames::size)))) {
  2404. log_parse_error();
  2405. if (m_parsing_fragment) {
  2406. goto AnyOtherStartTag;
  2407. }
  2408. TODO();
  2409. }
  2410. if (token.is_start_tag()) {
  2411. AnyOtherStartTag:
  2412. if (adjusted_current_node().namespace_() == Namespace::MathML) {
  2413. adjust_mathml_attributes(token);
  2414. } else if (adjusted_current_node().namespace_() == Namespace::SVG) {
  2415. adjust_svg_tag_names(token);
  2416. adjust_svg_attributes(token);
  2417. }
  2418. adjust_foreign_attributes(token);
  2419. insert_foreign_element(token, adjusted_current_node().namespace_());
  2420. if (token.is_self_closing()) {
  2421. if (token.tag_name() == SVG::TagNames::script && current_node().namespace_() == Namespace::SVG) {
  2422. token.acknowledge_self_closing_flag_if_set();
  2423. goto ScriptEndTag;
  2424. }
  2425. m_stack_of_open_elements.pop();
  2426. token.acknowledge_self_closing_flag_if_set();
  2427. }
  2428. return;
  2429. }
  2430. if (token.is_end_tag() && current_node().namespace_() == Namespace::SVG && current_node().tag_name() == SVG::TagNames::script) {
  2431. ScriptEndTag:
  2432. m_stack_of_open_elements.pop();
  2433. TODO();
  2434. }
  2435. if (token.is_end_tag()) {
  2436. RefPtr<DOM::Element> node = current_node();
  2437. // FIXME: Not sure if this is the correct to_lowercase, as the specification says "to ASCII lowercase"
  2438. if (node->tag_name().to_lowercase() != token.tag_name())
  2439. log_parse_error();
  2440. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2441. if (node == m_stack_of_open_elements.first()) {
  2442. VERIFY(m_parsing_fragment);
  2443. return;
  2444. }
  2445. // FIXME: See the above FIXME
  2446. if (node->tag_name().to_lowercase() == token.tag_name()) {
  2447. while (current_node() != node)
  2448. m_stack_of_open_elements.pop();
  2449. m_stack_of_open_elements.pop();
  2450. return;
  2451. }
  2452. node = m_stack_of_open_elements.elements().at(i - 1);
  2453. if (node->namespace_() != Namespace::HTML)
  2454. continue;
  2455. process_using_the_rules_for(m_insertion_mode, token);
  2456. return;
  2457. }
  2458. }
  2459. VERIFY_NOT_REACHED();
  2460. }
  2461. // https://html.spec.whatwg.org/multipage/parsing.html#reset-the-insertion-mode-appropriately
  2462. void HTMLDocumentParser::reset_the_insertion_mode_appropriately()
  2463. {
  2464. for (ssize_t i = m_stack_of_open_elements.elements().size() - 1; i >= 0; --i) {
  2465. bool last = i == 0;
  2466. // NOTE: When parsing fragments, we substitute the context element for the root of the stack of open elements.
  2467. RefPtr<DOM::Element> node;
  2468. if (last && m_parsing_fragment) {
  2469. node = m_context_element;
  2470. } else {
  2471. node = m_stack_of_open_elements.elements().at(i);
  2472. }
  2473. if (node->local_name() == HTML::TagNames::select) {
  2474. if (!last) {
  2475. for (ssize_t j = i; j > 0; --j) {
  2476. auto& ancestor = m_stack_of_open_elements.elements().at(j - 1);
  2477. if (is<HTMLTemplateElement>(ancestor))
  2478. break;
  2479. if (is<HTMLTableElement>(ancestor)) {
  2480. m_insertion_mode = InsertionMode::InSelectInTable;
  2481. return;
  2482. }
  2483. }
  2484. }
  2485. m_insertion_mode = InsertionMode::InSelect;
  2486. return;
  2487. }
  2488. if (!last && node->local_name().is_one_of(HTML::TagNames::td, HTML::TagNames::th)) {
  2489. m_insertion_mode = InsertionMode::InCell;
  2490. return;
  2491. }
  2492. if (node->local_name() == HTML::TagNames::tr) {
  2493. m_insertion_mode = InsertionMode::InRow;
  2494. return;
  2495. }
  2496. if (node->local_name().is_one_of(HTML::TagNames::tbody, HTML::TagNames::thead, HTML::TagNames::tfoot)) {
  2497. m_insertion_mode = InsertionMode::InTableBody;
  2498. return;
  2499. }
  2500. if (node->local_name() == HTML::TagNames::caption) {
  2501. m_insertion_mode = InsertionMode::InCaption;
  2502. return;
  2503. }
  2504. if (node->local_name() == HTML::TagNames::colgroup) {
  2505. m_insertion_mode = InsertionMode::InColumnGroup;
  2506. return;
  2507. }
  2508. if (node->local_name() == HTML::TagNames::table) {
  2509. m_insertion_mode = InsertionMode::InTable;
  2510. return;
  2511. }
  2512. if (node->local_name() == HTML::TagNames::template_) {
  2513. m_insertion_mode = m_stack_of_template_insertion_modes.last();
  2514. return;
  2515. }
  2516. if (!last && node->local_name() == HTML::TagNames::head) {
  2517. m_insertion_mode = InsertionMode::InHead;
  2518. return;
  2519. }
  2520. if (node->local_name() == HTML::TagNames::body) {
  2521. m_insertion_mode = InsertionMode::InBody;
  2522. return;
  2523. }
  2524. if (node->local_name() == HTML::TagNames::frameset) {
  2525. VERIFY(m_parsing_fragment);
  2526. m_insertion_mode = InsertionMode::InFrameset;
  2527. return;
  2528. }
  2529. if (node->local_name() == HTML::TagNames::html) {
  2530. if (!m_head_element) {
  2531. VERIFY(m_parsing_fragment);
  2532. m_insertion_mode = InsertionMode::BeforeHead;
  2533. return;
  2534. }
  2535. m_insertion_mode = InsertionMode::AfterHead;
  2536. return;
  2537. }
  2538. }
  2539. VERIFY(m_parsing_fragment);
  2540. m_insertion_mode = InsertionMode::InBody;
  2541. }
  2542. const char* HTMLDocumentParser::insertion_mode_name() const
  2543. {
  2544. switch (m_insertion_mode) {
  2545. #define __ENUMERATE_INSERTION_MODE(mode) \
  2546. case InsertionMode::mode: \
  2547. return #mode;
  2548. ENUMERATE_INSERTION_MODES
  2549. #undef __ENUMERATE_INSERTION_MODE
  2550. }
  2551. VERIFY_NOT_REACHED();
  2552. }
  2553. DOM::Document& HTMLDocumentParser::document()
  2554. {
  2555. return *m_document;
  2556. }
  2557. NonnullRefPtrVector<DOM::Node> HTMLDocumentParser::parse_html_fragment(DOM::Element& context_element, const StringView& markup)
  2558. {
  2559. auto temp_document = DOM::Document::create();
  2560. HTMLDocumentParser parser(*temp_document, markup, "utf-8");
  2561. parser.m_context_element = context_element;
  2562. parser.m_parsing_fragment = true;
  2563. parser.document().set_quirks_mode(context_element.document().mode());
  2564. if (context_element.local_name().is_one_of(HTML::TagNames::title, HTML::TagNames::textarea)) {
  2565. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RCDATA);
  2566. } else if (context_element.local_name().is_one_of(HTML::TagNames::style, HTML::TagNames::xmp, HTML::TagNames::iframe, HTML::TagNames::noembed, HTML::TagNames::noframes)) {
  2567. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2568. } else if (context_element.local_name().is_one_of(HTML::TagNames::script)) {
  2569. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::ScriptData);
  2570. } else if (context_element.local_name().is_one_of(HTML::TagNames::noscript)) {
  2571. if (context_element.document().is_scripting_enabled())
  2572. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::RAWTEXT);
  2573. } else if (context_element.local_name().is_one_of(HTML::TagNames::plaintext)) {
  2574. parser.m_tokenizer.switch_to({}, HTMLTokenizer::State::PLAINTEXT);
  2575. }
  2576. auto root = create_element(context_element.document(), HTML::TagNames::html, Namespace::HTML);
  2577. parser.document().append_child(root);
  2578. parser.m_stack_of_open_elements.push(root);
  2579. if (context_element.local_name() == HTML::TagNames::template_) {
  2580. parser.m_stack_of_template_insertion_modes.append(InsertionMode::InTemplate);
  2581. }
  2582. // FIXME: Create a start tag token whose name is the local name of context and whose attributes are the attributes of context.
  2583. parser.reset_the_insertion_mode_appropriately();
  2584. for (auto* form_candidate = &context_element; form_candidate; form_candidate = form_candidate->parent_element()) {
  2585. if (is<HTMLFormElement>(*form_candidate)) {
  2586. parser.m_form_element = downcast<HTMLFormElement>(*form_candidate);
  2587. break;
  2588. }
  2589. }
  2590. parser.run(context_element.document().url());
  2591. NonnullRefPtrVector<DOM::Node> children;
  2592. while (RefPtr<DOM::Node> child = root->first_child()) {
  2593. root->remove_child(*child);
  2594. context_element.document().adopt_node(*child);
  2595. children.append(*child);
  2596. }
  2597. return children;
  2598. }
  2599. }